Specific antibodies against the components of desmosomes, the adhesive junctions of epithelial cells, have been used to determine which components are located on the cell surface. Three criteria have been used: fluorescent antibody staining, immuno-gold labelling and electron microscopy, and quantitative measurements of antibody binding using [125I]protein A. When these techniques were applied to living Madin-Darby bovine kidney (MDBK) cells, antibodies against only two desmosomal components, glycoproteins of approximately 115 X 10(3) Mr and 100 X 10(3) Mr, bound to the cell surface. Antibodies against all other components, the 230 and 205 X 10(3) Mr proteins (desmoplakins), the 150 X 10(3) Mr glycoprotein and the 82 and 86 X 10(3) Mr proteins reacted in fluorescent antibody staining only after cells had been fixed and made permeable. MDBK cells were cultured in the presence of univalent fragments (Fab') of anti-desmosomal antibodies for periods from 24 h to 72 h. After these times cells were fixed, made permeable, and stained with anti-desmoplakin antibody to assay for desmosome formation. Fab' derived from anti-100 X 10(3) Mr protein specifically inhibited desmosome formation, whereas Fab's from anti-desmoplakin, anti-150 X 10(3) Mr and anti-82 and 86 X 10(3) Mr proteins were without effect. We conclude that the 100 X 10(3) Mr and the immunologically related 115 X 10(3) Mr components are located on the cell surface and are directly involved in cell-cell adhesion. We have named them desmocollins to denote that they are involved in the adhesive function of desmosomes. The modulation of desmocollin distribution during monolayer formation and establishment of epithelial polarity has also been studied. Fluorescent and immuno-gold labelling using Fab' or IgG at 4 degrees C revealed that desmocollins were initially evenly dispersed over the cell surface. Staining with IgG at 37 degrees C caused the desmocollins to “patch' but not to “cap'. With the establishment of confluency, desmocollins were gradually removed from the upper surfaces of the cells (or masked and rendered inaccessible to antibody) being confined to the lateral and probably basal regions of the cells. Treatment of confluent monolayers with 3 mM-EGTA rendered the desmocollins stainable, probably by causing their release from lateral constraint. Desmocollin staining at the cell surface was not appreciably reduced during 5 h of EGTA treatment, suggesting that desmocollins, unlike desmosomal plaques, may not be internalized after junction breakdown.
In previous work we used immunofluorescent staining with specific antibodies to study the distribution of five desmosomal antigens in the epithelia of different vertebrate animals. We showed that all five antigens were present in all epithelia studied in human, bovine, rat, guinea pig, chick and frog (Rana pipiens) tissues. It was concluded that desmosomes are highly conserved structures. This paper extends those studies: by including three other species, a lizard (Lacerta viridis), the axolotl (Ambystoma mexicanum) and the trout (Salmo trutta), and by looking at several tissues in more detail. The principal results are as follows. The epidermis of all species down to the frog stain with equal intensity for all desmosomal antigens. In the epidermis of axolotl and trout, staining for desmosomal plaque constituents is present, but staining for the desmosomal glycoproteins is greatly reduced or absent. Within mammalian species as well as chick, lizard and frog, staining for the 115 X 10(3) and 100 X 10(3) molecular weight desmosomal glycoproteins is less intense in non-epidermal tissues than in the epidermis, while staining for desmosomal plaque constituents and for the 150 X 10(3) molecular weight glycoprotein is undiminished. It is possible, therefore, that slight differences exist between certain glycoproteins of epidermis and non-epidermal epithelia. The hearts of lower vertebrates (lizard, frog, axolotl and trout) stain only for individual desmosomal plaque antigens. The pillar cells of trout gill stain, adjacent to their collagenous columns, for one desmosomal plaque antigen. There is a fibrous cytoplasmic mat in this position but no desmosomes. Thus one of the desmosomal antigens may have a function outside the desmosome.