Characterization of the retinoblastoma binding proteins RBP1 and RBP2.

The retinoblastoma gene product, pRB, regulates cell proliferation by binding to and inhibiting the activity of key growth promoting proteins. Several cellular proteins have been shown to bind directly to pRB and the genes encoding a number of them have been isolated. The protein product of one of these genes is the transcription factor E2F. We have now isolated cDNA clones that contain the full-length coding sequence of two other proteins, RBP1 and RBP2, cloned originally by their interaction with pRB. The products of the RBP1 and RBP2 genes are ubiquitously expressed, large (200 kDa for RBP1 and 195 kDa for RBP2) nuclear phosphoproteins with structural motifs that suggest a role in transcriptional regulation. In addition we have been able to identify complexes of pRB and RBP1 in vivo that are dissociated in the presence of purified human papillomavirus E7 protein.

Papillomavirus E7 protein binding to the retinoblastoma protein is not required for viral induction of warts.

Human papillomaviruses (HPVs) are the etiologic agents responsible for benign epithelial proliferative disorders including genital warts and are a contributory factor in the pathogenesis of cervical cancer. HPVs demonstrate strict species and cell-type specificity, which is manifested by the inability of these viruses to induce disease in any species other than humans. The natural history of HPV infection in humans is closely mimicked by cottontail rabbit papillomavirus (CRPV) infection in domestic laboratory rabbits. The CRPV E7 gene is known to play an essential role in virus-mediated induction of papillomas. We now show by mutational analysis that the CRPV E7 protein's biochemical and biological properties, including binding to the retinoblastoma suppressor protein (pRB), transcription factor E2F transactivation of the adenovirus E2 promoter, disruption of pRB-E2F complexes, and cellular transformation as measured by growth in soft agar, mimic those of the HPV E7 protein. Intradermal injection of CRPV DNA lacking E7 gene sequences critical for the binding of the CRPV E7 protein to pRB induced papillomas in rabbits. These studies indicate that E7 protein binding to pRB is not required in the molecular pathogenesis of virally induced warts and suggest that other properties intrinsic to the E7 protein are necessary for papilloma formation.

Comparison of the binding of the human papillomavirus type 16 and cottontail rabbit papillomavirus E7 proteins to the retinoblastoma gene product.

Binding of the human papillomavirus type 16 (HPV-16) E7 oncoprotein to the retinoblastoma protein (pRb) is thought to be involved in the cellular transformation mediated by HPV-16. Here we show that the E7 protein of the cottontail rabbit papillomavirus (CRPV) binds to the same C-terminal portion of human pRb as HPV-16 E7, and that both the CRPV and HPV-16 E7 proteins bind specifically through similar domains to rabbit pRb. Furthermore, a single amino acid substitution which reduces the binding of HPV-16 E7 to human pRb also abolishes binding of CRPV E7 to both human and rabbit pRb. The biochemical similarities observed between the HPV-16 and CRPV E7 proteins suggest that they are functionally conserved. These results further validate the use of CRPV as an animal model for the study of HPV-mediated disease.

Cloning of cDNAs for cellular proteins that bind to the retinoblastoma gene product.

The E7 transforming protein of human papilloma virus-16 binds to the retinoblastoma gene product (pRb) through a nine-amino-acid segment of E7 (21-29). This segment of E7 is homologous to the pRb-binding domains of the simian virus 40 large T and adenovirus E1A transforming proteins. Each of these viral transforming proteins bind to the same region of pRb. To isolate cellular proteins that interact with this viral protein-binding domain on pRb, we used recombinant pRb to screen a human complementary DNA expression library. Two cDNAs were isolated that encode retinoblastoma binding proteins (RBP-1 and RBP-2). We report here that these RBP genes exist in separate loci and produce discrete messenger RNAs. The predicted amino-acid sequence of these genes showed no homology to known proteins, but both RBPs contain the pRb binding motif conserved between E7, large T and E1A14. In vitro expression of the RBP cDNAs yielded proteins that specifically bound to pRb. Recombinant E7 protein, the E7 21-29 peptide and the homologous RBP-1 peptide inhibited RBP-pRb binding. Mutations introduced into the putative pRb-binding segment in RBP-1 impaired its binding activity. These studies indicate that the cellular RBP-1, RBP-2 and viral E7 proteins interact with pRb through similar domains.

Mammalian cell lines engineered to identify inhibitors of specific signal transduction pathways.

A variety of signal transduction pathways contribute to the regulation of transcription in mammalian cells. Several of these pathways ultimately rely upon the interaction of transcription factors with genetic sequences termed response elements in the promoter regions of some genes. The biochemical mechanisms that control the levels and state of activation of transcription factors are poorly understood. However, specific phosphorylation events mediated by protein kinase C, growth factor receptor-linked tyrosine kinases, and protein kinase A clearly participate in the regulation of these signal transduction pathways. To understand the relationship between activation and/or inhibition of these pathways and regulation of gene expression controlled by specific response elements, cell lines were prepared containing the TPA response element (TRE), serum response element (SRE), or cyclic AMP response element (CRE) fused to a gene encoding a secretable form of alkaline phosphatase (SEAP). These TRE-SEAP, SRE-SEAP, and CRE-SEAP cells exhibit dramatic increases in alkaline phosphatase (AP) activity following exposure to TPA, PDGF, or forskolin. Down regulation of protein kinase C or inhibition of tyrosine kinase activity blocked the stimulation of AP activity caused by TPA or PDGF. These cell lines can be used to characterize existing inhibitors, and to identify new agents that affect specific signal transduction pathways in mammalian cells.

Lovastatin selectively inhibits ras activation of the 12-O-tetradecanoylphorbol-13-acetate response element in mammalian cells.

To evaluate ras-mediated signal transduction, an alkaline phosphatase gene (SEAP) was placed under the control of the ras-inducible phorbol ester response element (TRE) in murine fibroblasts (TRE-SEAP cells). The Kirsten ras gene was placed under the control of the glucocorticoid-inducible mouse mammary tumor virus promoter and introduced into the TRE-SEAP cells. Dexamethasone increased ras expression in the TRE-SEAP cells carrying the Kirsten ras gene and stimulated SEAP activity 25-fold. Lavostatin blocked dexamethasone induction of SEAP activity (50% inhibitory concentration, 0.5 microM) but did not affect phorbol ester-induced SEAP activity in the same cells. Lovastatin also did not block forskolin induction of SEAP activity in cells expressing SEAP under the control of the cyclic AMP response element.

Genetic regulation of beta 2-adrenergic receptors in 3T3-L1 fibroblasts.

The beta 2-adrenergic receptor from mouse 3T3-L1 cells is up-regulated through genetic mechanisms by glucocorticoids and butyrate. To study the genetic regulation of these receptors, we sequenced a 5 kb region of genomic DNA from 3T3-L1 cells, containing the beta-adrenergic receptor gene and approx. 1.5 kb of both 5' and 3' flanking sequences. The sequence contained one copy of an 8 bp consensus sequence which can confer phorbol ester-responsiveness to genes. Phorbol esters attenuated the up-regulation of beta 2-adrenergic receptors by glucocorticoids but not by butyrate. This effect was probably due to a phorbol ester-induced decrease in glucocorticoid receptor number. Using methylation-sensitive restriction enzymes, we examined the methylation of a CG-rich region occurring 5' to the gene and did not detect any changes in methylation of this region upon dexamethasone or butyrate treatment. A total of 16 putative glucocorticoid response elements were found which may mediate the glucocorticoid-induced increase in beta 2-adrenergic receptors. A comparison of the regulatory sequences of the two beta-adrenergic receptor subtypes from human and mouse confirms the observed physiological controls of receptor subtype expression and offers an explanation as to why the subtypes differ in genetic regulation.