Generation of a new adenovirus type 12-inducible fragile site by insertion of an artificial U2 locus in the human genome.

Infection with adenovirus type 12 (Ad12) induces four fragile sites in the human genome (H.F. Stich, G.L. van Hoosier, and J.J. Trentin, Exp. Cell Res. 34:400-403, 1964; H. zur Hausen, J. Virol. 1:1174-1185, 1967). The major site, at 17q21-22, contains the U2 gene cluster, which is specifically disrupted by infection in at least a percentage of the cells (D.M. Durnam, J.C. Menninger, S.H. Chandler, P.P. Smith, and J.K. McDougall, Mol. Cell. Biol. 8:1863-1867, 1988). For direct assessment of whether the U2 locus is the target of the Ad12 effect, an artificial locus, constructed in vitro and consisting of tandem arrays of the U2 6-kbp monomer, was transfected into human cells. We report that integration of this artificial locus on the p arm of chromosome 13 creates a new Ad12-inducible fragile site.     

Analysis of adenovirus transforming proteins from early regions 1A and 1B with antisera to inducible fusion antigens produced in Escherichia coli.

Plasmid vectors were constructed which expressed three adenovirus tumor antigens fused to a portion of the trpE protein of Escherichia coli. Insertion of adenovirus type 2 DNA from early region 1A (E1A) into such a plasmid led to a fusion protein which contained the C-terminal 266 amino acids of the 289-amino acid protein encoded by the viral 13S mRNA. Similarly, insertion of adenovirus type 5 DNA corresponding to the E1B 55- and 21-kilodalton proteins led to production of fusion proteins containing amino acid sequences from these proteins. After induction with indoleacrylic acid, fusion proteins accumulated stably in the E. coli cells. By using a simple extraction of insoluble protein, 1 to 10 mg of fusion protein per liter of culture was obtained. The fusion proteins were purified on preparative polyacrylamide gels and used to immunize rabbits. Specific antisera for the E1A 289- and closely related 243-amino acid proteins and the E1B 55- and 21-kilodalton proteins were obtained. These sera were used to immunoprecipitate the tumor antigens in cells infected with wild-type and various mutants of adenovirus or to analyze them by an immunoblotting procedure. Mutant E1A proteins in which the C-terminal 70 amino acids are deleted were phosphorylated to much lower extents than the wild-type E1A proteins. This indicates that the deleted region is important for the process of phosphorylation. The E1A proteins were extracted, sedimented in glycerol gradients, analyzed by immunoprecipitation, and found to sediment primarily as monomers.     

Adenovirus E1B 19-kilodalton protein overcomes the cytotoxicity of E1A proteins.

Infection with adenovirus mutants carrying either point mutations or deletions in the coding region for the 19-kDa E1B gene product (19K protein) causes degradation of host cell and viral DNAs (deg phenotype) and enhanced cytopathic effect (cyt phenotype). Therefore, one function of the E1B 19K protein is to protect nuclear DNA integrity and preserve cytoplasmic architecture during productive adenovirus infection. When placed in the background of a virus incapable of expressing a functional E1A gene product, however, E1B 19K gene mutations do not result in the appearance of the cyt and deg phenotypes. This demonstrated that expression of the E1A proteins was responsible for inducing the appearance of the cyt and deg phenotypes. By constructing a panel of viruses possessing E1A mutations spanning each of the three E1A conserved regions in conjunction with E1B 19K gene mutations, we mapped the induction of the cyt and deg phenotypes to the amino-terminal region of E1A. Viruses that fail to express conserved region 3 (amino acids 140 to 185) and/or 2, (amino acids 121 to 185) or nonconserved sequences between conserved regions 2 and 1 of E1A (amino acids 86 to 120) were still capable of inducing cyt and deg. This indicated that activities associated with these regions, such as transactivation and binding to the product of the retinoblastoma susceptibility gene, were dispensable for induction of E1A-dependent cytotoxic effects. In contrast, deletion of sequences in the amino terminus of E1A (amino acids 22 to 107) resulted in extragenic suppression of the cyt and deg phenotypes. Therefore, a function affected by deletion of amino acids 22 to 86 of E1A is responsible for exerting cytotoxic effects in virally infected cells. Furthermore, transient high-level expression of the E1A region using a cytomegalovirus promoter plasmid expression vector was sufficient to induce the cyt and deg phenotypes, demonstrating that E1A expression alone is sufficient to exert these cytotoxic effects and that other viral gene products are not involved. Finally, placing E1A expression under the control of a strong promoter did not alter the requirement for E1B in the transformation of primary cells. One possibility is that the E1B 19K protein is required to overcome the cytotoxic effects of E1A protein expression and thereby enable primary cells to become transformed.     

Induction of apoptosis by human Nbk/Bik, a BH3-containing protein that interacts with E1B 19K.

The E1B 19-kilodalton protein (19K protein) is a potent apoptosis inhibitor and the adenovirus homolog of Bcl-2 (E. White, Genes Dev. 10:1-15, 1996). To obtain a better understanding of the biochemical mechanism by which the E1B 19K protein regulates apoptosis, proteins that interact with 19K have been identified; one of these is Bax (J. Han, P. Sabbatini, D. Perez, L. Rao, D. Mohda, and E. White, Genes Dev. 10:461-477, 1996), and another is Bak (S. N. Farrow, J. H. M. White, I. Martinou, T. Raven, K.-T. Pun, C. J. Grinham, J.-C. Martinou, and R. Brown, Nature (London) 374:731-733, 1995). Bax and Bak are Bcl-2 family members which contain Bcl-2 homology regions 1, 2, and 3 (BH1, BH2, and BH3), which interact with E1B 19K and Bcl-2 and promote apoptosis. Like Bax and Bak, Nbk was cloned from a yeast two-hybrid screen for proteins that interact with E1B 19K. Nbk contained BH3 but not BH1 or BH2. It also interacted with Bcl-2 but not with Bax. Both Bcl-2 and E1B 19K interacted with Nbk in vitro, and this interaction was highly specific. In vivo, the Nbk and E1B 19K proteins may colocalize with cytoplasmic and nuclear membranes. Nbk expression functionally antagonized 19K-mediated inhibition of apoptotic cell death and completely prevented transformation by E1A and E1B 19K. Nbk was sufficient for induction of apoptosis in the presence of mutant p53 and thus low levels of Bax, suggesting that Nbk functions independently of Bax to induce apoptosis. Nbk may therefore represent a novel death regulator which contains only a BH3 that interacts with and antagonizes apoptosis inhibitors such as the E1B 19K protein.     

Intranuclear location of the adenovirus type 5 E1B 55-kilodalton protein.

The intracellular location of the adenovirus type 5 E1B 55-kilodalton (kDa) protein, particularly the question of whether it is associated with nuclear pore complexes, was examined. Fractionation of adenovirus type 5-infected HeLa cell nuclei by an established procedure (N. Dwyer and G. Blobel, J. Cell. Biol. 70:581-591, 1976) yielded one population of E1B 55-kDa protein molecules released by digestion of nuclei with RNase A and a second population recovered in the pore complex-lamina fraction. Free and E1B 55-kDa protein-bound forms of the E4 34-kDa protein (P. Sarnow, C. A. Sullivan, and A. J. Levine, Virology 120:387-394, 1982) were largely recovered in the pore complex-lamina fraction. Nevertheless, the association of E1B 55-kDa protein molecules with this nuclear envelope fraction did not depend on interaction of the E1B 55-kDa protein with the E4 34-kDa protein. Comparison of the immunofluorescence patterns observed with antibodies recognizing the E1B 55-kDa protein or cellular pore complex proteins and of the behavior of these viral and cellular proteins during in situ fractionation suggests that the E1B 55-kDa protein does not become intimately or stably associated with pore complexes in adenovirus-infected cells.     

Adenovirus type 5 early region 4 is responsible for E1A-induced p53-independent apoptosis.

In the absence of E1B, the 289- and 243-residue E1A products of human adenovirus type 5 induce p53-dependent apoptosis. However, our group has shown recently that the 289-residue E1A protein is also able to induce apoptosis by a p53-independent mechanism (J. G. Teodoro, G. C. Shore, and P. E. Branton, Oncogene 11:467-474, 1995). Preliminary results suggested that p53-independent cell death required expression of one or more additional adenovirus early gene products. Here we show that both the E1B 19-kDa protein and cellular Bcl-2 inhibit or significantly delay p53-independent apoptosis. Neither early region E2 or E3 appeared to be necessary for such cell death. Analysis of a series of E1A mutants indicated that mutations in the transactivation domain and other regions of E1A correlated with E1A-mediated transactivation of E4 gene expression. Furthermore, p53-deficient human SAOS-2 cells infected with a mutant which expresses E1B but none of the E4 gene products remained viable for considerably longer times than those infected with wild-type adenovirus type 5. In addition, an adenovirus vector lacking both E1 and E4 was unable to induce DNA degradation and cell killing in E1A-expressing cell lines. These data showed that an E4 product is essential for E1A-induced p53-independent apoptosis.     

Plasmid-directed synthesis of genuine adenovirus 2 early-region 1A and 1B proteins in Escherichia coli.

The transforming region of human adenovirus 2 is located in the left 11.2% of the viral genome and is comprised of two distinct genetic units termed E1A and E1B. cDNAs containing the entire nucleotide sequence of the mature E1A 13S and E1B 22S mRNAs that are complementary to these genetic units have been introduced into bacterial plasmids a short distance downstream from the Escherichia coli lac promoter. Upon transformation into appropriate E. coli hosts, one of these plasmids, pKHAO, directed the synthesis of a 45-kilodalton (kd) protein, and the other, pKHBO, synthesized a protein of 54.9 kd. Both of these plasmid-encoded proteins constituted 0.1 to 0.3% of the total cellular protein and were virtually identical to the authentic adenovirus 2 E1A 42- to 50-kd and E1B 53- to 58-kd tumor antigens (T antigen) as determined by gel electrophoresis, immunoprecipitation, and tryptic fingerprint analysis. With the use of our pKHBO expression plasmid we were also able to demonstrate that the second AUG sequence appearing in the E1B 22S mRNA corresponded to the start of the gene encoding the large adenovirus 2 T antigen. This confirms theoretical deductions based on DNA sequencing analysis that translation of the large T antigen initiates translation at an internal ATG rather than at the 5'-proximal AUG.     

Adenovirus type 12 E1B 19-kilodalton protein is not required for oncogenic transformation in rats.

The adenovirus type 12 mutants in700 and pm700 carry site-specific mutations within the reading frame encoding the E1B 19-kilodalton protein (19K protein) which prevent the production of the intact 19K protein. In cultures of human A549 cells, these mutants grow just as well as the wild-type virus does, but they display a large-plaque (lp), cytocidal (cyt) phenotype. DNA in these infected cells is not degraded, but at late times in human KB cells infected by the mutants, the mutants display a DNA degradation (deg) phenotype. The transformation phenotype of these mutants is also host range. Although the mutants are defective for transformation of the 3Y1 rat cell line, they transform rat and mouse primary kidney cells in vitro at wild-type efficiency and are capable of inducing tumors in rats. These results support the view that the type 12 E1B 19K protein is not obligatory for oncogenic transformation.     

Constructing chimeric type 12/type 5 adenovirus E1A genes and using them to identify an oncogenic determinant of adenovirus type 12.

The E1A gene of highly oncogenic type 12 adenovirus (Ad12) possesses a segment unique to this serotype and comprising 60 base pairs contiguous with and separating conserved regions 2 and 3 in the gene. A similar but slightly longer segment is also present in the E1A gene of highly oncogenic simian adenovirus type 7 (D. Kimelman, J. S. Miller, D. Porter, and B. E. Roberts, J. Virol. 53:399-409, 1985). This segment is missing entirely from the E1A gene of type 5 adenovirus, which is nononcogenic. To test the hypothesis that this unique separating or "spacer" region influences the oncogenicity of Ad12, we constructed ClaI and SmaI restriction sites on either side of it, which allowed reciprocal exchange between this and the equivalent cassette from type 5 adenovirus E1A, bounded by the same restriction sites intrinsic to that gene. The resultant Ad12-based chimeric viruses, ch702 and ch704, in which the spacer region is replaced with (in-frame) type 5 sequence, grow normally on human A549 cells and display wild-type transformation frequencies on baby rat and mouse kidney cells. In contrast, the oncogenic capacity of these chimeric viruses, as measured by tumor induction following virus inoculation in Hooded Lister rats, is greatly reduced. Likewise, cells transformed by ch702 and ch704 display reduced tumorigenicity compared with wild-type transformants in syngeneic rats. These results, coupled with recent preliminary tests using a mutant with a point mutation in this region, support the view that the unique spacer region of type 12 is an oncogenic determinant of this virus.     

Heritable fragile sites on human chromosomes. XI. Factors affecting expression of fragile sites at 10q25, 16q22, and 17p12.

The fragile sites at 10q25, 16q22, and 17p12 can all be induced in lymphocyte culture by BrdU or BrdC added 6-12 hrs prior to harvest. Without induction, fra(10)(q25) is rarely expressed spontaneously, whereas fra(16)(q22) is frequently expressed spontaneously. Fra(17)(p12) is frequently expressed spontaneously but is probably expressed only after induction in some individuals. Distamycin A, netropsin, and Hoechst 33258 induced high levels of expression of fra(16)(q22) and fra(17)(p12) but did not enhance expression of fra(10)(q25). The mechanisms of induction of fra(16)(q22) by BrdU and distamycin A appear to be different, since the time of induction by BrdU reaches a maximum about 12 hrs prior to harvest whereas induction by distamycin A requires much longer exposure. The fragile sites at 10q25 and 16q22 were both induced in fibroblast culture by BrdU. Fra(17)(p12) is accepted as a fragile site because preliminary studies show that it behaves similarly in lymphocyte culture to fra(16)(q22); however, there is only limited evidence for fragility at 17p12.     

Induction of the cell cycle in baby rat kidney cells by adenovirus type 5 E1A in the absence of E1B and a possible influence of p53.

From previous studies on the induction of DNA synthesis in quiescent primary baby rat kidney cells by adenovirus type 5 (Ad5) E1A deletion mutants, we concluded that induction is prevented only when cellular proteins p300 and pRb are both uncomplexed with E1A (J.A. Howe, J.S. Mymryk, C. Egan, P.E. Branton, and S.T. Bayley, Proc. Natl. Acad. Sci. USA 87:5883-5887, 1990). We have now examined induction by these same mutants in virus lacking the E1B region, so that cellular p53 was no longer complexed to the E1B 55-kDa protein. E1A mutants that fail to bind pRb induced DNA synthesis at a significantly lower level in Ad5 lacking E1B than in Ad5 containing E1B. Apparently, therefore, uncomplexed p53 can partially replace p300 in cooperating with pRb to suppress DNA synthesis in baby rat kidney cells.     

Adenovirus 12 nononcogenic mutants: oncogenicity of transformed cells and viral proteins synthesized in vivo and in vitro

Several nononcogenic cyt mutants of adenovirus type 12 induced the same E1A (55,000 [55K] and 25K) and E1B polypeptides (55K, 19K, and 17K) as did the wild-type virus, except that cyt 68 did not induce the E1B 19K protein. Tumorigenicity tests showed cells transformed by cyt 68 to be highly oncogenic in vivo. Therefore, it was concluded that the E1B 19K polypeptide is not necessary for tumor induction but may be involved in the efficiency of transformation.     

Efficient nuclear localization and immortalizing ability, two functions dependent on the adenovirus type 5 (Ad5) E1A second exon, are necessary for cotransformation with Ad5 E1B but not with T24ras.

Expression of adenovirus type 5 E1A 12S is sufficient to immortalize primary baby rat kidney cells, but another viral or cellular oncogene, such as E1B or T24ras, is necessary for complete transformation. The regions of 12S sufficient for T24ras cotransformation have been well characterized and are located in the first exon. The second exon is dispensable for ras cotransformation, although it contains a region which appears to modulate the transforming phenotype. The same 12S first exon regions important in ras transformation are also necessary for E1B transformation. Analysis of an extensive series of second exon deletion and amino acid point mutations demonstrated that mutations affecting either the efficient nuclear localization and/or the immortalizing ability of the 12S protein also prevented cooperation with E1B. In general, the entire C-terminal half of 12S, including the nuclear localization signal, was necessary for efficient cotransformation with E1B. In addition to the differences between T24ras and E1B regarding 12S regions necessary for cotransformation, the characteristics of E1B-cotransformed foci differed from those of T24ras. The E1B foci took longer to appear and had a much slower growth rate. No hypertransformed foci were produced with E1B cotransfections, and established E1A-E1B lines exhibited minimal growth in soft agar compared with that of E1A-T24ras lines.     

p300 binding by E1A cosegregates with p53 induction but is dispensable for apoptosis.

E1A expression during adenovirus infection induces apoptosis. E1A expression causes accumulation of the p53 tumor suppressor protein, and E1A-induced apoptosis is p53 mediated in primary rodent cells, implying that p53 induction may be linked to apoptosis induction by E1A. Adenoviruses containing mutations in the E1A gene were tested for the ability to trigger both p53 accumulation and the appearance of enhanced cytopathy (cyt phenotype) and degradation of DNA (deg phenotype), indicative of apoptosis in infected HeLa cells. The adenoviruses had mutations which disrupted the pRb- and/or p300-binding activities of E1A so that the relationship between p53 induction and apoptosis and binding to these cellular proteins by E1A could be determined. An E1A mutation that specifically disrupted the p300-binding activity failed to induce p53 accumulation, whereas mutations in E1A which affected pRb binding induced p53 accumulation. Thus, p300 binding was required and pRb binding was dispensable for E1A-mediated accumulation of p53 in HeLa cells. All the E1A mutant viruses, regardless of the ability to induce p53 accumulation, induced the cyt and deg phenotypes, suggesting that p53 induction in infected HeLa cells was not essential for apoptosis, nor was binding of E1A to the pRb and/or p300 protein. The possibility that E1A induced a p53-independent apoptosis pathway was tested by analyzing the appearance of the cyt and deg phenotypes in Saos-2 cells, which were null for both alleles of p53, upon adenovirus infection. An adenovirus expressing wild-type 12S E1A induced both the cyt and deg phenotypes in Saos-2 cells, as did all the E1A mutant viruses. Thus, E1A expression during infection of human cells may trigger redundant p53-independent and -dependent apoptotic pathways.     

Mapping of cellular protein-binding sites on the products of early-region 1A of human adenovirus type 5.

The binding sites for the 300-, 107-, and 105-kilodalton cellular proteins which associate with human adenovirus type 5 E1A products were studied with E1A deletion mutants. All appeared to bind to the amino-terminal half of E1A products in regions necessary for oncogenic transformation. These results suggest that these cellular species may be important for the biological activity of E1A products.