Characterization of an autoantigen associated with chronic ulcerative stomatitis: the CUSP autoantigen is a member of the p53 family.

A unique clinical syndrome has been described in which patients have chronic oral ulceration and autoantibodies to nuclei of stratified squamous epithelium. We have characterized the autoantibodies from patients sera and found that the major autoantigen is a 70 kDa epithelial nuclear protein. Sequencing of the cDNA for this protein, chronic ulcerative stomatitis protein, revealed it to be homologous to the p53 tumor suppressor and to the p73 putative tumor suppressor, and to be a splicing variant of the KET gene. The p53-like genes, p73 and the several KET splicing variants, are recently described genes of uncertain biologic and pathologic significance. This study provides the first clear association of a p53-like protein with a disease process.

Retinoblastoma protein switches the E2F site from positive to negative element.

Originally E2F sites were identified as elements in the promoters of adenovirus early genes that are necessary for activation of these genes by the early protein E1a (ref. 1). E2F promoter elements have been shown to be important for transcriptional activation of several genes critical for progression through the cell cycle. During the G1 phase of the cell cycle, the E2F protein forms a complex with the cell-cycle protein Rb (ref. 5) and it has been suggested that this binding of Rb to E2F inactivates E2F (ref. 5). Here we show that Rb-E2F is an active complex that, when bound to the E2F site, inhibits the activity of other promoter elements and thus silences transcription. We propose that the ability of this complex to inhibit transcription is integral to the function of Rb and provide evidence that E2F is a positive element in the absence of an active form of Rb. It has been shown that binding of Rb to E2F depends on the phosphorylation state of Rb (only the underphosphorylated form binds) and that the phosphorylation state of Rb changes during progression through the cell cycle. We therefore suggest that the E2F site alternates between a positive and negative element with the phosphorylation/dephosphorylation cycle of Rb. This cyclic activity may be responsible for activating and then inhibiting genes during the cell cycle.

The properdin-like type I repeats of human thrombospondin contain a cell attachment site.

Thrombospondin (TS) is a modular adhesive glycoprotein that contains three domains previously implicated in the attachment of cells to TS. These include the amino-terminal heparin-binding domain, the carboxy terminal cell or platelet-binding domain, and an RGDA sequence of TS. We have characterized a mAb against human TS, designated A4.1, which inhibits the attachment of human melanoma cells (G361) to TS. The epitope for A4.1 lies within the amino terminal half of the central stalklike region of TS which is distinct from the three known cell attachment sites. This region of TS is recovered in a 50-kD peptide after chymotryptic digestion of TS in EDTA. It contains the procollagen-like domain of TS as well as three type I repeats of a 60-residue segment homologous to two malarial proteins and the complement proteins properdin, and factors C6 through C9. The purified chymotryptic fragment is an effective attachment factor for G361 cells. A4.1 blocks adhesion to the 50-kD domain, as do some sulfated glycoconjugates. RGD (and RGE) peptides and mAbs against other domains of TS are not inhibitory. Peptides (19 mers) based on the core homology sequence of the three type I repeats of TS are potent attachment factors for these cells, and this adhesion is also inhibited by sulfated glycoconjugates. A polyclonal antibody raised against one of these peptides inhibits adhesion of G361 cells to the peptides, to the 50-kD fragment and to intact TS. Thus a new cell-adhesion site has been identified in TS whose sequence is very similar to the site identified in region II of the circumsporozoite protein of malaria parasites (Rich, K. A., F. W. George IV, J. L. Law, and W. J. Martin. 1990. Science (Wash. DC) 249:1574-1577. Thus there may be a common receptor which binds TS, malarial proteins, and properdin.