Study of the E2 gene product of the cottontail rabbit papillomavirus reveals a common mechanism of transactivation among papillomaviruses

The rabbit papillomavirus model: a valuable tool to study viral-host interactions

Cottontail rabbit papillomavirus (CRPV) was the first DNA virus shown to be tumorigenic. The virus has since been renamed and is officially known as Sylvilagus floridanus papillomavirus 1 (SfPV1). Since its inception as a surrogate preclinical model for high-risk human papillomavirus (HPV) infections, the SfPV1/rabbit model has been widely used to study viral-host interactions and has played a pivotal role in the successful development of three prophylactic virus-like particle vaccines. In this review, we will focus on the use of the model to gain a better understanding of viral pathogenesis, gene function and host immune responses to viral infections. We will discuss the application of the model in HPV-associated vaccine testing, in therapeutic vaccine development (using our novel HLA-A2.1 transgenic rabbits) and in the development and validation of novel anti-viral and anti-tumour compounds. Our goal is to demonstrate the role the SfPV1/rabbit model has played, and continues to play, in helping to unravel the intricacies of papillomavirus infections and to develop tools to thwart the disease. This article is part of the theme issue 'Silent cancer agents: multi-disciplinary modelling of human DNA oncoviruses'.

A transactivator function of cottontail rabbit papillomavirus e2 is essential for tumor induction in rabbits

Infection of domestic rabbits with cottontail rabbit papillomavirus (CRPV) causes local papillomas which progress to carcinomas in more than 80% of cases. This animal model system therefore allows the identification of molecular mechanisms required for the induction and progression of epithelial tumors. The viral E2 protein stimulates both viral DNA replication and transcription, and these functions can be genetically separated. We introduced the respective mutations into CRPV E2 and found, in line with published data for other papillomavirus E2 proteins, that mutation of the highly conserved amino acid 37 or 73 resulted in replication-competent but transactivation-deficient E2 proteins, whereas E2 proteins with mutations at residue 39 were replication deficient and transactivation competent. The R37A, I73L, and I73A E2 mutants, showing a loss of transactivation function, and the R37K E2 mutant, which is still transactivation competent, were introduced into the whole genome of CRPV, which was then injected into the skin of rabbits. Strikingly, the ability to induce tumors within 6 weeks was abolished by each of the E2 mutations, in contrast to the tumor induction rate (93%) obtained with wild-type CRPV DNA. Two small papillomas induced by mutant E2 I73A CRPV DNA appeared as late as 12 or 24 weeks postinjection, were significantly smaller, and showed no further extension of growth. These data suggest that functionally conserved amino acids in the transactivation domain of E2 are also required for the induction and growth of epithelial tumors in rabbits infected with CRPV.

Ubiquitin-fused and/or multiple early genes from cottontail rabbit papillomavirus as DNA vaccines

Human papillomavirus (HPV) vaccines have the potential to prevent cervical cancer by preventing HPV infection or treating premalignant disease. We previously showed that DNA vaccination with the cottontail rabbit papillomavirus (CRPV) E6 gene induced partial protection against CRPV challenge and that the vaccine's effects were greatly enhanced by priming with granulocyte-macrophage colony-stimulating factor (GM-CSF). In the present study, two additional strategies for augmenting the clinical efficacy of CRPV E6 vaccination were evaluated. The first was to fuse a ubiquitin monomer to the CRPV E6 protein to enhance antigen processing and presentation through the major histocompatibility complex class I pathway. Rabbits vaccinated with the wild-type E6 gene plus GM-CSF or with the ubiquitin-fused E6 gene formed significantly fewer papillomas than the controls. The papillomas also required a longer time to appear and grew more slowly. Finally, a significant proportion of the papillomas subsequently regressed. The ubiquitin-fused E6 vaccine was significantly more effective than the wild-type E6 vaccine plus GM-CSF priming. The second strategy was to vaccinate with multiple CRPV early genes to increase the breadth of the CRPV-specific response. DNA vaccines encoding the wild-type CRPV E1-E2, E6, or E7 protein were tested alone and in all possible combinations. All vaccines and combinations suppressed papilloma formation, slowed papilloma growth, and stimulated subsequent papilloma regression. Finally, the two strategies were merged and a combination DNA vaccine containing ubiquitin-fused versions of the CRPV E1, E2, and E7 genes was tested. This last vaccine prevented papilloma formation at all challenge sites in all rabbits, demonstrating complete protection.

Epidemiologic and mucosal immunologic aspects of HPV infection and HPV-related cervical neoplasia in the lower female genital tract: a review

Human papillomavirus (HPV) infections are known to play an important role in the pathogenesis of cervical neoplasia. Considering the morbidity and mortality of cervical cancer, infection with HPV can be regarded as a worldwide problem, especially in developing countries. Currently, many studies focus on the development of both prophylactic and therapeutic HPV vaccines. Crucial for these vaccination protocols to be successful is that they will result in a long-lasting ability to generate an immune response that will eliminate the virus. HPV transmission and subsequent infection is a local event in the lower female genital tract and therefore the efficacy of vaccines against this locally transmitted infection can be best assessed by parameters of local immunity. In this review we describe both the epidemiology of HPV-related cervical neoplasia and the general aspects of mucosal immunity in the female genital tract while focusing on the local humoral immunity in HPV-related cervical neoplasia.

Granulocyte-macrophage colony-stimulating factor priming plus papillomavirus E6 DNA vaccination: effects on papilloma formation and regression in the cottontail rabbit papillomavirus--rabbit model

A cottontail rabbit papillomavirus (CRPV) E6 DNA vaccine that induces significant protection against CRPV challenge was used in a superior vaccination regimen in which the cutaneous sites of vaccination were primed with an expression vector encoding granulocyte-macrophage colony-stimulating factor (GM-CSF), a cytokine that induces differentiation and local recruitment of professional antigen-presenting cells. This treatment induced a massive influx of major histocompatibility complex class II-positive cells. In a vaccination-challenge experiment, rabbit groups were treated by E6 DNA vaccination, GM-CSF DNA inoculation, or a combination of both treatments. After two immunizations, rabbits were challenged with CRPV at low, moderate, and high stringencies and monitored for papilloma formation. As expected, all clinical outcomes were monotonically related to the stringency of the viral challenge. The results demonstrate that GM-CSF priming greatly augmented the effects of CRPV E6 vaccination. First, challenge sites in control rabbits (at the moderate challenge stringency) had a 0% probability of remaining disease free, versus a 50% probability in E6-vaccinated rabbits, and whereas GM-CSF alone had no effect, the interaction between GM-CSF priming and E6 vaccination increased disease-free survival to 67%. Second, the incubation period before papilloma onset was lengthened by E6 DNA vaccination alone or to some extent by GM-CSF DNA inoculation alone, and the combination of treatments induced additive effects. Third, the rate of papilloma growth was reduced by E6 vaccination and, to a lesser extent, by GM-CSF treatment. In addition, the interaction between the E6 and GM-CSF treatments was synergistic and yielded more than a 99% reduction in papilloma volume. Finally, regression occurred among the papillomas that formed in rabbits treated with the E6 vaccine and/or with GM-CSF, with the highest regression frequency occurring in rabbits that received the combination treatment.

New concepts on the role of human papillomavirus in cell cycle regulation

Human papillomaviruses (HPVs) are strictly host-specific and also show a distinct tropism to squamous epithelial cells. Upon HPV infection, only a portion of the virus reaching the nucleus seems to undergo replication, suggesting that HPV replication remains confined to a small number of cells. HPVs critically depend on the cellular machinery for the replication of their genome. Viral replication is restricted to differentiated keratinocytes that are normally growth arrested. Hence, HPVs have developed strategies to subvert cellular growth regulatory pathways and are able to uncouple cellular proliferation and differentiation. Endogenous growth factors and cellular oncogenes modify HPV E (early) and L (late) gene expression and influence on the pathogenesis of HPV infections. HPV oncoproteins (E5, E6, E7) are important proteins not only in cell transformation but also in the regulation of the mitotic cycle of the cell, thus allowing the continuous proliferation of the host cells. Cyclins are important regulators of cell cycle transitions through their ability to bind cyclin-dependent kinases (cdks). Cdks have no kinase activity unless they are associated with a cyclin. Several classes of cyclins exist which are thought to coordinate the timing of different events necessary for cell cycle progression. Each cdk catalytic subunit can associate with different cyclins, and the associated cyclin determines which proteins are phosphorylated by the cdk-cyclin complex. The effects of HPVs on the cell cycle are mediated through the inhibition of antioncogens (mostly p53 and retinoblastoma) and through interference with the cyclins and cdks, resulting in target cell proliferation, their delayed differentiation, and as a side-effect, in malignant transformation.

Human papillomavirus and head and neck cancer: epidemiology and molecular biology

BACKGROUND

Human papillomaviruses (HPV) are known to cause cancers of the cervix and other anogenital tract sites. Molecular biology has provided some evidence as to the specific mechanisms involved in the HPV-related carcinogenesis. Epidemiologic and molecular biology studies have also suggested that HPV infection may be associated with cancers of the head and neck.

METHODS

This review summarizes the biology of HPV and its potential etiologic role in head and neck cancer. Published reports were used to determine the prevalence of HPV in benign, precancerous, and neoplastic lesions of the oral cavity, pharynx, and larynx. The prevalence was also examined by head and neck site, HPV type, and method of HPV detection. In addition, the occurrence of HPV in normal head and neck tissue, epidemiologic factors related to HPV infection, and clinical implications are discussed.

RESULTS

Overall, the frequency of HPV in benign and precancerous lesions ranged from 18.5% to 35.9%, depending upon the detection methodology. Based upon the most sensitive method of detection, polymerase chain reaction (PCR), the overall prevalence of HPV in head and neck tumors was 34.5% (416 of 1205 tumors). The majority of HPV-positive tumors contained the "high risk" HPV types 16 (40.0%) and 18 (11.9%). Among head and neck sites, HPV was most often detected in tumors of the oral cavity (59%), followed by the pharynx (43%), and larynx (33%). The frequency of HPV positivity in oral samples from healthy individuals ranged from 1% to 60%. A limited number of descriptive and analytic epidemiologic studies have indicated that age (<60 years) and sex (male) were associated with the presence of HPV in the tumor, whereas tobacco and alcohol use were not. The relationship between HPV and survival is unclear, with few comprehensive studies currently available.

CONCLUSIONS

The prevalence of HPV, particularly the high-risk types, suggests a potential etiologic role for the virus in head and neck cancer. Molecular biology has provided important data on the interaction of the HPV oncoproteins with genes important in cell cycle control. Nonetheless, more basic research is needed to describe the physical state of the virus in a variety of cell types and the interaction with other genes. In addition, epidemiologic research is required to further understand the association between HPV and demographic and other risk factors as well as possible routes of transmission. Finally, much work is warranted to provide a definitive assessment of the prognostic significance of HPV in head and neck cancer.

Enhanced transcriptional activation by E2 proteins from the oncogenic human papillomaviruses

A systematic comparison of transcriptional activation by papillomavirus E2 proteins revealed that the E2 proteins from high-risk human papillomaviruses (human papillomavirus type 16 [HPV-16] and HPV-18) are much more active than are the E2 proteins from low-risk HPVs (HPV-6b and HPV-11). Despite the tropism of HPVs for particular epithelial cell types, this difference in transcriptional activation was observed in a number of different epithelial and nonepithelial cells. The enhanced activities of the E2 proteins from high-risk HPVs did not result from higher steady-state levels of protein in vivo, and in vitro DNA-binding assays revealed similar binding properties for these two classes of E2 proteins. These results demonstrate that the E2 proteins from high-risk HPVs have an intrinsically enhanced potential to activate transcription from promoters with E2-responsive elements. We found that there are also substantial differences between the activation properties of the bovine papillomavirus type 1 E2 protein and those of either of the two classes of HPV E2 proteins, especially with regard to requirements for particular configurations of E2 binding sites in the target promoter. Our results indicate that there are at least three distinct functional classes of E2 proteins and that these classes of E2 proteins may perform different roles during the respective viral life cycles.

Methylation of cottontail rabbit papillomavirus DNA and tissue-specific expression in transgenic rabbits

Transgenic rabbits carrying multiple copies of CRPV genomic DNA were described previously (Peng et al. (1993) J. Virol. 67, 1698-1701). CRPV DNA was detectable in all tissues of the transgenic rabbits, however, the transcripts of CRPV genes only were found in skin and skin tumors. Tumor development was also restricted to skin. To study the mechanism involving tissue-specific expression of CRPV genes in rabbits, cellular DNAs, isolated from different normal tissues and skin tumors, were digested with the two isoschizomeric restriction endonucleases MspI (methylation resistant) and HpaII (methylation sensitive), respectively, and analyzed by Southern blot. CRPV DNA, especially its upstream regulatory region (URR), was extensively methylated in all normal tissues, but methylation was remarkably reduced in skin tumors. These data suggest that extensive methylation of CRPV genome, especially in the URR, might be a factor in controlling its tissue-specific expression.

Papilloma formation by cottontail rabbit papillomavirus requires E1 and E2 regulatory genes in addition to E6 and E7 transforming genes

The present study used the cottontail rabbit papillomavirus DNA-rabbit system to evaluate whether the regulatory genes E1 and E2 and the transforming gene E6 are required for papilloma formation. Frameshift mutations were generated in the individual genes in the context of a full-length cottontail rabbit papillomavirus genome, and the mutant DNAs were intradermally inoculated into domestic rabbits. None of the mutants induced papillomas. Marker rescue experiments confirmed that the defects were due to mutations that we deliberately introduced. Marker rescue also confirmed our previous report that the upstream region of E7 around position 9 was critical for papilloma induction. These results demonstrate that the E1 and E2 regulatory genes as well as the E6 and E7 transforming genes are each required for papilloma formation. Each gene may provide molecular targets for therapeutic intervention.

Identification of three transforming proteins encoded by cottontail rabbit papillomavirus

Cottontail rabbit papillomavirus (CRPV) provides an animal model for human papillomaviruses associated with a high risk of cancer development. So far, nothing is known about the transforming functions of CRPV genes because of the lack of an assay system. We have recently developed two systems to assay for CRPV transforming functions. One is based on the finding that transformation of NIH 3T3 cells by CRPV is considerably increased by deleting sequences in open reading frame L2. The second one is based on the use of a cottontail rabbit skin epithelial cell line, sf1Ep (C. Meyers and F. O. Wettstein, Virology 181:637-646, 1991). Mutations were introduced which abolished expression of the full-length E6 protein (LE6), the short E6 protein (SE6) initiated at the second ATG of E6, the E7 protein, or the E5 protein. Mutations affecting LE6 or E7, but not SE6, reduced transformation of NIH 3T3 and sf1Ep cells. Transformed NIH 3T3 cell lines with mutations in LE6 and E7 did not grow in soft agar, while those with mutations in SE6 and E5 grew with a reduced efficiency. The cell lines with mutations in LE6, SE6, or E7 still did induce tumors in nude mice. These mutations, however, abolished the ability to induce papillomas in rabbits. When expressed individually with a retroviral vector, LE6, SE6, or E7, but not E5, conferred anchorage-independent growth. The level of viral protein expression in these cell lines was generally low, and a comparison of the abundance of virus-specific mRNA showed that cell lines contained 20 to 50 times less mRNA than a cottontail rabbit papilloma. These data demonstrate that CRPV encodes at least three transforming proteins.

Molecular and biological characterization of cottontail rabbit papillomavirus variant DNA sequences integrated in the VX7 carcinoma

The transplantable VX7 carcinoma was derived from a tumor induced by a recoverable strain of cottontail rabbit papillomavirus (CRPV) able to replicate in domestic rabbits. Low levels of late viral gene expression have been retained through serial propagation in rabbits. We have cloned and characterized the three major types of CRPV sequences integrated in this tumor, a genome-length 8-kb DNA molecule and two rearranged 9- and 3.8-kb molecules. The VX7 8-kb DNA displays only a few differences in its restriction map, when compared to the wild-type (wt) CRPV DNA. The VX7 9- and 3.8-kb DNAs derive from the VX7 8-kb DNA since they share the same restriction site polymorphism. The VX7 9-kb DNA contains a duplication of the E6 open reading frame. The VX7 3.8-kb DNA results from the deletion of most of the E region and the insertion, between the borders of the deletion, of 174-nucleotide-long segment of the long control region potentially driving the expression of a truncated L2 protein. Both VX7 9- and 3.8-kb species potentially allow the expression of abnormal E6 fusion proteins. Nineteen point mutations were detected in the 3.8-kb DNA, compared to the wt CRPV DNA. None of these molecules were able to induce warts in domestic rabbits, in contrast to wt CRPV DNA. Furthermore, when cloned VX7 DNAs were inoculated together with wt CRPV DNA, none of the VX7 CRPV sequences, as identifiable by their specific restriction enzyme cleavage patterns, could be detected in the resulting warts. This suggests that CRPV sequences integrated in the VX7 carcinoma are no longer able to replicate as episomes, which might be a prerequisite for the production of warts.

Molecular cloning and characterization of a late Tipula iridescent virus gene

Virions of the cytoplasmic, icosahedral insect virus, Tipula iridescent virus (TIV), contain two major DNA components (L, greater than 176 kb; and S1, 10.8 kb) and 25-30 proteins. We characterized a gene (L96) whose 3.6-kb transcript is expressed late in the course of TIV infection of cultured of Estigmene acrea (salt marsh caterpillar, permissive host) and Aedes albopictus (mosquito, semipermissive host) cells. The L96 gene has an open reading frame of 867 codons, predicting a protein of 96 kDa with a pI of 10.9. The C terminus of the L96 protein is rich in hydrophobic amino acids and contains a small region of homology spanning a proteolytic cleavage site within two mammalian viral (GAG) polyproteins. Additional identity with H5 lysine-rich histones in the same region and with other DNA-binding proteins suggests that this protein may be involved in TIV structure. The lengths of the 5'- and 3'-untranslated regions of the L96 transcript were determined to be 21 nucleotides (nt) and 700 nt, respectively. Comparison of the TIV L96- and capsid-encoding genes, both of which are expressed late in infection, revealed that their 5' and 3' regions are generally rich in A and T residues, and that their 3' ends encode at least one eukaryotic polyadenylation signal (AATAAA).

Cooperative activation of transcription by bovine papillomavirus type 1 E2 can occur over a large distance

The viral transcriptional factors encoded by the E2 open reading frame bind to the specific DNA sequence elements ACCGNNNNCGGT, allowing activation or repression of transcription. We have analyzed bovine papillomavirus type 1 E2 transactivation using recombinant genes containing E2-binding sites inserted at either 3' or 5' positions relative to the heterologous transcriptional initiation site of the herpes simplex virus thymidine kinase gene. In these hybrid plasmids, strong transactivation required the presence of a minimum of two E2-binding sites in close proximity to the promoter or five binding sites at a distance. The presence of a single E2-binding motif 5', close to the initiation site, increased the efficiency of E2 transactivation from a distance in a more-than-additive manner. Since each E2-binding site bound a dimer of the E2 protein, these experiments suggest that transactivation by E2 requires the interaction between several E2 dimers with other essential transcription factors. This interaction may be facilitated by DNA looping, which would bring E2 molecules close to the promoter.

Cooperative activation of transcription by bovine papillomavirus type 1 E2 can occur over a large distance

The viral transcriptional factors encoded by the E2 open reading frame bind to the specific DNA sequence elements ACCGNNNNCGGT, allowing activation or repression of transcription. We have analyzed bovine papillomavirus type 1 E2 transactivation using recombinant genes containing E2-binding sites inserted at either 3' or 5' positions relative to the heterologous transcriptional initiation site of the herpes simplex virus thymidine kinase gene. In these hybrid plasmids, strong transactivation required the presence of a minimum of two E2-binding sites in close proximity to the promoter or five binding sites at a distance. The presence of a single E2-binding motif 5', close to the initiation site, increased the efficiency of E2 transactivation from a distance in a more-than-additive manner. Since each E2-binding site bound a dimer of the E2 protein, these experiments suggest that transactivation by E2 requires the interaction between several E2 dimers with other essential transcription factors. This interaction may be facilitated by DNA looping, which would bring E2 molecules close to the promoter.

Human papillomavirus type 1 produces redundant as well as polycistronic mRNAs in plantar warts

Human papillomavirus type 1 (HPV-1) causes plantar warts. On the basis of previously mapped mRNAs and sequence homologies of HPV-1 to other papillomaviruses, we designed oligonucleotide primers and employed the polymerase chain reaction to recover HPV-1 cDNAs from plantar warts. Seven spliced RNA species were characterized, including three not previously detected, and the coding potentials of each were deduced. The most abundant viral mRNA encodes an E1i--E4 protein. One new species is predicted to encode the full-length E2 protein, and another can, theoretically, encode the E2-C or E1-M proteins, three products that regulate mRNA transcription and DNA replication. One RNA species originating from a novel HPV promoter in the upstream regulatory region has the potential to encode the minor capsid protein L2. A newly recognized E5a open reading frame (ORF) is contained in all mRNAs that are polyadenylated at the E-region poly(A) site and also in a putative L2 mRNA. Three distinct species, two of which are derived from the upstream regulatory region promoter, have the potential to encode the L1 protein; the third species also contains the entire coding region of the E1i--E4 protein 5' to the L1 ORF. Both the E1i--E4 mRNA and the potentially bicistronic L1 mRNA are derived from a promoter located in the E7 ORF. We uncovered no evidence of alternatively spliced mRNAs that could account for the multiple, abundant E4 proteins in plantar warts, suggesting that posttranslational modification is mainly responsible for the observed protein heterogeneity.

Phenotypic analysis of bovine papillomavirus type 1 E2 repressor mutants

The bovine papillomavirus type 1 (BPV-1) E2 open reading frame encodes three proteins: the E2 transcriptional transactivator, the E2 transcriptional repressor (E2-TR), and the E8/E2 fusion peptide. In this study, we describe the phenotypes of BPV-1 mutants which are disrupted in their capacity to encode either the E2 transcriptional repressor or the E8/E2 fusion peptide. We also describe experiments which demonstrate that the E8/E2 gene product functions similarly to E2-TR. In the context of the entire viral genome, disruption of E8/E2 expression had little effect on viral processes, whereas disruption of E2-TR expression resulted in a complex phenotype marked by a 10- to 20-fold increase in viral DNA plasmid copy number as well as increased transformation potential. A double mutant, defective in the expression of both E8/E2 and E2-TR proteins, had high levels of E2 transactivation activity yet had reduced plasmid replication capacity and a delayed capacity to transform rodent cells.

Characterization of cDNAs of spliced HPV-11 E2 mRNA and other HPV mRNAs recovered via retrovirus-mediated gene transfer

Human papillomaviruses (HPVs) are associated with hyperproliferations of cutaneous or mucosal epithelium. These viruses cannot be propagated in any cell culture system. Because cloning cDNA copies of HPV mRNAs recovered from human lesions has met with only very limited success, the characterization of HPV mRNAs has been problematic. Using the Moloney murine leukemia virus vector system (C.L. Cepko, B.E. Roberts, and R.C. Mulligan, 1984, Cell 37, 1053-1062), we have recovered cDNAs of spliced E2 mRNAs of human papillomavirus type 11 and additional mRNAs of type 11 and type 18 and determined the utilization of open reading frames (ORFs) in the DNA sequences. The recovery of cDNA copies of messages with splice sites identical to those previously described strongly suggests that the newly characterized splice donors and acceptors are also authentic. The HPV-11 E2 cDNA contains the intact E6 and E7 ORFs and the beginning of the E1 ORF in the first exon, which is then spliced from nt 847 to the second exon at nt 2622, 100 nucleotides upstream from the initiation codon for the E2 ORF. The initiation codon in the E1 ORF is followed by four additional in-frame AUG triplets and an in-frame termination codon positioned 30 nucleotides upstream from the initiation codon for the E2 protein. The authenticity of this putative E2 cDNA was shown by its ability to provide enhancer transactivating activity in chloramphenicol acetyltransferase (CAT) assays in several cell lines. A mutation in the genomic DNA at this splice acceptor site eliminates its activity, demonstrating that the splice is essential for the expression of the E2 protein. We conclude that the translation of the HPV-11 E2 protein requires internal initiation.

The human papillomavirus type 18 (HPV18) E2 gene product is a repressor of the HPV18 regulatory region in human keratinocytes

The human papillomavirus type 18 (HPV18) long control region (LCR) harbors transcriptional promoter and enhancer elements. Recombinant plasmids bearing all or part of the HPV18 LCR cloned in enhancer or promoter configuration upstream of the chloramphenicol acetyltransferase (CAT) gene were transfected into human fibroblasts and keratinocytes. Although the HPV18 enhancer can function in the absence of E2 gene products in both fibroblasts and keratinocytes, the promoter activity of the HPV18 LCR is detectable in keratinocytes but not in fibroblasts, suggesting that it is tissue specific. This promoter activity was repressed in human keratinocytes not only by the bovine papillomavirus type 1 E2 gene product but also by the homologous HPV18 E2 gene product. The promoter involved in the HPV18 E2 repression is located within a 230-base-pair domain directly upstream of the E6 open reading frame of the HPV18 LCR and is probably the previously identified E6 promoter. Although one cannot rule out the possibility that this repressing effect is mediated by a truncated form of HPV18 E2 protein, as was previously demonstrated for bovine papillomavirus type 1, a more likely explanation would be that the full-length HPV18 E2 protein behaves as a repressor. Indeed, at the same doses at which it inhibits transcription from the homologous HPV18 LCR, the HPV18 E2 gene product activates transcription from constructs bearing E2-binding palindromes cloned in enhancer configuration upstream of a heterologous promoter. The fact that the homologous HPV18 E2 gene product acts as a transcriptional repressor of the HPV18 LCR suggests a possible explanation for the overexpression of E6 and E7 open reading frames in cervical carcinoma cells and in cell lines derived from them.

Identification of a novel constitutive enhancer element and an associated binding protein: implications for human papillomavirus type 11 enhancer regulation

The human papillomavirus type 11 enhancer, when linked to the minimal simian virus 40 early promoter, has been dissected into two domains in monkey kidney CV-1 cells, one being constitutive (designated CEI) and the other inducible by trans-acting E2 proteins encoded by homologous and heterologous papillomaviruses (H. Hirochika, T.R. Broker, and L.T. Chow, J. Virol. 61:2599-2606, 1987; H. Hirochika, R. Hirochika, T.R. Broker, and L.T. Chow, Genes Dev. 2:54-67, 1988). We have demonstrated that the natural promoter regulated by this enhancer is located immediately upstream of the E6 open reading frame (the E6 promoter). We have mapped the cap site to nucleotide 99 by RNase protection. We further demonstrate a second constitutive enhancer element, CEII, which is required for transcription from the E6 promoter in the human cervical carcinoma cell lines C-33A and HeLa but not in CV-1 cells. By deletion mapping, we have localized this cell type-specific domain to 71 base pairs by using chloramphenicol acetyltransferase assays. Deletion of either CEI or CEII dramatically decreased the constitutive activity of the enhancer and the E6 promoter, whereas multimerization of either domain in the absence of the other could independently restore expression. Furthermore, when either of these elements was deleted, the full-length E2 protein of human papillomavirus type 11 abolished the remaining basal E6 promoter activity, demonstrating for the first time that the enhancer-activating E2 protein of human papillomaviruses can also function as a transcriptional repressor for the homologous E6 viral promoter. The presence of multiple copies of each element in tandem overcomes the repression by the E2 protein. The effects of CEII are at the level of transcription, without changing the cap site. By gel shift assay, we have shown that a protein present in nuclear extracts of C-33A and HeLa cervical carcinoma cells binds to the newly identified constitutive element II. This protein did not bind the simian virus 40 enhancer, nor did it bind to the enhancer region of many other papillomaviruses tested. UV cross-linking experiments revealed major 44-kilodalton and minor 34-kilodalton proteins that bound specifically to CEII. These two proteins are either related or bind to CEII with high cooperativity. We conclude that transcriptional activities directed by the enhancer and E6 promoter reflect an intricate balance among viral and cellular factors. We present a model on the regulation of the E6 promoter by host and viral transcription factors.

Genetic assignment of multiple E2 gene products in bovine papillomavirus-transformed cells

The E2 open reading frame of bovine papillomavirus type 1 has been shown genetically to encode at least three transcriptional regulatory factors, and three E2 specific proteins have been recently identified in virally transformed rodent cells. In this study, the genes encoding these E2 specific proteins have been determined. The 48-kilodalton (kDa) protein was identified as the product of a full-length E2 open reading frame cDNA, which confirmed that this polypeptide is the E2 transactivator. The 31-kDa E2 protein species, which is the most abundant E2 specific polypeptide, was identified by analysis of both bovine papillomavirus type 1 mutants and cDNAs to be the previously identified E2 transcriptional repressor, E2-TR, which results from translation initiation at an internal E2 ATG codon. The smallest E2 protein species, the 28-kDa polypeptide, was identified as the product of the E8/E2 fusion gene which results from translation of a spliced mRNA species.

Identification of human papillomavirus type 18 transforming genes in immortalized and primary cells

The selective retention and expression of the E6-E7 region of human papillomavirus (HPV) types 16 and 18 in cervical carcinomas suggests that these viral sequences play a role in the development of genital neoplasia. Each of three possible gene products, E6, E6*, and E7, from this region of HPV-18 were examined for transforming properties in several types of rodent cells. We have found that in immortalized fibroblasts, both E6 and E7 (but not E6*) are capable of inducing anchorage-independent growth. In rat embryo cells, the HPV-18 E7 open reading frame was an effective immortalizing agent and complemented an activated ras oncogene for transformation. In both immortalized and primary cells, transformation was observed when the HPV-18 sequences were expressed from either the HPV-18 promoter or a heterologous promoter. The E6-E7 region is not, however, the sole transforming domain of HPV-18, since another portion of the early region, possibly E5, also exhibited transforming capability in immortalized fibroblasts. The development of human cervical carcinomas may therefore involve a series of steps involving multiple viral and cellular gene products.

Functional analysis of the papilloma virus E2 trans-activator in Saccharomyces cerevisiae

The papilloma virus E2 transcriptional trans-activator is representative of a class of transcriptional modulators that activate transcription through direct binding to cis-acting DNA sequences. In this study we measured the capacity for this mammalian virus factor to function in Saccharomyces cerevisiae. When expressed in the yeast, the bovine papilloma virus E2 trans-activator could stimulate transcription from a yeast promoter having E2 DNA-binding sites present in cis. Whereas a single E2 DNA-binding site was sufficient for trans-activation, a strong cooperative effect was observed with two E2 DNA-binding sites. The level of trans-activation was dependent on the position of the E2 DNA-binding sites in relation to the yeast promoter, with the maximal effect demonstrated when the binding sites were positioned upstream. Deleted E2 proteins, lacking part of the trans-activation or DNA-binding domains, failed to activate transcription in yeast, similar to their behavior in mammalian cells. Replacement of the amino-terminal region of the E2 trans-activation domain with a synthetic amphipathic helix partially restored the trans-activation function; however, it did not result in a molecule that exhibited cooperativity between neighboring E2 DNA-binding sites.

E2 polypeptides encoded by bovine papillomavirus type 1 form dimers through the common carboxyl-terminal domain: transactivation is mediated by the conserved amino-terminal domain

The E2 open reading frame (ORF) of bovine papillomavirus type 1 (BPV-1) encodes positive- and negative-acting factors that regulate viral gene expression. The full-length ORF encodes a transactivator, and two transcriptional repressors are expressed from the 3' half of the ORF. Previous analysis has shown that a conserved C-terminal region of 101 amino acids, which is shared by E2 transactivator and repressor proteins, contains the specific DNA binding activity. Further analysis of the E2 transactivator shows that a conserved N-terminal domain of approximately 220 amino acids is crucial for the transcriptional activation function, whereas the variable internal region is not required. The E2 proteins bind to a sequence, ACCGN4CGGT, several copies of which are sufficient to constitute an E2-dependent enhancer. By using a gel retardation assay and proteins derived by in vitro transcription and translation, we were able to show that the E2 polypeptides bind as dimers to a single DNA binding site. The dimeric E2 proteins are stable in the absence of DNA and dimerization is mediated through the DNA binding domain. This may reveal an additional mechanism of repression that could potentially result from the formation of inactive heterodimers between transactivator and repressor species.

Interplay of viral and cellular proteins along the long control region of human papillomavirus type 18

The long control region of human genital papillomavirus type 18 harbors transcription regulatory elements, such as the E6 promoter and a cell type-specific enhancer independent of the E2 protein. By performing DNase I footprint experiments in vitro with protein extracts from different cell lines and tissues we searched for cellular factors interacting with the totality of the control region. We detected a total of eight different protected sites; most of them were found with all the extracts. Two of these sites, one in the enhancer and the other in the sequences proximal to the E6 cap site, interact with a member of the activator protein 1 family. However, in the absence of fine mutational analysis, we cannot readily discern an exact functional role for the different binding sites. We also characterized two regions, which are protected only in the presence of E2, corresponding to the perfect palindromes ACCGN4CGGT and present either 500 nucleotides or tandemly repeated about 70 nucleotides upstream of the E6 cap site. This last protected area covers a large part of the putative TATA box of the E6 promoter and could explain its repression by bovine papilloma virus type 1 E2, which could interfere with binding of the TFII D ubiquitous transcription factor to the TATA sequence of the promoter.