Sequence-independent autoregulation of the adenovirus type 5 E1A transcription unit

Targeting the Retinoblastoma/E2F repressive complex by CDK4/6 inhibitors amplifies oncolytic potency of an oncolytic adenovirus

CDK4/6 inhibitors (CDK4/6i) and oncolytic viruses are promising therapeutic agents for the treatment of various cancers. As single agents, CDK4/6 inhibitors that are approved for the treatment of breast cancer in combination with endocrine therapy cause G1 cell cycle arrest, whereas adenoviruses induce progression into S-phase in infected cells as an integral part of the their life cycle. Both CDK4/6 inhibitors and adenovirus replication target the Retinoblastoma protein albeit for different purposes. Here we show that in combination CDK4/6 inhibitors potentiate the anti-tumor effect of the oncolytic adenovirus XVir-N-31 in bladder cancer and murine Ewing sarcoma xenograft models. This increase in oncolytic potency correlates with an increase in virus-producing cancer cells, enhanced viral genome replication, particle formation and consequently cancer cell killing. The molecular mechanism that regulates this response is fundamentally based on the reduction of Retinoblastoma protein expression levels by CDK4/6 inhibitors.

Neither CDK4/6 inhibitors nor oncolytic adenoviruses show high efficiency as monotherapy in the treatment of cancer. Authors show here that when combined, CDK4/6 inhibitors deplete Retinoblastoma protein levels, which leads to more efficient virus replication and an increase in oncolytic virus-producing cancer cells and thus to efficient anti-tumor response in mouse xenograft sarcoma models.

Local character of readthrough activation in adenovirus type 5 early region 1 transcription control

Wild-type early activity of the adenovirus 5 E1b gene promoter requires readthrough transcription originating from the adjacent upstream E1a gene. This unusual mode of viral transcription activation was identified by genetic manipulation of the mouse beta(maj)-globin gene transcription termination sequence (GGT) inserted into the E1a gene. To facilitate further study of the mechanism of readthrough activation, the activities of GGT and a composite termination sequence CT were tested in recombinant adenoviruses containing luciferase reporters driven by the E1b promoter. There was a strict correlation between readthrough and substantial downstream gene expression, indicating that interference with downstream transcription was not a unique property of GGT. Blockage of readthrough transcription of E1a had no apparent effect on early expression of the major late promoter, the next active promoter downstream of E1b. A test for epistatic interaction between termination sequence insertions and E1a enhancer mutations suggested that readthrough activation and E1a enhancer activation of the E1b promoter are mechanistically distinct. In addition, substitution of the human cytomegalovirus major immediate-early promoter for the E1b promoter suppressed the requirement for readthrough. These results suggest that readthrough activation is a "local" effect of a direct interaction between the invading transcription elongation complex and the E1b promoter. DNase I hypersensitivity footprinting provided evidence that this interaction altered an extensive E1b promoter DNA-protein complex that was assembled in the absence of readthrough transcription.

Cre levels limit packaging signal excision efficiency in the Cre/loxP helper-dependent adenoviral vector system

Helper-dependent (HD) adenovirus vectors devoid of all viral coding sequences have a large cloning capacity and provide long-term transgene expression in vivo with negligible toxicity, making them attractive vectors for gene therapy. Currently, the most efficient means of producing HD vectors involves coinfecting 293 cells expressing Cre with the HD vector and a helper virus bearing a packaging signal flanked by loxP sites. Cre-mediated packaging signal excision renders the helper virus genome unpackageable but still able to replicate and provide helper functions for HD vector propagation. Typically, helper virus contamination is < or =1% pre- and < or =0.1% postpurification by CsCl banding. While these contamination levels are low, further reduction is desirable. However, this objective has not been realized since the Cre/loxP system was first developed. This lack of progress is due, at least in part, to our lack of understanding of the origins of the contaminating helper virus, thus rendering its reduction or elimination difficult to achieve. This study was designed to investigate the possible sources of contaminating helper virus persisting during HD vector amplification. The results revealed that Cre is limiting in helper virus-infected Cre-expressing 293 cells, thereby permitting helper viruses to escape packaging signal excision and propagate. The results of this study should provide a foundation for developing rational strategies to further reduce or possibly eliminate the contaminating helper virus.

E1B 55-kilodalton-associated protein: a cellular protein with RNA-binding activity implicated in nucleocytoplasmic transport of adenovirus and cellular mRNAs

The adenovirus type 5 (Ad5) early 1B 55-kDa protein (E1B-55kDa) is a multifunctional phosphoprotein that regulates viral DNA replication and nucleocytoplasmic RNA transport in lytically infected cells. In addition, E1B-55kDa provides functions required for complete oncogenic transformation of rodent cells in cooperation with the E1A proteins. Using the far-Western technique, we have isolated human genes encoding E1B-55kDa-associated proteins (E1B-APs). The E1B-AP5 gene encodes a novel nuclear RNA-binding protein of the heterogeneous nuclear ribonucleoprotein (hnRNP) family that is highly related to hnRNP-U/SAF-A. Immunoprecipitation experiments indicate that two distinct segments in the 55-kDa polypeptide which partly overlap regions responsible for p53 binding are required for complex formation with E1B-AP5 in Ad-infected cells and that this protein interaction is modulated by the adenovirus E4orf6 protein. Expression of E1B-AP5 efficiently interferes with Ad5 E1A/E1B-mediated transformation of primary rat cells. Furthermore, stable expression of E1B-AP5 in Ad-infected cells overcomes the E1B-dependent inhibition of cytoplasmic host mRNA accumulation. These data suggest that E1B-AP5 might play a role in RNA transport and that this function is modulated by E1B-55kDa in Ad-infected cells.

Mouse adenovirus type 1 early region 1A is dispensable for growth in cultured fibroblasts

Mouse adenovirus type 1 (MAV-1) mutants with deletions of conserved regions of early region 1A (E1A) or with point mutations that eliminate translation of E1A were used to determine the role of E1A in MAV-1 replication. MAV-1 E1A mutants expressing no E1A protein grew to titers comparable to wild-type MAV-1 titers on mouse fibroblasts (3T6 fibroblasts and fibroblasts derived from Rb+/+, Rb+/-, and Rb-/- transgenic embryos). To test the hypothesis that E1A could induce a quiescent cell to reenter the cell cycle, fibroblasts were serum starved to stop DNA replication and cellular replication and then infected with the E1A mutant and wild-type viruses. All grew to equivalent titers. Steady-state levels of MAV-1 early mRNAs (E1A, E1B, E2, E3, and E4) from 3T6 cells infected with wild-type or E1A mutant virus were examined by Northern analysis. Steady-state levels of mRNAs from the mutant-infected cells were comparable to or greater than the levels found in wild-type virus infections for most of the early regions and for two late genes. The E2 mRNA levels were slightly reduced in all mutant infections relative to wild-type infections. E1A mRNA was not detected from infections with the MAV-1 E1A null mutant, pmE109, or from infections with similar MAV-1 E1A null mutants, pmE112 and pmE113. The implications for the lack of a requirement of E1A in cell culture are discussed.

Readthrough activation of early adenovirus E1b gene transcription

In cells productively infected with adenovirus type 5, transcription is not terminated between the E1a gene and the adjacent downstream E1b gene. Insertion of the mouse beta(maj)-globin transcription termination sequence (GGT) into the E1a coding region dramatically reduces early, but not late, E1b expression (E. Falck-Pedersen, J. Logan, T. Shenk, and J. E. Darnell, Jr., Cell 40:897-905, 1985). In the study described herein, we showed that base substitution mutations in the globin DNA that specifically relieved transcription termination also restored early E1b promoter activity in cis, establishing that maximal early E1b expression requires readthrough transcription originating from the adjacent upstream gene. To identify potential targets of readthrough activation, a series of recombinant viruses with double mutations was constructed. Each double-mutant virus strain had the transcription termination sequences in the first exon of E1a and a deletion within the transcription control region of E1b. Early E1b expression from the double-mutant strains was more defective than that from strains containing either mutation alone, indicating that the deleted regions (positions -362 to -35) are not the target for readthrough activation. Two findings suggested that a cis-dominant property of early viral templates is important for readthrough activation. First, the early E1b defect caused by the GGT insertion was not complemented in trans by factors present in late-infected cells. Second, restoration of E1b transcription at late times occurred concurrently with viral DNA replication. Readthrough activation may help convert virion DNA into a transcriptionally competent template prior to DNA replication and late transcription.

Adenoviral E1B-55kDa protein inhibits yeast mRNA export and perturbs nuclear structure

The mechanisms of export of RNA from the nucleus are poorly understood; however, several viral proteins modulate nucleocytoplasmic transport of mRNA. Among these are the adenoviral proteins E1B-55kDa and E4-34kDa. Late in infection, these proteins inhibit export of host transcripts and promote export of viral mRNA. To investigate the mechanism by which these proteins act, we have expressed them in Saccharomyces cerevisiae. Overexpression of either or both proteins has no obvious effect on cell growth. By contrast, overexpression of E1B-55kDa bearing a nuclear localization signal (NLS) dramatically inhibits cell growth. In this situation, the NLS-E1B-55kDa protein is localized to the nuclear periphery, fibrous material is seen in the nucleoplasm, and poly(A)+ RNA accumulates in the nucleus. Simultaneous overexpression of E4-34kDa bearing or lacking an NLS does not modify these effects. We discuss the mechanisms of selective mRNA transport.

Mammalian nonsense codons can be cis effectors of nuclear mRNA half-life

Frameshift and nonsense mutations within the gene for human triosephosphate isomerase (TPI) that generate a nonsense codon within the first three-fourths of the protein coding region have been found to reduce the abundance of the product mRNA that copurifies with nuclei. The cellular process and location of the nonsense codon-mediated reduction have proven difficult to elucidate for technical reasons. We show here, using electron microscopy to judge the purity of isolated nuclei, that the previously established reduction to 25% of the normal mRNA level is evident for nuclei that are free of detectable cytoplasmic contamination. Therefore, the reduction is likely to be characteristic of bona fide nuclear RNA. Fully spliced nuclear mRNA is identified by Northern (RNA) blot hybridization and a reverse transcription-PCR assay as the species that undergoes decay in experiments that used the human c-fos promoter to elicit a burst and subsequent shutoff of TPI gene transcription upon the addition of serum to serum-deprived cells. Finally, the finding that deletion of a 5' splice site of the TPI gene results predominantly but not exclusively in the removal by splicing (i.e., skipping) of the upstream exon as a part of the flanking introns has been used to demonstrate that decay is specific to those mRNA products that maintain the nonsense codon. This result, together with our previous results that implicate translation by ribosomes and charged tRNAs in the decay mechanism, indicate that nonsense codon recognition takes place after splicing and triggers decay solely in cis. The possibility that decay takes place during the process of mRNA export from the nucleus to the cytoplasm is discussed.

Mammalian nonsense codons can be cis effectors of nuclear mRNA half-life

Frameshift and nonsense mutations within the gene for human triosephosphate isomerase (TPI) that generate a nonsense codon within the first three-fourths of the protein coding region have been found to reduce the abundance of the product mRNA that copurifies with nuclei. The cellular process and location of the nonsense codon-mediated reduction have proven difficult to elucidate for technical reasons. We show here, using electron microscopy to judge the purity of isolated nuclei, that the previously established reduction to 25% of the normal mRNA level is evident for nuclei that are free of detectable cytoplasmic contamination. Therefore, the reduction is likely to be characteristic of bona fide nuclear RNA. Fully spliced nuclear mRNA is identified by Northern (RNA) blot hybridization and a reverse transcription-PCR assay as the species that undergoes decay in experiments that used the human c-fos promoter to elicit a burst and subsequent shutoff of TPI gene transcription upon the addition of serum to serum-deprived cells. Finally, the finding that deletion of a 5' splice site of the TPI gene results predominantly but not exclusively in the removal by splicing (i.e., skipping) of the upstream exon as a part of the flanking introns has been used to demonstrate that decay is specific to those mRNA products that maintain the nonsense codon. This result, together with our previous results that implicate translation by ribosomes and charged tRNAs in the decay mechanism, indicate that nonsense codon recognition takes place after splicing and triggers decay solely in cis. The possibility that decay takes place during the process of mRNA export from the nucleus to the cytoplasm is discussed.

Adenovirus E4orf4 protein reduces phosphorylation of c-Fos and E1A proteins while simultaneously reducing the level of AP-1

Adenovirus E1A protein and cyclic AMP cooperate to induce transcription factor AP-1 and viral gene expression in mouse S49 cells. We report that a protein encoded within the viral E4 gene region acts to counterbalance the induction of AP-1 DNA-binding activity by E1A and cyclic AMP. Studies with mutant adenoviruses demonstrated that in the absence of E4orf4 protein, AP-1 DNA-binding activity is induced to substantially higher levels than in wild-type virus-infected cells. The induction is the result of increased production of JunB and c-Fos proteins. Hyperphosphorylated forms of c-Fos and E1A proteins accumulate in the absence of functional E4orf4 protein. We propose that the E4orf4 protein acts to inhibit the activity of a cellular kinase that phosphorylates both the E1A and c-Fos proteins. Phosphorylation-dependent alterations in the activity of c-Fos, E1A, or some unidentified protein might, then, lead to decreased synthesis of AP-1 components. This E4 function likely plays an important role in natural infections, since a mutant virus unable to express the E4orf4 protein is considerably more cytotoxic than the wild-type virus.

Negative autoregulation of the neu gene is mediated by a novel enhancer

In an attempt to study potential feedback regulation of the neu oncogene, we have found that the neu oncogene product specifically represses its own promoter activity. Deletion analysis indicated a 140-bp region (nucleotides -312 to -173 relative to the ATG initiation codon) in the rat neu promoter responsible for neu autorepression. Gel shift assays and methylation interference analysis further demonstrated that a GGTGGGGGGG sequence (nucleotides -243 to -234 relative to the ATG initiation codon) in this 140-bp region interacts with specific protein complexes. The GGTGGGGGGG sequence (GTG element), which functions as an enhancer, is sufficient to cause neu-mediated repression in a heterologous promoter. Furthermore, it produces different gel shift patterns with nuclear extracts from neu-transformed cell lines and their parental lines, suggesting that a transcriptional factor(s) interacting with this enhancer element has been perturbed by the introduction of neu. Taken together, the data presented in this report show that (i) the neu oncogene product autorepresses its own promoter, (ii) the neu promoter contains a novel enhancer, and (iii) neu autorepression is mediated through this enhancer, likely by inhibition of the enhancer activity.

Negative autoregulation of the neu gene is mediated by a novel enhancer

In an attempt to study potential feedback regulation of the neu oncogene, we have found that the neu oncogene product specifically represses its own promoter activity. Deletion analysis indicated a 140-bp region (nucleotides -312 to -173 relative to the ATG initiation codon) in the rat neu promoter responsible for neu autorepression. Gel shift assays and methylation interference analysis further demonstrated that a GGTGGGGGGG sequence (nucleotides -243 to -234 relative to the ATG initiation codon) in this 140-bp region interacts with specific protein complexes. The GGTGGGGGGG sequence (GTG element), which functions as an enhancer, is sufficient to cause neu-mediated repression in a heterologous promoter. Furthermore, it produces different gel shift patterns with nuclear extracts from neu-transformed cell lines and their parental lines, suggesting that a transcriptional factor(s) interacting with this enhancer element has been perturbed by the introduction of neu. Taken together, the data presented in this report show that (i) the neu oncogene product autorepresses its own promoter, (ii) the neu promoter contains a novel enhancer, and (iii) neu autorepression is mediated through this enhancer, likely by inhibition of the enhancer activity.

Functional diversity of E1A gene autoregulation among human adenoviruses

Autoregulation of the adenovirus E1A gene involves its constitutive expression and positively and negatively regulated transcription. Dissection of this process will identify basal-level cis elements and autoregulatory targets of the E1A promoter and functional domains within the trans-acting E1A gene products. In this report, the DNA sequence of the human subgroup B adenovirus type 3 (Ad3) E1A gene is presented and compared with that of the E1A genes of similar and distantly related human adenoviruses. The cDNA forms of the Ad3 E1A gene, corresponding to two major early mRNA species, are cloned, sequenced, and subcloned into plasmid expression vectors. Cotransfections of cell cultures are performed with Ad5 or Ad3 E1A gene expression plasmids and a reporter gene under control of the Ad5 or Ad3 E1A promoter. The Ad5 and Ad3 E1A promoters are similarly repressed by either serotype's 12S cDNA gene products. The Ad3 E1A promoter responds much more strongly than the Ad5 E1A promoter to transactivation by 13S cDNA gene products. In contrast, the 13S cDNA gene of Ad5 has greater transactivation activity than that of Ad3. Experiments with missense mutations of the Ad5 E1A gene indicate that transactivation of the Ad5 E1A promoter is weak, just reversing or balancing negative autorepression. Single amino acid substitutions in the conserved, repressive functional domain 2 of the E1A gene modulate transactivating activity that is usually associated with the separate and distal conserved functional domain 3. These results suggest a strong structure-function relationship influenced by the variable sequences separating these conserved domains.

Nonsense codons in human beta-globin mRNA result in the production of mRNA degradation products

Human beta zero-thalassemic beta-globin genes harboring either a frameshift or a nonsense mutation that results in the premature termination of beta-globin mRNA translation have been previously introduced into the germ line of mice (S.-K. Lim, J.J. Mullins, C.-M. Chen, K. Gross, and L.E. Maquat, EMBO J. 8:2613-2619, 1989). Each transgene produces properly processed albeit abnormally unstable mRNA as well as several smaller RNAs in erythroid cells. These smaller RNAs are detected only in the cytoplasm and, relative to mRNA, are longer-lived and are missing sequences from either exon I or exons I and II. In this communication, we show by using genetics and S1 nuclease transcript mapping that the premature termination of beta-globin mRNA translation is mechanistically required for the abnormal RNA metabolism. We also provide evidence that generation of the smaller RNAs is a cytoplasmic process: the 5' ends of intron 1-containing pre-mRNAs were normal, the rates of removal of introns 1 and 2 were normal, and studies inhibiting RNA synthesis with actinomycin D demonstrated a precursor-product relationship between full-length mRNA and the smaller RNAs. In vivo, about 50% of the full-length species that undergo decay are degraded to the smaller RNAs and the rest are degraded to undetectable products. Exposure of erythroid cells that expressed a normal human beta-globin transgene to either cycloheximide or puromycin did not result in the generation of the smaller RNAs. Therefore, a drug-induced reduction in cellular protein synthesis does not reproduce this aspect of cytoplasmic mRNA metabolism. These data suggest that the premature termination of beta-globin mRNA translation in either exon I or exon II results in the cytoplasmic generation of discrete mRNA degradation products that are missing sequences from exon I or exons I and II. Since these degradation products appear to be the same for all nonsense codons tested, there is no correlation between the position of translation termination and the sites of nucleolytic cleavage.

Nonsense codons in human beta-globin mRNA result in the production of mRNA degradation products

Human beta zero-thalassemic beta-globin genes harboring either a frameshift or a nonsense mutation that results in the premature termination of beta-globin mRNA translation have been previously introduced into the germ line of mice (S.-K. Lim, J.J. Mullins, C.-M. Chen, K. Gross, and L.E. Maquat, EMBO J. 8:2613-2619, 1989). Each transgene produces properly processed albeit abnormally unstable mRNA as well as several smaller RNAs in erythroid cells. These smaller RNAs are detected only in the cytoplasm and, relative to mRNA, are longer-lived and are missing sequences from either exon I or exons I and II. In this communication, we show by using genetics and S1 nuclease transcript mapping that the premature termination of beta-globin mRNA translation is mechanistically required for the abnormal RNA metabolism. We also provide evidence that generation of the smaller RNAs is a cytoplasmic process: the 5' ends of intron 1-containing pre-mRNAs were normal, the rates of removal of introns 1 and 2 were normal, and studies inhibiting RNA synthesis with actinomycin D demonstrated a precursor-product relationship between full-length mRNA and the smaller RNAs. In vivo, about 50% of the full-length species that undergo decay are degraded to the smaller RNAs and the rest are degraded to undetectable products. Exposure of erythroid cells that expressed a normal human beta-globin transgene to either cycloheximide or puromycin did not result in the generation of the smaller RNAs. Therefore, a drug-induced reduction in cellular protein synthesis does not reproduce this aspect of cytoplasmic mRNA metabolism. These data suggest that the premature termination of beta-globin mRNA translation in either exon I or exon II results in the cytoplasmic generation of discrete mRNA degradation products that are missing sequences from exon I or exons I and II. Since these degradation products appear to be the same for all nonsense codons tested, there is no correlation between the position of translation termination and the sites of nucleolytic cleavage.

A protein kinase is present in a complex with adenovirus E1A proteins

A kinase activity can be immunoprecipitated in a complex that includes adenovirus E1A proteins. In vitro, this activity phosphorylated other E1A-associated proteins, as well as added E1A and histone H1 proteins. The E1A-associated kinase activity was cleared from extracts with an antibody to cyclin A, but not with antibody to cyclin B. The formation of a complex that included the kinase activity required amino acids 30-60 and 122-129 on the E1A proteins, sequences needed for association of E1A proteins with cyclin A and the retinoblastoma protein and implicated in control of cell growth. The complex of E1A-associated proteins included a 33-kDa ATP-binding protein, similar in size to a cyclin A-associated cdc2 kinase family member. Sucrose gradient analysis revealed two distinct E1A-containing complexes with the kinase activity. We suggest that E1A proteins may affect cellular proliferation by interacting with a member of the cdc2 kinase family and thereby influencing its activity.

Cooperative binding of EF-1A to the E1A enhancer region mediates synergistic effects on E1A transcription during adenovirus infection

A cellular nuclear factor, EF-1A, binds to a sequence motif which is repeated in the adenovirus type 5 E1A transcriptional control region. Previous genetic analyses demonstrated that two of these binding sites are predominant functional elements of the E1A enhancer region in vivo. In this report, we demonstrate that the cooperative binding of EF-1A to neighboring sites in the E1A enhancer region results in a synergistic activation of E1A transcription in infected cells.

Induction of c-fos mRNA and AP-1 DNA-binding activity by cAMP in cooperation with either the adenovirus 243- or the adenovirus 289-amino acid E1A protein

Products of the adenovirus E1A gene can act synergistically with cAMP to activate transcription of several viral early genes and the cellular genes c-fos and jun-B. Transcription factor AP-1-binding activity is also induced by the combined action of E1A and cAMP. Mouse S49 cells were infected with adenovirus variants expressing either the 243- or 289-amino acid E1A protein and treated with the cAMP analog dibutyryl-cAMP. Significant E1A-dependent induction of c-fos mRNA and AP-1-binding activity was observed in cells expressing either E1A protein. These effects absolutely required the presence of cAMP. In contrast, the 243-amino acid protein was a poor activator of the viral early genes E2 and E4 compared with the 289-amino acid protein. These data suggest that the 243- and 289-amino acid E1A proteins both interact functionally with the cAMP signaling system to activate transcription of a cellular gene and AP-1-binding activity. The mechanism involved in this process is probably different from the mechanism of transcriptional activation of viral genes.

Expression and interactions of human adenovirus oncoproteins

Images

Translation to near the distal end of the penultimate exon is required for normal levels of spliced triosephosphate isomerase mRNA

The translation of human triosephosphate isomerase (TPI) mRNA normally terminates at codon 249 within exon 7, the final exon. Frameshift and nonsense mutations within the TPI gene that cause translation to terminate prematurely at or upstream of codon 189, within exon 6, result in a decreased level of TPI mRNA (I.O. Daar and L.E. Maquat, Mol. Cell. Biol. 8:802-813, 1988). For all mutations in this group, the decrease is to the same extent, i.e., to approximately 20% of the normal level. We show here that a second group of nonsense mutations that cause translation to terminate prematurely at or downstream of codon 208, in exon 6, did not affect TPI mRNA abundance. Deletion analysis demonstrated that the abundance of translationally active TPI mRNA is a function of both the distance and the polarity of the nonsense codon relative to the final intron in TPI pre-mRNA. Our results indicate that if translating ribosomes are unable to progress to at least a certain position within the penultimate exon relative to the final intron, then the level of the corresponding mRNA will be abnormally low. Studies inhibiting RNA synthesis with dactinomycin demonstrated that a block in translation does not affect the half-life of mature TPI mRNA. The simplest interpretation of our data is that the translation of TPI mRNA in the cytoplasm facilitates the splicing of TPI pre-mRNA or the transport of TPI mRNA across the nuclear envelope or both.

Differential requirement for adenovirus type 12 E1A gene products in oncogenic transformation

During the early period of infection, adenovirus type 12 E1A gene is expressed as overlapping, spliced mRNAs of 12 and 13S, which encode in-frame proteins of 235 and 266 amino acid residues (235R and 266R), respectively. To define the functions of these related products in the infection of human cells and transformation of rodent cells, we created single T-to-C transitions at the second base of each mRNA intron which specifically prevent splicing of the respective mRNAs. Mutant pm712 expresses only the 13S mRNA and 266R protein, while pm713 expresses only the 12S mRNA and 235R protein. By using these mutants, we showed that only the larger product is required for growth in human cells, including growth-arrested W138 cells, that the capacity to activate other viral genes (in human cells, at least) lies primarily with that protein, and that the 266R product is not required for autoregulation of its own transcription. In the presence of the 266R protein the 235R product was not required for complete and efficient transformation of a variety of rodent cells or for direct induction of tumors in rats, whereas in its absence the smaller product was insufficient for transformation or tumor induction. Finally, we showed that transformants resulting from infection of rodent cells with pm712 possess a fully-transformed phenotype and are tumorigenic. Previous studies with group C adenoviruses led to the conclusion that both E1A products are required for complete transformation; we conclude that with oncogenic serotype 12, only the 266R product is required for this process.

Translation to near the distal end of the penultimate exon is required for normal levels of spliced triosephosphate isomerase mRNA

The translation of human triosephosphate isomerase (TPI) mRNA normally terminates at codon 249 within exon 7, the final exon. Frameshift and nonsense mutations within the TPI gene that cause translation to terminate prematurely at or upstream of codon 189, within exon 6, result in a decreased level of TPI mRNA (I.O. Daar and L.E. Maquat, Mol. Cell. Biol. 8:802-813, 1988). For all mutations in this group, the decrease is to the same extent, i.e., to approximately 20% of the normal level. We show here that a second group of nonsense mutations that cause translation to terminate prematurely at or downstream of codon 208, in exon 6, did not affect TPI mRNA abundance. Deletion analysis demonstrated that the abundance of translationally active TPI mRNA is a function of both the distance and the polarity of the nonsense codon relative to the final intron in TPI pre-mRNA. Our results indicate that if translating ribosomes are unable to progress to at least a certain position within the penultimate exon relative to the final intron, then the level of the corresponding mRNA will be abnormally low. Studies inhibiting RNA synthesis with dactinomycin demonstrated that a block in translation does not affect the half-life of mature TPI mRNA. The simplest interpretation of our data is that the translation of TPI mRNA in the cytoplasm facilitates the splicing of TPI pre-mRNA or the transport of TPI mRNA across the nuclear envelope or both.

Transcriptional repression of the neu protooncogene by the adenovirus 5 E1A gene products

Amplification/overexpression of the human neu protooncogene has been frequently found in human primary breast and ovarian cancers and is correlated with the number of axillary lymph nodes positive for metastasis in breast cancer patients. Identification of the factors controlling transcription of the neu gene is essential for understanding the mechanisms of neu gene regulation and its role in tumorigenicity. The adenovirus early region 1A (E1A) gene products are pleiotropic transcription regulators of viral and cellular genes and have been identified as a viral suppressor gene for metastasis. Here we demonstrate that transcription of neu can be strongly repressed by the E1A gene products. The 13S and 12S products of E1A gene are effective at repressing neu transcription and the transcriptional repression requires the conserved region 2 of the E1A proteins. The target for E1A repression was localized within a 139-base-pair DNA fragment in the upstream region of the neu promoter. In addition, competition experiments suggest that the sequence TGGAATG, within the 139-base-pair fragment, is an important element for the E1A-induced repression. These results indicate that E1A negatively regulates neu gene expression at the transcriptional level by means of a specific DNA element.

The adenovirus E4 17-kilodalton protein complexes with the cellular transcription factor E2F, altering its DNA-binding properties and stimulating E1A-independent accumulation of E2 mRNA

E2F is a cellular DNA-binding factor. Its binding activity is changed within adenovirus-infected cells so that it binds cooperatively to pairs of properly spaced and oriented E2F recognition sites. In the work described in this report, the conversion to cooperative binding was shown to require the adenovirus E4 17-kilodalton (kDa) polypeptide. Mutant viruses carrying alterations within the E4 17-kDa coding region failed to generate the infection-specific, cooperatively binding form of E2F. It was possible to alter E2F from uninfected cells so that it bound cooperatively by incubation with a partially purified fraction obtained from infected cells. The E4 17-kDa protein copurified with this activity and was also found to be present in a complex containing E2F. Consistent with its ability to alter the binding of E2F to its recognition sites within the E2 promoter, the E4 17-kDa polypeptide contributed to maximal expression of E2 mRNAs in some cell types. Its ability to enhance E2 transcription did not require expression of the E1A transactivator protein. These results are consistent with a model which proposes that the E4 17-kDa polypeptide binds to the cellular E2F factor, altering its binding behavior and thereby enhancing its ability to stimulate transcription.

Nuclear factor EF-1A binds to the adenovirus E1A core enhancer element and to other transcriptional control regions

We have identified a cellular enhancer-binding protein, present in nuclear extracts prepared from human and rodent cells, that binds to the adenovirus E1A enhancer element I sequence. The factor has been termed EF-1A, for enhancer-binding factor to the E1A core motif. EF-1A was found to bind to two adjacent, related sequence motifs in the E1A enhancer region (termed sites A and B). The binding of EF-1A to these adjacent sites, or to synthetic dimerized sites of either motif, was cooperative. The cooperative binding of EF-1A to these sites was not subject to strict spacing constraints. EF-1A also bound to related sequences upstream of the E1A enhancer region and in the polyomavirus and adenovirus E4 enhancer regions. The EF-1A-binding region in the E1A enhancer stimulated expression of a linked gene in human 293 cells when multimerized. Based on the contact sites for EF-1A binding determined by chemical interference assays, this protein appears to be distinct from any previously characterized nuclear binding protein.

Topoisomerase I and II cleavage of adenovirus DNA in vivo: both topoisomerase activities appear to be required for adenovirus DNA replication

Sites of topoisomerase I and II cleavage across large portions of the adenovirus type 5 genome were mapped by using the drugs camptothecin and VM26, respectively. These drugs prolong the half-lives of the covalent DNA-protein intermediates in which the DNA is transiently cleaved, and so treatment with protein denaturants after exposure to the drugs leads to DNA strand scission at the site of topoisomerase cleavage. Strong topoisomerase II cleavage sites occurred in clusters throughout the regions examined, including both transcribed regions and transcriptional control regions. The efficiency of topoisomerase II cleavage increased as the rate of adenovirus DNA replication increased and then decreased with the decreasing rate of replication late in the infection cycle. The increase was not dependent on expression of the E1A gene, whose products activate transcription of the early viral genes. Positions of topoisomerase II cleavage sites did not vary during the infection. Topoisomerase I cleavage sites were also found throughout the examined regions, with the strongest sites occurring near the ends of the transcription units. Topoisomerase I cleavage in the E1 region occurred much more frequently than topoisomerase II cleavage, was not dependent on E1A gene expression, and remained at a similar level from the early viral phase into the late viral phase. Treatment of infected cells with either drug prevented efficient replication of adenovirus DNA. Inhibition of topoisomerase I activity led to an immediate cessation of adenovirus DNA replication, while inhibition of topoisomerase II blocked replication only after completion of approximately one additional round.

Nuclear factor EF-1A binds to the adenovirus E1A core enhancer element and to other transcriptional control regions

We have identified a cellular enhancer-binding protein, present in nuclear extracts prepared from human and rodent cells, that binds to the adenovirus E1A enhancer element I sequence. The factor has been termed EF-1A, for enhancer-binding factor to the E1A core motif. EF-1A was found to bind to two adjacent, related sequence motifs in the E1A enhancer region (termed sites A and B). The binding of EF-1A to these adjacent sites, or to synthetic dimerized sites of either motif, was cooperative. The cooperative binding of EF-1A to these sites was not subject to strict spacing constraints. EF-1A also bound to related sequences upstream of the E1A enhancer region and in the polyomavirus and adenovirus E4 enhancer regions. The EF-1A-binding region in the E1A enhancer stimulated expression of a linked gene in human 293 cells when multimerized. Based on the contact sites for EF-1A binding determined by chemical interference assays, this protein appears to be distinct from any previously characterized nuclear binding protein.

The regions important for the activator and repressor functions of herpes simplex virus type 1 alpha protein ICP27 map to the C-terminal half of the molecule

The herpes simplex virus type 1 (HSV-1) alpha or immediate-early proteins ICP4 (IE175), ICP0 (IE110), and ICP27 (IE63) are trans-acting proteins which affect HSV-1 gene expression. We previously showed that ICP27 in combination with ICP4 and ICP0 could act as a repressor or an activator in transfection assays, depending on the target gene (R. E. Sekulovich, K. Leary, and R. M. Sandri-Goldin, J. Virol. 62:4510-4522, 1988). To investigate the regions of the ICP27 protein which specify these functions, we constructed a series of in-frame insertion and deletion mutants in the ICP27 gene. These mutants were analyzed in transient expression assays for the ability to repress or to activate two different target genes. The target plasmids used consisted of the promoter regions from the HSV-1 beta or early gene which encodes thymidine kinase and from the beta-gamma or leaky late gene. VP5, which encodes the major capsid protein, each fused to the chloramphenicol acetyltransferase gene. Our previous studies showed that induction of pTK-CAT expression by ICP4 and ICP0 was repressed by ICP27, whereas the stimulation of pVP5-CAT expression seen with ICP4 and ICP0 was significantly increased when ICP27 was also added. In this study, a series of transfection assays was performed with each of the ICP27 mutant plasmids in combination with plasmids containing the ICP4 and ICP0 genes with each target. The results of these experiments showed that mutants containing insertions or deletions in the region from amino acids 262 to 406 in the carboxy-terminal half of the protein were unable to stimulate expression of pVP5-CAT but were able to repress induction of pTK-CAT activity by ICP4 and ICP0. Mutants in the carboxy-terminal 78 amino acids lost both activities; that is, these mutants did not show repression of pTK-CAT activity or stimulation of pVP5-CAT activity, whereas mutants in the hydrophilic amino-terminal half of ICP27 were able to perform both functions. These results show that the carboxy-terminal half of ICP27 is important for the activation and repression functions. Furthermore, the carboxy-terminal 62 amino acids are required for the repressor activity, because mutants with this region intact were able to repress. Analysis of the DNA sequence showed that there are a number of cysteine and histidine residues encoded by this region which have some similarity to zinc finger metal-binding regions found in other eucaryotic regulatory proteins. These results suggest that the structural integrity of this region is important for the function of ICP27.

Multiple transcription factor binding sites mediate adenovirus E1A transactivation

We studied the response of simple synthetic promoter regions to transactivation by the adenovirus early region 1A (E1A) protein. Binding sites for one or two host cell transcription factors were substituted for the E1B promoter region in reconstructed virus mutants, and the response to E1A transactivation was assayed during the early phase of infection. We found that a single CREB/ATF binding site resulted in a surprisingly strong promoter which responded to E1A. A CREB/ATF binding site placed upstream of the E1B TATA box behaved much like the wild-type E1B promoter, which is composed of a single Sp1 binding site plus a TATA box. A single E2F binding site resulted in an extremely weak promoter which did not respond to E1A, much like a single Sp1 site. Two E2F sites in an inverted orientation with the same spacing as in the adenovirus type 2 E2 early promoter produced a strong, E1A-responsive promoter. Substitution of the E4 TATA box region for the E1B TATA box region produced a promoter about five times stronger than the wild-type E1B promoter in the absence of E1A. However, the E4 TATA box substitution did not respond significantly to E1A transactivation. These results directly demonstrate that many different transcription factor binding sites, including the E1B TATA box, a CREB/ATF binding site, and two E2F sites, can mediate E1A transactivation. Other transcription factor binding sites cannot mediate an E1A response; these other sites include the E4 TATA box, a single Sp1 binding site, and a single E2F binding site. Implications of these findings for the mechanism of E1A transactivation are discussed.

Upstream DNA sequences determine different autoregulatory responses of the adenovirus types 5 and 3 E1A promoters

Adenovirus types 5 and 3 (Ad5 and Ad3), two human adenovirus serotypes of evolutionarily divergent subgroups, show very different levels of E1A gene expression early after infection of permissive cells. Since adenovirus E1A gene expression is known to be transcriptionally autoregulated, we have investigated the difference between Ad3 and Ad5 by monitoring transient expression of a reporter gene under transcriptional control of the E1A promoter of Ad5 or Ad3. There was only a modest difference between the basal levels of transcription driven by these two E1A promoters. This difference was amplified from 10 to 100 times by the different net responses of the E1A promoters to concomitantly expressed E1A genes. Each promoter had a characteristic net response to positive and negative regulation by E1A gene products. The Ad5 E1A promoter was more strongly repressed, whereas the Ad3 E1A promoter was more strongly activated by E1A gene products. Experiments with a chimeric Ad5/3 E1A promoter indicated that these different autoregulatory responses are determined by DNA sequences which are more than 50 base pairs upstream from the E1A transcriptional start site. A plausible target DNA sequence for positive and negative autoregulation by E1A gene products is discussed.

The herpes simplex virus type 1 alpha protein ICP27 can act as a trans-repressor or a trans-activator in combination with ICP4 and ICP0

The herpes simplex virus type 1 (HSV-1) alpha proteins ICP4, ICP0, and ICP27 are trans-acting proteins which affect HSV-1 gene expression. To investigate potential interactions between these alpha products and to determine the specificity of action of the alpha proteins in combination with each other compared with their activities individually, we performed a series of transient-expression assays. In these assays we used plasmids containing the alpha genes encoding ICP4, ICP0, and ICP27 either singly or in combination as effectors and HSV-1 genes of different kinetic classes and heterologous genes as targets. The HSV-1 targets consisted of promoter-regulatory domains from alpha (ICP0 and ICP27), beta (thymidine kinase and alkaline exonuclease), beta-gamma (glycoprotein D, glycoprotein B, and VP5), and gamma (glycoprotein C) genes, each fused to the chloramphenicol acetyltransferase (CAT) gene. The heterologous target genes consisted of the simian virus 40 early promoter with enhancer and the Rous sarcoma virus long terminal repeat promoter and enhancer each fused to the CAT gene. Target promoter activity was measured by the assay of CAT activity in extracts of transfected cells and by Northern (RNA) blot hybridization of CAT mRNA. The results of these experiments showed that ICP4 activated only HSV-1 target genes, whereas ICP0 activated all of the targets and ICP27 had little effect on any of the targets. ICP4 and ICP0 had a synergistic effect when inducing HSV-1 targets, but they did not have this effect on the heterologous targets pSV2-CAT or pRSV-CAT. In fact, lower levels of CAT activity and CAT mRNA were found in the presence of both effectors than with ICP0 alone. Most interestingly, although the effector plasmid containing the ICP27 gene had little effect on its own, two different and marked effects depending on the target were observed when ICP27 was combined with ICP4 or ICP0 or both. A trans-repression of the induction seen with ICP4 and ICP0 was found when ICP27 was present in the transfections with pSV2-CAT, pRSV-CAT, pICP0-CAT, pICP27-CAT, pTK-CAT, pgD-CAT, pgB-CAT, and pgC-CAT. This resulted in CAT activity levels which were similar to or lower than the basal level of expression of the target genes in the absence of effector plasmids. This trans-repression occurred over a wide range of concentrations of input ICP27 plasmid. In contrast to this repressive effect of ICP27, a trans-activation was seen when ICP4, ICP0, and ICP27 plasmids were combined in transfections with pAE-CAT and pVP5-CAT as targets. This trans-activation also occurred over a 10-fold range of input ICP27 plasmid. These results suggest that ICP27 can facilitate both down

Factors responsible for the higher transcriptional activity of extracts of adenovirus-infected cells fractionate with the TATA box transcription factor

Extracts of adenovirus-infected HeLa cells have 5- to 10-fold-higher activity for transcription from the major late promoter in vitro than do extracts of mock-infected or E1A mutant-infected cells (K. Leong and A. J. Berk, Proc. Natl. Acad. Sci. USA 83:5844-5848, 1986). In this study, we analyzed extracts from mock-infected cells and from cells infected with an E1A mutant, pm975, which expresses principally the large E1A protein responsible for the stimulation of transcription. These extracts were fractionated by phosphocellulose chromatography, a procedure which separates factors required for transcription from this promoter (J. D. Dignam, B. S. Shastry, and R. G. Roeder, Methods Enzymol. 101:582-589, 1983), allowing the quantitative assay of individual factors (M. Samuels, A. Fire, and P. A. Sharp, J. Biol. Chem. 257:14419-14427, 1982). Fractions eluted with 0.04, 0.35, and 0.6 M KCl, which contained RNA polymerase II, the upstream factor MLTF, and three general polymerase II transcription factors, had similar activities when prepared from virus-infected or from mock-infected cells. The sequence-specific DNA-binding activity of MLTF was also similar in the virus-infected- and mock-infected-cell extracts. In contrast, the 1.0 M KCl fraction prepared from virus-infected cells consistently exhibited activity severalfold higher than that of the equivalent fraction prepared in parallel from mock-infected cells. E1A protein eluted principally (greater than 80%) in the 0.35 M KCl fraction. Results of others (M. Sawadogo and R. G. Roeder, Cell 43:165-175, 1985) have shown that the 1.0 M KCl fraction, containing 2 to 5% of the unfractionated protein extract, contains a factor which binds specifically to the major late promoter TATA box. These results, together with a recent genetic analysis of the E1B promoter which demonstrated that the TATA box was required for its efficient transcriptional activation (transactivation) by E1A (L. Wu, D. S. E. Rosser, M. Schmidt, and A. J. Berk, Nature (London) 326:512-515, 1987), are consistent with the model that E1A protein indirectly activates the TATA box transcription factor. Consistent with this model was the finding that mutants of the major late promoter containing only the TATA box and cap site region were transcribed at higher rates with extracts from virus-infected cells than with extracts from mock-infected cells. Other models consistent with the results are also discussed.

Induced heat shock mRNAs escape the nucleocytoplasmic transport block in adenovirus-infected HeLa cells

Under conditions in which cytoplasmic accumulation of HeLa cell mRNAs has been blocked by adenovirus infection, hsp70 family mRNAs are transported from the nucleus to the cytoplasm at near normal efficiency subsequent to heat shock. Heat shock does not reverse the general virus-induced block to host cell mRNA transport. The heat shock mRNAs are translated within the cytoplasm of the infected cell but at substantially reduced efficiency compared with that of uninfected cells. Thus, the hsp70 family of mRNAs can escape the transport block but not the translational block instituted late after adenovirus infection. The beta-tubulin gene family is induced by the viral E1A gene after infection, and its mRNAs also accumulate in the cytoplasmic compartment. Given these two examples, it seems likely that the process of transcriptional induction allows the resulting mRNA to escape the viral block of transport.

Factors responsible for the higher transcriptional activity of extracts of adenovirus-infected cells fractionate with the TATA box transcription factor

Extracts of adenovirus-infected HeLa cells have 5- to 10-fold-higher activity for transcription from the major late promoter in vitro than do extracts of mock-infected or E1A mutant-infected cells (K. Leong and A. J. Berk, Proc. Natl. Acad. Sci. USA 83:5844-5848, 1986). In this study, we analyzed extracts from mock-infected cells and from cells infected with an E1A mutant, pm975, which expresses principally the large E1A protein responsible for the stimulation of transcription. These extracts were fractionated by phosphocellulose chromatography, a procedure which separates factors required for transcription from this promoter (J. D. Dignam, B. S. Shastry, and R. G. Roeder, Methods Enzymol. 101:582-589, 1983), allowing the quantitative assay of individual factors (M. Samuels, A. Fire, and P. A. Sharp, J. Biol. Chem. 257:14419-14427, 1982). Fractions eluted with 0.04, 0.35, and 0.6 M KCl, which contained RNA polymerase II, the upstream factor MLTF, and three general polymerase II transcription factors, had similar activities when prepared from virus-infected or from mock-infected cells. The sequence-specific DNA-binding activity of MLTF was also similar in the virus-infected- and mock-infected-cell extracts. In contrast, the 1.0 M KCl fraction prepared from virus-infected cells consistently exhibited activity severalfold higher than that of the equivalent fraction prepared in parallel from mock-infected cells. E1A protein eluted principally (greater than 80%) in the 0.35 M KCl fraction. Results of others (M. Sawadogo and R. G. Roeder, Cell 43:165-175, 1985) have shown that the 1.0 M KCl fraction, containing 2 to 5% of the unfractionated protein extract, contains a factor which binds specifically to the major late promoter TATA box. These results, together with a recent genetic analysis of the E1B promoter which demonstrated that the TATA box was required for its efficient transcriptional activation (transactivation) by E1A (L. Wu, D. S. E. Rosser, M. Schmidt, and A. J. Berk, Nature (London) 326:512-515, 1987), are consistent with the model that E1A protein indirectly activates the TATA box transcription factor. Consistent with this model was the finding that mutants of the major late promoter containing only the TATA box and cap site region were transcribed at higher rates with extracts from virus-infected cells than with extracts from mock-infected cells. Other models consistent with the results are also discussed.

The abundance and in vitro DNA binding of three cellular proteins interacting with the adenovirus EIIa early promoter are not modified by the EIa gene products

Specific protein binding on the EIa-inducible adenovirus EIIa early (EIIaE) promoter was analyzed by the sensitive electrophoretic band-shift assay and by protection against DNase I digestion. Three factors were identified, and precise mapping of the cognate-binding sites revealed their correspondence to promoter elements essential for constitutive EIIaE transcription. One binds to the major upstream element located between -82 and -64 (with respect to the major EIIaE cap site), another appears to interact with sequences on either side of this region, and the last one binds to an element located further upstream. Comparison of the binding activities of the factors present in extracts from cells infected with wild-type adenovirus (adenovirus type 5) or with the EIa deletion mutant dl312 did not reveal striking differences. Not only were the general binding patterns indistinguishable, but the concentration of each of the identified factors as well as their affinity for the cognate-binding sites were unchanged. Our results suggest that the EIa-mediated activation of the EIIaE transcription complexes involves appropriate interactions between transcription factors, rather than their increased binding to DNA.

Adenovirus E1A gene autorepression: revertants of an E1A promoter mutation encode altered E1A proteins

Revertants have been isolated from Ad3hr15, a mutant of human adenovirus type 3 that carries a defective E1A promoter. Transcription of these revertant E1A genes is restored--from nil for Ad3hr15 mutant to levels exceeding that of the wild-type virus. The mutant Ad3hr15 virus and the revertants all have an aberrant E1A promoter that contains two short tandem duplications of viral DNA sequence. The E1A gene-coding region of the mutant is the same as that for wild-type adenovirus type 3, whereas the revertants are characterized by short in-frame deletions within the 5' exon region of their E1A genes. Location of these reverting, second-site deletions is discussed in relation to E1A gene autoregulation and the evolved diversity of E1A-related oncogenic potential among different human adenoviruses.

Induced heat shock mRNAs escape the nucleocytoplasmic transport block in adenovirus-infected HeLa cells

Under conditions in which cytoplasmic accumulation of HeLa cell mRNAs has been blocked by adenovirus infection, hsp70 family mRNAs are transported from the nucleus to the cytoplasm at near normal efficiency subsequent to heat shock. Heat shock does not reverse the general virus-induced block to host cell mRNA transport. The heat shock mRNAs are translated within the cytoplasm of the infected cell but at substantially reduced efficiency compared with that of uninfected cells. Thus, the hsp70 family of mRNAs can escape the transport block but not the translational block instituted late after adenovirus infection. The beta-tubulin gene family is induced by the viral E1A gene after infection, and its mRNAs also accumulate in the cytoplasmic compartment. Given these two examples, it seems likely that the process of transcriptional induction allows the resulting mRNA to escape the viral block of transport.

The abundance and in vitro DNA binding of three cellular proteins interacting with the adenovirus EIIa early promoter are not modified by the EIa gene products

Specific protein binding on the EIa-inducible adenovirus EIIa early (EIIaE) promoter was analyzed by the sensitive electrophoretic band-shift assay and by protection against DNase I digestion. Three factors were identified, and precise mapping of the cognate-binding sites revealed their correspondence to promoter elements essential for constitutive EIIaE transcription. One binds to the major upstream element located between -82 and -64 (with respect to the major EIIaE cap site), another appears to interact with sequences on either side of this region, and the last one binds to an element located further upstream. Comparison of the binding activities of the factors present in extracts from cells infected with wild-type adenovirus (adenovirus type 5) or with the EIa deletion mutant dl312 did not reveal striking differences. Not only were the general binding patterns indistinguishable, but the concentration of each of the identified factors as well as their affinity for the cognate-binding sites were unchanged. Our results suggest that the EIa-mediated activation of the EIIaE transcription complexes involves appropriate interactions between transcription factors, rather than their increased binding to DNA.

Adenovirus E1A 12S protein induces DNA synthesis and proliferation in primary epithelial cells in both the presence and absence of serum

Infection of primary baby rat kidney (BRK) cells with an adenovirus that carries an E1A 12S cDNA in place of the normal E1A region (adenovirus 5 [Ad5] 12S) resulted in the induction of cellular DNA synthesis and proliferation of the epithelial cells in the population, even in the absence of serum. Increased cellular DNA synthesis was first detectable by 12 h after infection and was maintained at a 10- to 20-fold higher level than in mock-infected cells. By 5 days after infection there was a 10-fold-greater number of 12S virus-infected BRK cells. These infected BRK cells retained many of their normal epithelial cell characteristics and were not transformed. The expression of the E1A 12S protein(s) occurred early after infection. There was no induction of adenoviral gene expression or viral DNA replication in these cells. The early effects of a fully transforming gene product(s) were also examined. The Ad5-simian virus 40 hybrid virus, Ad5.SVR4, in which the early region of simian virus 40 has replaced the E1 region of Ad5, was used to infect BRK cells. The kinetics of expression of the T antigens were similar to those of the 12S polypeptides. Infection with Ad5.SV4 also resulted in the induction of cellular DNA synthesis and cell proliferation at levels similar to those observed with the 12S virus. However, infection with Ad5.SVR4 resulted in cells that had lost some of their epithelial cell characteristics and were fully transformed. Thus, although the early cellular events induced by the two genes were similar, they did not yield the same final cellular phenotype.

Regulation of adenovirus gene expression in human WI38 cells by an E1B-encoded tumor antigen

Adenovirus mutants carrying alterations in the gene encoding the E1B 19-kilodalton tumor antigen (19K protein) cause enhanced cytopathic effect (cyt phenotype) and the degradation of host-cell chromosomal DNA (deg phenotype) upon infection of human HeLa or KB cells. Furthermore, E1B 19K gene mutant viruses are defective for cellular transformation. We report that these mutant viruses possess a host-range phenotype for growth in human cells. In human HeLa cells the mutant viruses grew to the same levels as the wild-type virus, but they were severely defective for growth in KB cells. In human WI38 cells, the E1B 19K gene mutant viruses had a substantial growth advantage over the wild-type virus, yielding 500-fold-higher titers. Viral DNA synthesis was reduced 10- to 20-fold in WI38 cells infected with the wild-type virus relative to that synthesized by the E1B mutant viruses. Viral early and late protein synthesis was similarly reduced in wild type- relative to mutant-infected cells. These reduced levels of early gene expression in wild-type virus-infected cells were paralleled by comparably reduced levels of early cytoplasmic mRNA. The primary cause of this host-range phenotype appeared at the level of early gene transcription, since transcription of viral early genes in the mutant-infected cells was substantially greater than levels found in cells infected with the wild-type virus. These results implicate the E1B 19K tumor antigen in the regulation of adenovirus early gene expression. Specifically, the E1B 19K protein directly or indirectly exerts a negative effect on early gene transcription accounting for efficient gene expression from the E1B mutant viruses in WI38 cells. Based on these findings it is probable that the cyt and deg phenotypes observed in mutant-infected HeLa and KB cells are the result of the pleiotropic effect of this altered gene regulation.

Cellular promoters incorporated into the adenovirus genome: effects of viral regulatory elements on transcription rates and cell specificity of albumin and beta-globin promoters

In the accompanying paper (Friedman et al., Mol. Cell. Biol. 6:3791-3797, 1986), hepatoma-specific expression of the rat albumin promoter within the adenovirus genome was demonstrated. However, the rate of transcription was very low compared with that of the endogenous chromosomal albumin gene. Here we show that in hepatoma cells the adenovirus E1A enhancer, especially in the presence of E1A protein, greatly stimulates transcription from the albumin promoter but not the mouse beta-globin promoter. This enhancer-dependent stimulation did not occur in myeloma cells in which a virus containing a immunoglobulin promoter and enhancer did function. These experiments suggest a limited distribution in cultured differentiated cells of cell-specific transcription factors. However, either the regulation of such cell-specific factors breaks down in other cultured cells, or strictly cell-specific factors are not at play in controlling cell-specific transcription, because HeLa cells could transcribe the albumin promoter from the same start site about 10% as well as hepatomas could and 293 cells could transcribe both albumin and globin promoters.

Identification of separate domains in the adenovirus E1A gene for immortalization activity and the activation of virus early genes

The transformation and early adenovirus gene transactivation functions of the E1A region were analyzed with deletion and point mutations. Deletion of amino acids from position 86 through 120 had little effect on the lytic or transforming functions of the E1A products, while deletion of amino acids from position 121 through 150 significantly impaired both functions. The sensitivity of the transformation function to alterations in the region from amino acid position 121 to 150 was further indicated by the impairment of transforming activity resulting from single amino acid substitutions at positions 124 and 135. Interestingly, conversion of a cysteine residue at position 124 to glycine severely impaired the transformation function without affecting the early adenovirus gene activating functions. Single amino acid substitutions in a different region of the E1A gene had the converse effect. All the mutants produced polypeptides of sufficient stability to be detected by Western immunoblot analysis. The single amino acid substitutions at positions 124 and 135, although impairing the transformation functions, did not detectably alter the formation of the higher-apparent-molecular-weight forms of the E1A products.

Regulation of adenovirus gene expression in human WI38 cells by an E1B-encoded tumor antigen

Adenovirus mutants carrying alterations in the gene encoding the E1B 19-kilodalton tumor antigen (19K protein) cause enhanced cytopathic effect (cyt phenotype) and the degradation of host-cell chromosomal DNA (deg phenotype) upon infection of human HeLa or KB cells. Furthermore, E1B 19K gene mutant viruses are defective for cellular transformation. We report that these mutant viruses possess a host-range phenotype for growth in human cells. In human HeLa cells the mutant viruses grew to the same levels as the wild-type virus, but they were severely defective for growth in KB cells. In human WI38 cells, the E1B 19K gene mutant viruses had a substantial growth advantage over the wild-type virus, yielding 500-fold-higher titers. Viral DNA synthesis was reduced 10- to 20-fold in WI38 cells infected with the wild-type virus relative to that synthesized by the E1B mutant viruses. Viral early and late protein synthesis was similarly reduced in wild type- relative to mutant-infected cells. These reduced levels of early gene expression in wild-type virus-infected cells were paralleled by comparably reduced levels of early cytoplasmic mRNA. The primary cause of this host-range phenotype appeared at the level of early gene transcription, since transcription of viral early genes in the mutant-infected cells was substantially greater than levels found in cells infected with the wild-type virus. These results implicate the E1B 19K tumor antigen in the regulation of adenovirus early gene expression. Specifically, the E1B 19K protein directly or indirectly exerts a negative effect on early gene transcription accounting for efficient gene expression from the E1B mutant viruses in WI38 cells. Based on these findings it is probable that the cyt and deg phenotypes observed in mutant-infected HeLa and KB cells are the result of the pleiotropic effect of this altered gene regulation.

Cellular promoters incorporated into the adenovirus genome: effects of viral regulatory elements on transcription rates and cell specificity of albumin and beta-globin promoters

In the accompanying paper (Friedman et al., Mol. Cell. Biol. 6:3791-3797, 1986), hepatoma-specific expression of the rat albumin promoter within the adenovirus genome was demonstrated. However, the rate of transcription was very low compared with that of the endogenous chromosomal albumin gene. Here we show that in hepatoma cells the adenovirus E1A enhancer, especially in the presence of E1A protein, greatly stimulates transcription from the albumin promoter but not the mouse beta-globin promoter. This enhancer-dependent stimulation did not occur in myeloma cells in which a virus containing a immunoglobulin promoter and enhancer did function. These experiments suggest a limited distribution in cultured differentiated cells of cell-specific transcription factors. However, either the regulation of such cell-specific factors breaks down in other cultured cells, or strictly cell-specific factors are not at play in controlling cell-specific transcription, because HeLa cells could transcribe the albumin promoter from the same start site about 10% as well as hepatomas could and 293 cells could transcribe both albumin and globin promoters.

Identification of separate domains in the adenovirus E1A gene for immortalization activity and the activation of virus early genes

The transformation and early adenovirus gene transactivation functions of the E1A region were analyzed with deletion and point mutations. Deletion of amino acids from position 86 through 120 had little effect on the lytic or transforming functions of the E1A products, while deletion of amino acids from position 121 through 150 significantly impaired both functions. The sensitivity of the transformation function to alterations in the region from amino acid position 121 to 150 was further indicated by the impairment of transforming activity resulting from single amino acid substitutions at positions 124 and 135. Interestingly, conversion of a cysteine residue at position 124 to glycine severely impaired the transformation function without affecting the early adenovirus gene activating functions. Single amino acid substitutions in a different region of the E1A gene had the converse effect. All the mutants produced polypeptides of sufficient stability to be detected by Western immunoblot analysis. The single amino acid substitutions at positions 124 and 135, although impairing the transformation functions, did not detectably alter the formation of the higher-apparent-molecular-weight forms of the E1A products.

Positive and negative autoregulation of the adeno-associated virus type 2 genome

The defective human parvovirus, adeno-associated virus (AAV), requires multiple functions provided by a coinfecting helper virus for viral replication. In addition, it has recently been shown that at least one AAV gene is also required for AAV DNA replication. In this paper, we investigate the autoregulation of the AAV genome by analyzing the expression of mutant AAV genomes upon transfection into adenovirus-infected human cells. Evidence is presented which indicates that the AAV genome regulates its own gene expression in at least two ways. First, either the AAV p5 gene or both the p5 and p19 genes appear to encode a trans activator of AAV transcription. Frameshift mutations within the p5 or p19 gene severely inhibited the synthesis and accumulation of all AAV transcripts. The defective accumulation of transcripts could be complemented in trans, in a manner independent of DNA replication, by cotransfection with a capsid deletion mutant. Second, evidence is presented which suggests that the p5 and p19 genes contain negative cis-active regulatory elements. Deletion of sequences within the p5 and p19 genes enhanced the accumulation of the p5 transcript in cis upon complementation with an AAV capsid deletion mutant, whereas certain deletions enhanced p40 RNA accumulation in the absence of trans activation by the p5 gene.

Promiscuous trans activation of gene expression by an Epstein-Barr virus-encoded early nuclear protein

We identified an Epstein-Barr virus (EBV) gene product which functions in transient-expression assays as a nonspecific trans activator. In Vero cells, cotransfection of the BglII J DNA fragment of EBV together with recombinant constructs containing the bacterial chloramphenicol acetyltransferase (CAT) gene gave up to a 100-fold increased expression of CAT activity over that in cells transfected with the recombinant CAT constructs alone. The BglII J fragment acted promiscuously, in that increased CAT synthesis was observed regardless of whether the promoter sequences driving the CAT gene were of EBV, simian virus 40, adenovirus, or herpes simplex virus origin. Cleavage of cloned BglII-J plasmid DNA before transfection revealed that activation was dependent upon the presence of an intact BMLF1 open reading frame. This was confirmed with subclones of BglII-J and with hybrid promoter-open reading frame constructs. This region of the genome is also present in the rearranged P3HR-1-defective DNA species, and defective DNA clones containing these sequences produced a similar activation of CAT expression in cotransfection experiments. The heterogeneous 45-60-kilodalton polypeptide product of BMLF1 may play an important regulatory role in expression of lytic-cycle proteins in EBV-infected lymphocytes.

Adenovirus early region 1A protein increases the number of template molecules transcribed in cell-free extracts

Protein encoded by adenovirus early region 1A (E1A) stimulates transcription from adenovirus promoters in vivo. Here we show that this effect can be observed in vitro. In a run-off transcription assay from the adenovirus serotype 2 (Ad2) major late promoter, extracts prepared 20 hr postinfection were 5-15 times more active than mock-infected-cell extracts prepared in parallel. Similar results were observed for in vitro transcription from the protein IX and E3 adenovirus promoters, whereas a 2-fold increase was observed for the human beta-globin promoter. The increased activities of infected-cell extracts did not depend on the expression of viral late proteins or the small E1A-encoded proteins but did require expression of the large E1A protein. These results are consistent with the large E1A protein stimulating transcription in vitro as it does in vivo. By limiting in vitro transcription to one initiation per template, we found that the higher activity of an infected-cell extract was due to an increase in the number of templates transcribed. These results suggest that the large E1A protein either increases the number of active transcription factors in infected cells or facilitates the interaction of cellular transcription factors with promoter DNA.