Adenovirus E1A proteins are closely associated with chromatin in productively infected and transformed cells.

The adenovirus E1A 243R oncoprotein encodes a potent transcription-repression function within the N-terminal 80 amino acids. Our proposed model of E1A repression predicts that E1A interacts with important cellular proteins on chromatin. Consistent with this idea, we report here that E1A proteins from in vivo formaldehyde cross-linked 293 cells are closely associated with chromatin even after several stringent purification steps including double isopycnic CsCl density gradient centrifugation and size exclusion chromatography. Likewise, E1A proteins expressed from virus during productive infection of HeLa cells are closely associated with chromatin starting at early times after infection. No other adenoviral proteins are necessary for E1A 243R protein to associate with chromatin. Analyses of chromatin from HeLa cells infected with adenovirus vectors expressing E1A 243R protein with deletions in different E1A functional domains indicate that sequences within the E1A N-terminal repression domain are needed for the majority of E1A's interactions with chromatin.

Oligomerization properties of the viral oncoproteins adenovirus E1A and human papillomavirus E7 and their complexes with the retinoblastoma protein.

Human papillomavirus 16 E7 (HPV16 E7) and adenovirus 5 E1A (Ad5 E1A) are encoded by highly divergent viruses yet are functionally similar in their ability to bind the retinoblastoma (pRB) tumor suppressor protein, causing the aberrant displacement of E2F trancription factors. The amino acid residues of HPV16 E7 that are necessary for stability, for inhibition of pRB function, and for cell transformation are also necessary for E7 oligomerization. However, neither the specific oligomerization state of HPV16 E7 nor of Ad5 E1A as a function of pRB-binding has been characterized. To gain insight into HPV16 E7 and Ad5 E1A oligomerization properties, sedimentation equilibrium experiments were performed with recombinant HPV16 E7 and Ad5 E1A proteins. These studies reveal that, despite the overall functional similarities between these proteins, monomers, dimers, and tetramers of HPV16 E7 were detected while only reversible monomer-dimer association was identified for Ad5 E1A. The apparent K(d(monomer)-(dimer)) of HPV16 E7 is approximately 100-fold lower than that of a comparable region of Ad5 E1A, and it is concluded that under physiological protein concentrations HPV16 E7 exists primarily as a dimer. Sedimentation equilibrium experiments of pRB/Ad5 E1A and of pRB/HPV16 E7 complexes demonstrate that the tight association of pRB with the viral oncoproteins does not disturb their inherent oligomerization properties. Taken together, this study demonstrates significant differences between the Ad5 E1A and HPV16 E7 oligomerization states that are potentially related to their distinct structures and specific mechanisms of pRB-inactivation.

Regeneration of the binding properties of adenovirus 12 early region 1A proteins after preparation under denaturing conditions.

Adenovirus 12 early region 1A (Ad12 E1A) was expressed in Escherichia coli. Protein was purified in good yield in the presence of 8 M urea and then renatured by dialysis against dilute NH4HCO3 buffer. The affinity of this protein for pRb, C-terminal binding protein (CtBP), TATA binding protein (TBP), and SUG1 was similar to, or greater than, that of Ad12 E1A prepared by immunoaffinity chromatography under nondenaturing conditions. While the binding of the 266- and 235-amino-acid (aa) E1A components to TBP showed similar characteristics the larger E1A protein had a higher affinity for CtBP, pRb, and SUG1. Using nuclear magnetic resonance (NMR) spectroscopy it was shown that structural perturbations occurred in the 266-aa protein in the presence of Zn2+ consistent with binding--no such changes were seen for the 235-aa protein. Limited proteolysis of the 266- and 235-aa E1A proteins gave rise to comparable polypeptide products, suggesting overall similarities in structure. However, the different affinities of the 266- and 235-aa proteins for the partner proteins and the differences seen in the NMR spectra from the two proteins suggested structural differences.

The transactivation domain of adenovirus E1A interacts with the C terminus of human TAF(II)135.

The CR3 activation domain of the human adenovirus E1A protein stimulates transcription by forming protein-protein interactions with DNA sequence-specific binding factors and components of the TFIID complex. Here, we demonstrate that CR3 can complex with the extreme C-terminal 105 amino acids of the human TATA box binding-factor-associated protein, hTAF(II)135. Furthermore, the C-terminal region of hTAF(II)135 can block transcriptional stimulation from an E1A-inducible promoter in vivo. This ability of the C terminus of hTAF(II)135 to bind CR3 and to inhibit E1A-inducible activation is highly specific. These results demonstrate for the first time that a discrete fragment of a mammalian TBP-associated factor which targets a specific activator can impair the stimulation of transcription.

Cyclin-dependent kinases phosphorylate the adenovirus E1A protein, enhancing its ability to bind pRb and disrupt pRb-E2F complexes.

The adenovirus E1A protein of 243 amino acids has been shown to affect a variety of cellular functions, most notably the immortalization of primary cells and the promotion of quiescent cells into S phase. The activity of E1A is derived, in part, from its association with various cellular proteins, many of which play important roles in regulating cell cycle progression. E1A is known to have multiple sites of phosphorylation. It has been suggested that cell cycle-dependent phosphorylation may also control some of E1A's functions. We find now that immune complexes of cyclin-dependent kinases such as cdk4, cdk2, and cdc2 are all capable of phosphorylating E1A in vitro. Additionally, the sites on E1A phosphorylated by these kinases in vitro are similar to the E1A sites phosphorylated in vivo. We have also found that a phosphorylated E1A is far more efficient than an unphosphorylated E1A in associating with pRB and in disrupting E2F/DP-pRB complexes as well. On the basis of our findings and the differences in timing and expression levels of the various cyclins regulating cdks, we suggest that E1A functions at different control points in the cell cycle and that phosphorylation controls, to some extent, its biological functions.

Repression of a matrix metalloprotease gene by E1A correlates with its ability to bind to cell type-specific transcription factor AP-2.

Adenovirus E1A 243-amino acid protein can repress a variety of enhancer -linked viral and cellular promoters. This repression is presumed to be mediated by its interaction with and sequestration of p3OO, a transcriptional coactivator. Type IV 72-kDa collagenase is one of the matrix metalloproteases that has been implicated in differentiation, development, angiogenesis, and tumor metastasis. We show here that the cell type-specific transcription factor AP-2 is an important transcription factor for the activation of the type IV 72-kDa collagenase promoter and that adenovirus E1A 243-amino acid protein represses this promoter by targeting AP-2. Glutathione S-transferase-affinity chromatography studies show that the E1A protein interacts with the DNA binding/dimerization region of AP-2 and that the N-terminal amino acids of E1A protein are required for this interaction. Further, E1A deletion mutants which do not bind to p3OO can repress this collagenase promoter as efficiently as the wildtype E1A protein. Because the AP-2 element is present in a variety of viral and cellular enhancers which are repressed by E1A, these studies suggest that E1A protein can repress cellular and viral promoter/enhancers by forming a complex with cellular transcription factors and that this repression mechanism may be independent of its interaction with p3OO.

E1A promotes association between p300 and pRB in multimeric complexes required for normal biological activity.

The oncogenes of the small DNA tumor viruses encode transforming proteins with multiple domains that influence the cell cycle and aspects of the transformed phenotype. Like other gene products of this type, the adenovirus E1A proteins influence the cell by binding to specific cell growth control proteins. These include members of the retinoblastoma gene product (pRB) family, which are bound by the E1A region 2-specific site, and p300, which is bound at the E1A amino terminus. Binding at these two sites is largely independent, and discrete transcription-regulating functions remain intact in E1A products when only one or the other binding site is functional. In this report, immunoprecipitation with p300 antibodies reveals the presence of the pRB family proteins in p300 complexes when E1A is expressed in host cells, indicating that E1A can mediate physical contact between p300 and the pRB-related proteins. The ability of E1A to induce proliferation efficiently in quiescent primary cells correlates closely with the ability to bind p300 and individual members of the pRB family simultaneously in multimeric complexes, even though the E1A active sites can bind their target proteins efficiently when separated on different molecules. Conservation of a spacer region between the two binding sites that is required for simultaneous binding and efficient induction of proliferation supports the concept that the E1A protein structure has evolved to facilitate simultaneous binding. These results indicate that the E1A proteins are designed not merely to sequester these cellular products, but also to bring them into proximal association with each other in biologically significant complexes.

Identification of distinct roles for separate E1A domains in disruption of E2F complexes.

The adenovirus E1A protein can disrupt protein complexes containing the E2F transcription factor in association with cellular regulatory proteins such as the retinoblastoma gene product (Rb) and the Rb-related p107 protein. Previous experiments have shown that the CR1 and CR2 domains of E1A are required for this activity. We now demonstrate that the CR2 domain is essential for allowing E1A to interact with the E2F-Rb or the E2F-p107-cyclin A-cdk2 complex. Multimeric complexes containing E1A can be detected when the CR1 domain has been rendered inactive by mutation. In addition, the E1A CR1 domain, but not the CR2 domain, is sufficient to prevent the interaction of E2F with Rb or p107. On the basis of these results, we suggest a model whereby the CR2 domain brings E1A to the E2F complexes and then, upon a normal equilibrium dissociation of Rb or p107 from E2F, the E1A CR1 domain is able to block the site of interaction on Rb or p107, thereby preventing the re-formation of the complexes.

TATA-binding protein and the retinoblastoma gene product bind to overlapping epitopes on c-Myc and adenovirus E1A protein.

Using a protein binding assay, we show that the amino-terminal 204 amino acids of the c-Myc protein interact directly with a key component of the basal transcription factor TFIID, the TATA box-binding protein (TBP). Essentially the same region of the c-Myc protein also binds the product of the retinoblastoma gene, the RB protein. c-Myc protein coimmunoprecipitates with TBP in lysates of mammalian cells, demonstrating that the proteins are also complexed in vivo. A short peptide that spans the RB binding site of the E7 protein of human papilloma virus type 16 interferes with the binding of c-Myc to TBP. The same peptide also blocks binding of adenovirus E1A protein to TBP, suggesting that c-Myc and E1A bind to RB and TBP through overlapping epitopes. Furthermore, we show that binding of RB to E1A prevents association of E1A with TBP. Our data suggest that one of the functions of RB and RB-like proteins is to prevent interaction of viral and cellular oncoproteins, such as c-Myc and E1A, with TBP.

The E1A products of oncogenic adenovirus serotype 12 include amino-terminally modified forms able to bind the retinoblastoma protein but not p300.

The cell growth-regulating properties of the adenovirus type 5 (Ad5) E1A oncogene correlate closely with the binding of the E1A products to specific cellular proteins. These proteins include the products of the retinoblastoma tumor susceptibility gene and a 300-kDa product, p300. pRB binds to E1A sequences that are highly conserved among the E1A products of various serotypes, while p300 binding requires sequences in the E1A amino terminus, a region that is not highly conserved. To help evaluate the roles of the E1A-associated proteins in cell growth control, we have compared the p300-binding abilities of the E1A products of Ad5 and of the more oncogenic Ad12 serotype. We show here that despite encoding a sequence that varies somewhat from the p300-binding sequences of Ad5 E1A, the Ad12 E1A products associate with p300 with an affinity similar to that of the Ad5 E1A products. Both the 12S and 13S splice products of Ad12 E1A, like those of Ad5 E1A, encode proteins able to associate with p300. Interestingly, though, both also give rise to prominent forms that are amino terminally modified and unable to associate with p300. This modification, at least in the 13S product, does not appear to diminish the affinity of this product for the retinoblastoma protein.

Isolation of a cDNA encoding the adenovirus E1A enhancer binding protein: a new human member of the ets oncogene family.

The cDNA encoding adenovirus E1A enhancer-binding protein E1A-F was isolated by screening a HeLa cell lambda gt11 expression library for E1A-F site-specific DNA binding. One cDNA clone produced recombinant E1A-F protein with the same DNA binding specificity as that endogenous to HeLa cells. Sequence analysis of the cDNA showed homology with the ETS-domain, a region required for sequence-specific DNA binding and common to all ets oncogene members. Analysis of the longest cDNA revealed about a 94% identity in amino acids between human E1A-F and mouse PEA3 (polyomavirus enhancer activator 3), a recently characterized ets oncogene member. E1A-F was encoded by a 2.5kb mRNA in HeLa cells, which was found to increase during the early period of adenovirus infection. In contrast, ets-2 mRNA was significantly reduced in infected HeLa cells. The results indicate that E1A enhancer binding protein E1A-F is a member of the ets oncogene family and is probably a human homologue of mouse PEA3.

Homologous sequences in adenovirus E1A and human papillomavirus E7 proteins mediate interaction with the same set of cellular proteins.

Studies of adenovirus E1A oncoprotein mutants suggest that the association of E1A with the retinoblastoma protein (pRB) is necessary for E1A-mediated transformation. Mutational analysis of E1A indicates that two regions of pRB are required for E1A to form stable complexes with the retinoblastoma protein. In addition to pRB binding, these regions are necessary for E1A association with several other cellular proteins, including p130, p107, cyclin A, and p33cdk2. Here we show that short synthetic peptides containing the pRB-binding sequences of E1A are sufficient for interaction with p107, cyclin A, and p130. The E7 protein of human papillomavirus type 16 contains an element that binds to pRB and appears to be functionally homologous to the E1A sequences. Peptides containing this region of the E7 protein were able to interact with p107, cyclin A, and p130 in addition to pRB. These findings suggest that the common mechanism of transformation used by these viral oncogenes involves their association with a set of polypeptides.

The adenovirus E1A 243R protein purified from Escherichia coli under nondenaturing conditions is found in association with dnaK.

The adenovirus E1A 243R protein immortalizes primary cells in culture and induces part of the phenotypes required for transformation. It has also been shown to interact with a number of cellular polypeptides, including the product of the retinoblastoma gene. To understand more fully the molecular activities of the E1A 243R protein in association with these proteins as well as its role in the processes of cellular growth, we have developed a method for rapidly purifying this protein from genetically engineered Escherichia coli under nondenaturing conditions. The plasmid-encoded E1A protein, when expressed in a protease-deficient mutant, is found to have the same length and amino acid sequence as that which is produced in a mammalian cell. The procedure for purifying the E1A 243R protein from bacteria relies primarily upon immunoaffinity chromatography and the use of a peptide comprising the epitope recognized by an E1A-specific antibody. Elution of the E1A protein under this condition allows for gentle isolation and a purity that ranges from 90 to 96%. However, without the addition of micromolar amounts of ATP prior to its elution from the antibody column, the E1A protein is found in association with an E. coli protein of 70 kDa. Immunoblot analysis with a specific antibody showed that this bacterial protein was the heat shock protein dnaK, which is known to have extensive homology with the hsp-hsc70 family of proteins in mammalian cells. Recognition of E1A by the dnaK protein may very well reflect a situation that also occurs between the mammalian heat shock proteins and the E1A 243R protein after adenovirus infection.