Multiple effects of the 72-kDa, adenovirus-specified DNA binding protein on the efficiency of cellular transformation

The early region 2A gene (E2A) of adenovirus types 2 and 5 encodes a 72-kDa DNA binding protein (DBP) which contains two physical domains comprising approximately the amino-terminal one-third and carboxyl-terminal two-thirds of the protein, respectively. Previous work has shown that some Ad5 mutants containing temperature-sensitive (ts) mutations in the carboxyl-terminal domain of DBP, such as Ad5ts125, show a 3- to 8-fold enhanced ability to transform rat cells. We have examined the transformation characteristics of a series of Ad5 E2A deletion mutants, Ad5dl801-5, which encode either no functional DBP or encode truncated, defective DBPs. The E2A deletion mutants transformed rat embryo cells at frequencies similar to wild-type (wt) Ad5. These results suggest that the high transformation phenotype of carboxyl-terminal E2A mutants like Ad5ts125 is not due to the simple inactivation of DBP function, but rather results from an activity possessed by an altered DBP. This hypothesis is supported by the fact that the transformation phenotype of Adsts125 and similar mutants is dominant over the wild-type phenotype. A number of additional Ad2 and Ad5 E2A mutants were examined with respect to their ability to transform primary rat embryo cells. It was found that a carboxyl-terminal E2A mutant, Ad2+ND1ts23, also showed the enhanced transformation phenotype. In contrast, several amino-terminal E2A host-range (hr) mutants, originally isolated on the basis of their ability to replicate in monkey cells, transformed rat embryo cells at a frequency similar to wild-type virus. Ad2ts400, and E2A mutant with alterations in both DBP domains, showed a wild-type frequency of transformation, while two similar mutants, Ad5ts125 X 405 and Ad5ts125 X 404, showed an enhanced frequency. Last, it was found that coinfection of primary rat embryo cells with the hr mutants plus Ad5ts125 or Ad2+ND1ts23 resulted in a wild-type frequency of transformation, demonstrating that the hr mutants are dominant to the ts mutants with regard to transformation phenotype. Thus, DBP can both positively and negatively affect viral transformation in this system.     

A hamster temperature-sensitive G1 mutant, tsBN250 has a single point mutation in histidyl-tRNA synthetase that inhibits an accumulation of cyclin D1

Background: We have previously isolated a series of temperature-sensitive mutants for cell-proliferation from the BHK21 cell line, derived from the golden hamster. These mutants proliferate at 33.5 °C, the permissive temperature, but not at 39.5 °C the restrictive temperature. Using DNA-mediated gene transfer, human genes complementing these ts mutants were cloned.
Results: At 39.5 °C the tsBN250 cell line, a temperature-sensitive mutant of the BHK21 cell line, had a defect in the G1 phase, but not in the S phase. The human gene complementing tsBN250 cells was found to encode histidyl-tRNA synthetase. Indeed, the tsBN250 cell line had a single base change—guanine to adenine at the second position of the 362nd codon of hamster histidyl-tRNA-synthetase, converting arginine to histidine. Following release from serum starvation, cyclin E, but not cyclin D1, was accumulated, while, at 39.5 °C, the mRNA of cyclin D1 was normally expressed in tsBN250 cells. A similar inhibition of cyclin D1 accumulation was observed in another ts mutant, tsBN269, which has a single point mutation in lysyl-tRNA synthetase. Overexpression of cyclin D1 enabled tsBN250 cells to enter the S phase.
Conclusion: tsBN250 cells have a single point mutation in histidyl tRNA synthetase that causes a loss of histidyl-tRNA synthetase activity which in turn reduces the content of cyclin D1, but not of cyclin E, thereby resulting in G1 arrest.     

Caspase-mediated cleavage of adenovirus early region 1A proteins

Adenovirus 2 and 12 early region 1A (Ad2 and Ad12 E1A) proteins were cleaved during cisplatin-induced apoptosis of Ad-transformed rat and human cells. Cleavage was inhibited in the presence of caspase inhibitors such as Z-VAD-FMK. In Ad12 transformants both 13S and 12S E1A proteins were cleaved at a similar rate. In Ad2 transformants the E1A 13S component was appreciably less stable than the 12S component. In in vitro studies Ad2 and Ad12 E1A 13S and Ad2 12S proteins were rapidly cleaved by caspase 3 whereas Ad12 12S E1A and Ad12 13S E1A were rapidly degraded by caspase 7. Cleavage sites in Ad12 13S proteins for caspase 3 have been determined. Initial cleavage occurred at D24 and D150; this was followed by cleavage at D204 and D242. Caspase-3-mediated cleavage of Ad12 13S E1A destroyed its ability to bind to CBP and TBP but interaction between C terminal E1A polypeptides and CtBP was observed. During viral infection Ad5 and Ad12 E1A 12S proteins were markedly more stable than 13S proteins but no difference was observed in Ad E1A levels in the absence or presence of the caspase inhibitors Z-VAD-FMK or Z-D(OMe)-E(OMe)-V-D(OMe)-CH(2)F. Limited caspase 3 and 10 activation occurred during infection with the E1B 19K(-) virus Ad2 pm1722 but little or no activation of caspase 3 was observed during wt virus infection. Examination of protein cleavage during viral infection of A549 cells showed proteolysis of lamin B and PARP in response to Ad5 wt and Ad2 pm1722. Protein degradation in response to both viruses was partially inhibited by Z-VAD-FMK. Following infection of human skin fibroblasts lamin B was degraded, although only limited changes in PARP levels were observed. We have concluded that Ad E1A is cleaved by caspases during apoptosis but not during viral infection. However, some of the processes commonly associated with apoptosis occur during viral infection, particularly with E1B 19K(-) mutants, although apoptosis per se is not evident.     

Deletion mutants of region E1a of AD12 E1 plasmids: effect on oncogenic transformation

Plasmids containing the E1 region of Ad12 DNA can transform certain rodent cells into oncogenic cells. To study the role of the E1a subregion in the process of oncogenic transformation, Ad12 region E1 mutants carrying deletions in the E1a region were constructed. Deletion mutants pR7 and pR8 affect only the 13 S mRNA species encoded by region E1a, whereas deletion mutants pR11 and pR15 damage both the 12 S and 13 S E1a mRNA. All four mutants have lost their capacity to transform primary cultures of baby rat kidney cells, indicating that the E1a gene product encoded by 13 S mRNA is essential for transformation. It was further found that the mutated E1a regions of both pR7 and pR11 can induce expression of region E1b, which implies that the transformation deficiency of these mutants is not due to the inability to activate E1b expression. Surprisingly, the transforming capacity of pR7 and pR11 is restored when these mutant E1 regions are covalently coupled to the SV40 "enhancer" region. Cells transformed by these hybrids plasmids, however, were not tumorigenic in nude mice.     

Intercistronic complementation between adenovirus 2 temperature-sensitive mutants.

The mechanism of complementation, dominance, and the effects of gene dosage were examined by means of sodium dodecyl sulfate-polyacrylamide gel analysis of infected cell proteins and purified virions synthesized in cells coinfected with adenovirus type 2 temperature-sensitive mutants. In particular the synthesis of a 50K core-polypeptide, related to protein V, and the processing of core protein PVII were examined. The efficiency of complementation was unique to each mutant. Coinfection with wild-type failed to suppress completely the mutant phenotype suggesting that the mutants were defective in non-catalytic proteins. Interserotype complementation between Ad2ts3, a hexon mutant, and Ad5ts22, a fiber mutant, resulted in mosaic virions with Ad2 fibers and Ad5 hexons, rather than in mosaic capsomeres. The putative hexon mutant, ts3, was found to exert a gene frequency-dependent dominance. Evidence was also obtained suggesting that the ts3-specific 50K polypeptide may be a precursor to core protein V. We conclude that complementation is profoundly influenced by the nature of the participating gene products and their function in virus infection.     

Functional relatedness between the E1a and E1b regions of group C and group D human adenoviruses

The functional relatedness of the transforming genes (E1a and E1b) of adenovirus type 9 (group D) which induces mammary tumors in rats and those of the non-tumorigenic adenoviruses, Ad2 and Ad5 (group C) was examined. Transfection of established rat embryo cells with a DNA segment containing the E1a and E1b regions of Ad9 resulted in efficient transformation; similar results have been shown for group A, B and C Ads. In contrast to Ads of group A, B and C, Ad9 DNA containing the E1 region or the entire viral genome was unable to transform primary baby rat kidney (BRK) cells. The functional relatedness of genes encoded within the E1 region was compared using a mutant complementation assay in which various group C mutants defective in the entire E1 region or in the E1a or E1b regions alone as well as mutants defective exclusively within the 19K or 58K T antigens coding regions of E1b were coinfected with wild type (wt) Ad9 and tested for group C mutant DNA replication, virus production, or expression of early and late genes. These studies have shown that a defect in the entire E1 region of Ad2 could only be complemented poorly by Ad9; our earlier studies have shown that coinfection with Ad12 (group A) or Ad7 (group B) resulted in efficient complementation (Brusca and Chinnadurai (1981) J. Virol. 39, 300-305). Further analysis indicated that a defect in the E1a region could be complemented by the group D E1a region. The level of E1a complementation as judged by mutant DNA replication and activation of expression of mutant early viral genes was about one-fourth to one-fifth the level in 293 cells that constitutively express Ad5 E1a and E1b regions. Our results indicate that a defect in the E1b 19K T antigen, which leads to degradation of intracellular DNA in infected cells, could be complemented by the group D protein. However, a defect in the E1b 58K T antigen could not be efficiently complemented by the group D protein. Coinfection of group C mutants defective in the 58K T antigen and Ad9 wt did not lead to an increase in the mutant viral production. Furthermore, in cells coinfected with the 58K T antigen mutants and Ad9 wt there was a large reduction in the accumulation of group C late cytoplasmic RNA. The observed complementation defect of Ad9 in supporting multiplication of group C mutants defective in the entire E1 region may therefore be a cumulative effect of both E1a and E1b regions.     

The structure of cauliflower mosaic virus. II. Identity and location of the viral polypeptides.

Adenovirus type 12 (Ad 12) tumor-specific transplantation antigen (TSTA) and the surface (S) antigen were examined using rat cells transformed with Ad 12 DNA and its fragments. WY3 (3Y1 cells transformed with Ad 12 whole DNA), CY1 (3Y1 cells transformed with the EcoRI-C fragment of Ad 12 DNA), and GY cells (3Y1 cells transformed with the HindIII-G fragment of Ad 12 DNA) contained TSTA and S antigen, but HY cells (3Y1 cells transformed with the BpaI-H fragment of Ad12 DNA) did not. These results suggest that TSTA and S antigens contain a protein(s) coded for by a portion of the transforming gene.     

An early function of the adenoviral E1B 55 kDa protein is required for the nuclear relocalization of the cellular p53 protein in adenovirus-infected normal human cells

It is well established that the human subgroup C adenovirus type 5 (Ad5) E1B 55 kDa protein can regulate the activity and concentration of the cellular tumor suppressor, p53. However, the contribution(s) of these functions of the E1B protein to viral reproduction remains unclear. To investigate this issue, we examined properties of p53 in normal human cells infected by E1B mutant viruses that display defective entry into the late phase or viral late mRNA export. The steady-state concentrations of p53 were significantly higher in cells infected by the E1B 55 kDa null mutant Hr6 or three mutants carrying small insertions in the E1B 55 kDa protein coding sequence than in Ad5-infected cells. Nevertheless, none of the mutants induced apoptosis in infected cells. Rather, the localization of p53 to E1B containing nuclear sites observed during infection by Ad5 was prevented by mutations that impair interaction of the E1B protein with p53 and/or with the E4 Orf6 protein. These results indicate that the E1B protein fulfills an early function that correlates efficient entry into the late phase with the localization of E1B and p53 in the nucleus of Ad5-infected normal human cells.     

Isolation and characterization of insertion mutants in E1A of adenovirus type 5.

We have constructed a series of insertion mutations at 18 sites in the coding sequences of early region 1A (E1A) of human adenovirus type 5 (Ad5). At each site we have introduced three types of mutation: a 39-bp insertion specifying a 13-aa residue oligopeptide, a 39-bp insertion containing chain termination codons in all three reading frames, and a "collapsed" insert of 6-bp forming a conventional linker insertion mutation. All mutants were sequenced to determine the precise location, structure, and orientation of the inserts. The mutants were assayed for their abilities to trans-activate and to repress using transient expression assays in HeLa cells cotransfected with the E1A mutant plasmids and a reporter plasmid containing the bacterial beta-galactosidase (lac Z) gene under the control of Ad5 early promoters. The mutants were also tested for their ability to transform baby rat kidney cells in cooperation with either E1B or the ras oncogene. Each mutant was rescued into virus and infectivity was compared in HeLa and 293 cells. In addition, E1A protein synthesis was analyzed in cells infected with the mutant viruses and the insertions were found to have pronounced but unpredictable effects on electrophoretic mobility of E1A proteins in SDS-polyacrylamide gels. The results of functional assays indicated that only mutations mapping in, or deleting, the unique region of the 13 S mRNA product had any effect on ability to trans-activate and that a perfect correlation existed between ability of a mutant to trans-activate and to replicate efficiently in HeLa cells or to transform baby rat kidney cells in an E1A plus E1B mediated assay. In contrast, insertions near conserved region 2 of exon I and in the NH2-terminal portion of exon II significantly reduced repression activity but left transforming activity with E1B or with ras essentially unaffected suggesting that the repression function of E1A is separate from, or at least nonessential in, transformation.     

Transforming properties of a 15-kDa truncated Ad12 E1A gene product

A mutant Ad12 E1A region (Ad12 R11E1A) was constructed, which directs the synthesis of only a 15-kDa N-terminal E1A product. When controlled by the SV40 early promoter plus enhancer region (SVR11E1A) this mutant E1A region is capable of immortalizing primary baby rat kidney (BRK) cells, showing that the information essential for immortalization is located in the N-terminal part of region E1A and is shared by the 13 S and 12 S mRNA gene products. This immortalization is thought to be an essential step in the process of oncogenic transformation. Primary BRK cells transformed by SVR11E1A in the presence of Ad12 E1B are nononcogenic. This implies that the E1A region also codes for activities required for oncogenicity. However, in the presence of an activated c-Ha-ras oncogene the SVR11E1A region can oncogenically transform primary BRK cells, showing that the c-Ha-ras oncogene not only can complement for the Ad12 E1B region, but also for the E1A function lost by the R11 deletion.     

C-terminal domain of the adenovirus E1A oncogene product is required for induction of cytotoxic T lymphocytes and tumor-specific transplantation immunity

Adenovirus genes required for the elicitation of adenovirus group C-specific cytolytic T lymphocytes (CTLs) and for the induction of adenovirus-specific transplantation antigen (TSTA) were identified by immunization with a library of adenovirus mutants. The group C Ad-specific CTL response was elicited by immunization with wild-type adenovirus type 5 (Ad5) or with recombinant adenoviruses containing Ad5 E1A gene. The specific CTL response was also elicited by Ad5 virus constructs which express only the 12 S or 13 S E1A early mRNA, but not with viruses unable to express E1A protein sequences normally encoded by the E1A early messages. The induction of transplantation immunity against tumorigenic Ad-transformed cells was studied next. The product encoded by either 13 S and 12 S E1A mRNA alone was sufficient for strong TSTA activity. A series of viruses with mutations within the first exon of the E1A message also induced strong TSTA, while Ad5 mutants with lesions within the second exon failed to induce syngraft immunity. These results provide strong evidence that amino acid sequence encoded by the second exon of the Ad5 E1A message is required, either directly or indirectly, for the induction of both Ad-specific CTL and Ad TSTA.     

Assignment of avian reovirus temperature-sensitive mutant recombination groups E, F, and G to genome segments

Avian reoviruses (ARV) are less well understood than their mammalian counterparts. ARV are ubiquitous in commercial poultry and frequently isolated from acutely infected chickens. We previously described isolation of ARV temperature-sensitive (ts) mutants after nitrosoguanidine mutagenesis of wild-type ARV138, their assignment to 7 recombination groups (A-G), and genetic mapping of mutants in groups A-D to specific gene segments. For this study, wild-type serotype ARV176 was crossed with ts mutants tsE158 (Group E), tsF206 (Group F), or tsG247 (Group G) and reassortant progenies analyzed. Reassortant temperature-sensitivities were determined by efficiency of plating at permissive and non-permissive temperatures. Mapping results indicated tsE158, tsF206, and tsG247 mapped to the L1, S4, and L3 genes, respectively, which encode the lambdaA core shell, sigmaNS non-structural, and lambdaC core spike proteins, respectively. Specific amino acid substitutions in each mutant were determined and locations of structural protein alterations were placed within the 3-dimensional structure of homologous mammalian reovirus proteins. Mapping recombination groups E-G marks completion of gene assignments for all seven ts mutant groups previously generated.     

Genetic analysis of control of proliferation in fibroblastic cells in culture. II. Alteration in proliferative and survival phenotypes in a set of temperature-sensitive mutants of rat 3Y1 cells after infection or transformation with simian virus 40

Mutants of rat 3Y1 fibroblasts, temperature sensitive for proliferation or survival and which represent each of eight complementation groups, were examined to determine whether cells made quiescent at confluence at 33.8C (permissive temperature) can be stimulated to enter S phase at 39.8C (nonpermissive temperature) by 20% serum or by infection with simian virus 40 (SV40). Three mutants with a short survival at 39.8 C did not enter S phase at 39.8 C under either condition. The remaining five entered S at 39.8C by infection with SV40. However, only one of these five entered S in response to high serum. After transformation with SV40, three mutants accumulating at 39.8C with a predominantly 2n (G₁) DNA content did not proliferate, there was a rapid and extensive cell death, and the cells had a DNA content similar to that seen in randomly proliferating populations. The other two mutants, accumulating at 39.8 C with a predominantly 2n or 2n.4n DNA content, proliferated at this temperature after transformation with SV40. These results clearly indicate that SV40 interacts closely with cellular ts lesions related to control of proliferation and cell survival.On leave of absence from the Division of Child Neurology, Institute of Neurological Sciences, Tottori University, School of Medicine.