An Enhanced Epithelial Response of a Papillomavirus Promoter to Transcriptional Activators*

Mucosal epitheliotropic papillomaviruses have a similar long control region (LCR) organization: a promoter region, an enhancer region, and a highly conserved distribution of E2 DNA binding sites. The enhancer of these viruses is epithelial-specific, as it fails to activate transcription from heterologous promoters in nonepithelial cell types (Gloss, B., Bernard, H. U., Seedorf, K., and Klock, G. (1987) EMBO J. 6, 3735–3743; Morgan, I. M., Grindlay, G. J., and Campo, M. S. (1999) J. Gen. Virol. 80, 23–27). Studies on E2 transcriptional regulation of the human mucosal epitheliotropic papillomaviruses have been hindered by poor access to the natural target cell type and by the observation that some of the human papillomavirus promoters, including human papillomavirus-16, are repressed in immortalized epithelial cells. Here we present results using the bovine papillomavirus-4 (BPV-4) LCR and a bovine primary cell system as a model to study the mechanism of E2 transcriptional regulation of mucosal epitheliotropic papillomaviruses and the cell type specificity of this regulation. E2 up-regulates transcription from the BPV-4 LCR preferentially in epithelial cells (Morgan, I. M., Grindlay, G. J., and Campo, M. S. (1998) J. Gen. Virol. 79, 501–508). We demonstrate that the epithelial-specific enhancer element of the BPV-4 LCR is not required for the enhanced activity of E2 in epithelial cells and that the BPV-4 promoter is more responsive, not only to E2, but to other transcriptional activators in epithelial cells. This is the first time a level of epithelial specificity has been shown to reside in a papillomavirus promoter region.

Human papillomaviruses (HPV) 1 are a family of small double-stranded DNA viruses that infect epithelial cells causing benign proliferative lesions. Certain types have been implicated in the development of carcinomas with over 90% of human cervical cancers containing viral sequences of a high risk HPV subtype, predominantly HPV-16 and -18 (3). All papillomaviruses have a noncoding region of 500 -1000 bp called the long control region (LCR). The LCR is the transcriptional control unit of the virus regulating expression of the viral-transforming proteins and of the proteins essential for the viral life cycle. The LCR is a typical transcriptional control unit with a promoter region and an upstream enhancer region (4). Mucosal epitheliotropic LCRs, for example HPV-16 and -18, contain an epithelial-specific enhancer, and it is therefore proposed that one of the restrictions of these viruses to their target cell type is at the transcriptional level.
Bovine papillomavirus-4 (BPV-4) is a mucosal epitheliotropic papillomavirus that infects the upper alimentary canal of cattle causing benign papillomas with a high risk of progressing to carcinoma in cattle feeding on bracken fern (5). The BPV-4 LCR has an organization similar to other mucosal epitheliotropic papillomaviruses. The epithelial-specific enhancer is of a similar size and position as that in the HPV-16 and -18 LCRs. There are two main regions contributing to enhancer activity. One of these sites, Site 2, has a corresponding sequence in the LCR of HPV-16 (6).
All papillomavirus LCRs contain binding sites for virally encoded E2; expression of which is controlled by the LCR. The E2 gene product is a 45-48-kDa protein that binds to the 12-bp palindromic DNA sequence -ACCGNNNNCGGT-as a dimer (for a review, see Ref. 7). All E2 proteins have three functional domains: an amino-terminal acidic transactivation domain, a carboxyl-terminal DNA binding and dimerization domain, and a central variable hinge region (8). As well as regulating transcription, E2 is required for replication of the viral genome (9 -11) and is therefore essential for the viral life cycle. The amino terminus of E2 can interact with E1, the major viral replication factor, directing it to the origin of replication. Disruption of this interaction is sufficient to prevent viral DNA replication in vitro (12). Mutational analysis of the E2 transactivation domain has demonstrated that the ability of E2 to regulate transcription and replication can be separated, indicating that the cellular proteins with which E2 interacts to carry out these two functions are different (13,14).
Several E2 molecules, including BPV-1 E2, BPV-4 E2, and HPV-16 E2, up-regulate transcription from the BPV-4 LCR in an epithelial-specific manner (2). 2 The inability of E2 to upregulate transcription in fibroblasts is promoter-specific as E2 can activate transcription from heterologous promoters in a variety of cell types (15)(16)(17). When studying the ability of E2 to up-regulate transcription from HPV LCRs, the cell type employed seems crucial. In primary human keratinocytes E2 upregulates transcription from the HPV-16 promoter (18). However, most studies have been carried out using transformed keratinocyte cell lines where E2 represses transcription from HPV LCRs. This repression is mediated by the E2 binding site 3 bp upstream from the TATA box (16,19).
Although the overall LCR sequence homology between BPV-4 and HPV-16 and -18 is low, the distribution of E2 binding sites within these LCRs is identical. Immediately upstream from the TATA box are two E2 binding sites separated from each other and the TATA box by 3 or 4 base pairs. There are also two upstream sites; one is beside the E1 DNA binding site involved in the regulation of viral DNA replication, and one is a further 300 -400 bp upstream. The conservation of the organization of these sites through evolution strongly suggests that the mechanism E2 uses to regulate transcription from these LCRs is conserved.
To study the mechanisms of E2 transcriptional regulation of mucosal epitheliotropic papillomaviruses and the tissue specificity of this regulation, we used the BPV-4 LCR and primary bovine palate keratinocytes (PalK) and palate fibroblasts (PalF). Comparisons are made between the results in PalK cells, the natural target cell type for transformation by BPV-4, with PalF cells from the same source. E2 up-regulates transcription from the BPV-4 LCR in PalK cells but not in PalF cells (2). Low to intermediate levels of E2 up-regulate transcription from the BPV-4 LCR, whereas at elevated levels of E2, transcription is down-regulated. This down-regulation is mediated by E2 interaction with the TATA proximal E2 DNA binding site, presumably disrupting the TATA box-binding protein-TATA box contact and therefore decreasing the rate of transcriptional initiation (2). The results presented here demonstrate that the BPV-4 LCR epithelial-specific enhancer element is not required for the enhanced activity of E2 in keratinocytes. We show that the BPV-4 promoter is more responsive to transcriptional activators in PalK cells than in PalF cells, whereas the tk promoter shows no such epithelial preference. This is the first time a level of epithelial specificity has been shown to reside in a papillomavirus promoter region.
When annealed together, a pair of E2 sites separated by 4 base pairs with BamHI/BglI compatible ends is generated. After end-labeling with T4 kinase and ligation, the double-stranded oligonucleotides were restriction digested with BamHI/BglII. This ensures that only concatamers of E2 binding sites facing in the same orientation are generated as ligation in the wrong orientation reconstitute a restriction site. LexA binding sites were generated in the same manner. Oligonucleotides used were as follows: upper strand, 5Ј-GATCCTGCTGTATATAAAAC-CAGTGGTTATATGTACAGTAA; and lower strand, 5Ј-GATCTTACTG-TACATATAACCACTGGTTTTATATACAGCAG, which when annealed generates a single colE1 operator with cut BamHI/BglII ends. The oligonucleotides were separated on a 6% polyacrylamide gel; the bands corresponding to 2, 4, 6, and 8 binding sites were excised, eluted overnight in minimal volume of elution buffer (0.1% SDS, 0.5 M NH 4 OAc, 10 mM MgOAc), and purified by LiCl and EtOH precipitation. The tk promoter from nucleotide 75-199 was polymerase chain reaction cloned as a BglII-HindIII fragment into pGL3. The fidelity of all plasmid constructions was verified using an ABI 373A automated sequencer. pCMV HPV16-E2, pCGBPV1-E2, and pCGVP16-E2 have been described previously (2,18). VP16-LexA was a gift from Dr. Chris Bartholomew (Glasgow Caledonian University).
Cell Culture-PalK cells were prepared from fetal biopsies as described for human cervical keratinocytes (21). They were cultured on irradiated Swiss 3T3 feeders in the conditions described (20). The Swiss 3T3 feeders were grown in simple liquid medium (Life Technologies, Inc.) with 10% fetal calf serum and irradiated with 60 grays prior to use as feeders. PalF cells were prepared as described (22) from the same palate biopsy used to prepare the PalKs and cultured in Dulbecco's modified Eagle's medium with 10% fetal calf serum. At least two different sources of these primary cell types were used in the experiments described.
Transfection-PalK cells were transfected using the polybrene-Me 2 SO technique as described (23). Briefly, 5 ϫ 10 5 cells were seeded on a 60-mm tissue culture dish without feeders. The following day the medium was replaced with 10 mg/ml polybrene containing medium, and the DNA was added. After 6 h the DNA-polybrene containing medium was removed, and a 35% Me 2 SO solution was added to the cells for 3 min. Following this incubation the cells were washed two times with phosphate-buffered saline and then refed with normal medium. The cells were harvested 44 -48 h later. PalF cells were transfected using a standard calcium phosphate precipitation technique. Briefly, cells were plated out at 2 ϫ 10 5 /60-mm tissue culture disc. The following day a calcium phosphate precipitate containing the DNA was added to the cells overnight. The following morning the cells were washed twice with phosphate-buffered saline and refed with normal medium. The cells were harvested 28 -32 h later.
Transcription Assay-PalK and PalF cells were lysed directly on the tissue culture plates. The medium was removed, and the cells were washed twice with phosphate-buffered saline. 300 l of reporter lysis buffer (Promega) was added to the plate and left for 10 min. The cell lysate was then scraped from the dish and placed in a 1.5-ml centrifuge tube. The lysate was cleared by centrifuging the sample for 10 min and removing the supernatant to a fresh tube. 80 l of the supernatant were then assayed for luciferase activity using the luciferase assay system (Promega) with a BioOrbit 1251 or a Tropix TR717 microplate luminometer (Figs. 5b and 7). To standardize for cell number, the protein concentration was determined. pGL3CONT, which contains the SV40 promoter-and enhancer-driving expression of the luciferase gene, was transfected in parallel to confirm efficient transfection. This construct demonstrates high levels of transcriptional activity in both keratinocytes and fibroblasts. All transfections were repeated at least three times in duplicate.
Western Blotting-PalK and PalF cells were transfected with increasing amounts of pCMV HPV16-E2 expression vector. Cells were lysed on ice in 50 l of SDS-polyacrylamide gel electrophoresis lysis buffer (100 mM Tris-HCl, pH 6.8, 2% SDS, 20% glycerol), sonicated, and clarified by centrifugation for 10 min at 4°C, and the supernatant was removed. Protein concentration was determined by absorbance measurement at 280 nm. 50 g of each sample were electrophoresed on 10% SDS-polyacrylamide gel electrophoresis and transferred onto a nitrocellulose membrane (ECL Hybond, Amersham Pharmacia Biotech). Membranes were incubated with a TVG261 monoclonal antibody directed against an epitope in the amino-terminal region of HPV-16 E2 (24) followed by an anti-mouse secondary antibody conjugated with peroxidase. Western blots were developed by enhanced chemiluminescence (Amersham Pharmacia Biotech).

RESULTS
The BPV-4 Promoter Retains the Epithelial-specific Response to Transcriptional Up-regulation by E2-The BPV-4 LCR has an enhancer that functions in epithelial cells but not in fibroblasts (6) (Fig. 1a). Also, E2 up-regulates transcription from the BPV-4 LCR in epithelial cells but not in fibroblasts (2). There are three possible reasons for this epithelial-specific function of E2: the E2 protein functions more efficiently in epithelial cells, E2 may interact with the epithelial-specific enhancer to regulate transcription, or the LCR promoter may show an enhanced epithelial response to transcriptional activators. To determine which of these possibilities are responsible, the epithelial-specific enhancer was removed and a series of concatamers of increasing numbers of E2 binding sites were cloned upstream from the BPV-4 promoter in the position of E2 BS3 (Fig. 1b). Mutations were introduced into the TATA box proximal E2 sites preventing E2 binding (Fig. 1b) as these sites have previously been shown to mediate down-regulation of transcription at high levels of E2 (2,20). This yielded a series of plasmids with the following nomenclature: the PV series represents the wild type LCR promoter sequence; and the PV1 series has the TATA proximal E2 DNA binding site mutated, whereas the PV2 series has both the TATA proximal and the adjacent E2 site mutated (see Fig. 1b for details). Increasing numbers of E2 DNA binding sites were then inserted upstream from these promoters: 2, 4, 6, and 8 copies of E2 DNA binding sites inserted PV 2E2 (2 E2 sites), PV 4E2 (4 E2 sites), etc.
In PalK and PalF these plasmids have similar background transcriptional activity (i.e. there is no enhanced activity in the epithelial cells). Also, in the absence of E2 protein the E2 sites do not enhance transcription, and without the LCR promoter sequences E2 does not activate transcription, indicating that there is no cryptic promoter in the reporter plasmid that is responsive to E2 (data not shown). HPV-16 E2, which functions in an identical manner to BPV-4 E2, 2 can specifically bind to the E2 sites inserted upstream from the LCR promoter constructs. We chose to study the effects of HPV-16 E2 on these promoter constructs in PalK and PalF as HPV-16 E2 is functionally identical to BPV-4 E2 in regulating transcription from the LCR. HPV-16 is a causative agent in a number of human diseases, and understanding the function of HPV-16 proteins will present opportunities for interfering with the viral life cycle.
In PalK cells, HPV-16 E2 up-regulates transcription from the LCR promoter in an additive manner increasing as the number of E2 binding sites increases (Fig. 2). A maximum of 20-fold activation of transcription (over the background activity of no E2) is observed with the PV1 8E2 and PV2 8E2 constructs. PV1 6E2 and PV2 6E2, which both have the TATA proximal E2 site (BS1) mutated, have a 2-3-fold elevated response over the PV 6E2 construct (Fig. 2). This is in agreement with previous studies showing that BS1 is responsible for mediating down-regulation of transcription by elevated levels of E2.
To determine if the epithelial-specific response of the fulllength LCR is retained by the promoter region, the transcriptional response to HPV-16 E2 of the PV1 6E2 construct was assayed in both PalK and PalF cells. PV1 6E2 was chosen as E2 BS1 is mutated in this construct eliminating down-regulation of transcription by elevated levels of E2 as a complicating factor. Low levels of E2 are sufficient to up-regulate transcription from the BPV-4 promoter in PalK cells. 10-fold activation is observed at a 1:1 ratio of E2 expression vector to reporter plasmid (Fig. 3a). Over a range of increasing E2 concentrations the BPV-4 promoter is not up-regulated more than 2-fold in PalF cells (Fig. 3a). Fig. 3b shows that E2 is being expressed at similar levels as a doublet of approximately 42 kDa in both PalK and PalF cells. These results demonstrate that the BPV-4 promoter mimics the response of the full-length LCR to transcriptional up-regulation by E2, that is, the response is much enhanced in PalK cells. They also demonstrate that the epithelial-specific enhancer element of the BPV-4 LCR is not required for the enhanced activity of E2 in PalK cells.
The BPV-4 Promoter Shows an Enhanced Epithelial Response to Transcriptional Activators-It has been proposed that the epithelial-specific transcriptional regulation of the LCR promoter may be mediated by the E2 transactivation domain (2). To test this hypothesis we assayed the ability of a chimeric molecule that has the VP16 transactivation domain fused to the BPV1 E2 DNA binding domain and of wild type BPV1 E2 to activate transcription from the PV2 6E2 construct. VP16 is a strong acidic transactivator from the herpes simplex virus. BPV-1 E2 activates transcription 18-fold in PalK cells but no more than 2-fold in PalF cells, demonstrating that the preferential activation in epithelial cells is retained with BPV-1 E2 (Fig. 4a). VP16-E2 also activates transcription from PV2 6E2 better in PalK cells than in PalF (Fig. 4b). At low levels VP16-E2 activates transcription 180-fold in fibroblasts and 900-fold in keratinocytes. At high levels of VP16-E2 transcription is down-regulated in both cell types. This may be because of squelching where excess VP16-E2 proteins not bound to DNA sequester factors required for transcriptional activation. These results demonstrate that the E2 transactivation domain is not solely responsible for the enhanced function of E2 on the BPV-4 promoter in epithelial cells.
Two components of the cellular basal transcription initiation complex, TATA box-binding protein and transcription factor IIB, have been shown to interact directly with the carboxylterminal DNA binding domain of E2 (25). To analyze the contribution of the DNA binding domain of E2 to cell type-specific transcriptional activation, four copies of the LexA site from the colE1 promoter were cloned upstream of the BPV-4 promoter with both TATA box proximal E2 binding sites mutated (Fig.  5a). The ability of a chimeric molecule, which has the VP16 transactivation domain fused to the bacterial LexA DNA binding domain, to activate transcription from this construct was assayed in PalK and PalF cells. At low to intermediate levels VP16-LexA activates transcription to a similar degree in both cell types. However, the BPV-4 promoter shows a 5-fold enhanced epithelial response to activation by high levels of VP16-LexA. A 1:1 ratio of expression vector to reporter plasmid up-regulates transcription approximately 200-fold in fibroblasts and 1000-fold in keratinocytes (Fig. 5b). This difference is similar to that observed with VP16-E2. These results suggest that the carboxyl-terminal region of E2 is not involved in mediating epithelial-specific transcriptional regulation of the LCR promoter but functions to localize active E2 dimers to the target promoter.
To address the possibility that E2 binding to its target sequences was being blocked in fibroblasts and that the enhanced epithelial response to transcriptional activators was a specific property of the BPV-4 promoter, six E2 binding sites were cloned upstream from the herpes simplex virus tk promoter (Fig. 6a). The tk promoter has similar background transcrip- FIG. 4. The BPV-4 promoter shows an enhanced epithelial response to transcriptional activators. PalK and PalF cells were cotransfected with 1 g of PV2 6E2, which has both the TATA box proximal E2 BS1 and BS2 mutated, and the indicated amounts of pCG BPV-1 E2 and pCG VP16 E2 expression vectors. pCG was used to make the total amount of DNA used 2 g in each case. Results are expressed as fold transactivation over the luciferase activity in the absence of expression vector. FIG. 5. a, PV2 4LexA reporter construct. Four copies of the LexA site from the colE1 promoter were cloned upstream of the BPV-4 promoter with both TATA proximal E2 binding sites mutated. b, the PV2 promoter shows an enhanced epithelial response to activation by VP16-LexA. PalK and PalF cells were cotransfected with 1 g of PV2 4LexA reporter construct and the indicated amounts of pCGVP16-LexA expression vector. The total amount of DNA transfected was made equal in each case. Results are expressed as fold transactivation over the luciferase activity in the absence of VP16-LexA expression vector. tional activity in both PalK and PalF cells in the absence of E2 (data not shown) and is approximately the same length as the PV promoter so that the E2 molecules are operating from a similar distance. The ability of VP16-E2 to up-regulate transcription from the tk 6E2 construct was assayed, and the results are shown in Fig. 6b. It is clear that the VP16-E2 chimera is being expressed, binding to its target sites and activating transcription in both PalK and PalF cells. At low levels VP16-E2 activates transcription from the tk promoter preferentially in PalF cells. At intermediate levels VP16-E2 activates transcription to a similar degree in both cell types, whereas at high levels of VP16-E2 transcriptional activity is down-regulated probably because of a squelching mechanism.
To further investigate the contribution of the E2 transactivation domain to epithelial specificity, the ability of E2 to up-regulate transcription from the tk 6E2 construct was assayed in PalK and PalF cells. Fig. 7 shows that E2 up-regulates transcription from the tk promoter in a cell type-independent manner. A maximum of approximately 90-fold activation is observed with a 0.1:1 ratio of expression vector to reporter plasmid in both cell types. These results demonstrate that the tk promoter does not show an epithelial preference to activation by both VP16-E2 and E2 itself. DISCUSSION In PalK cells, the natural target cell type for transformation by BPV-4, E2 up-regulates transcription from the BPV-4 promoter in an additive manner increasing as the number of E2 binding sites increases (Fig. 2). Such an additive effect of pairs of E2 sites has been observed previously (26 -28). Over a range of increasing E2 concentrations the BPV-4 promoter is not transactivated more than 2-fold in PalF cells (Fig. 3a). Failure to function in PalF cells is not because of a lack of expression, as Fig. 3b shows E2 is being expressed at similar levels as a doublet of approximately 42 kDa in both PalK and PalF cells.
The BPV-4 promoter shows an enhanced epithelial response to activation, not only by E2, but also by VP16-E2 (Fig. 4b) and VP16-LexA (Fig. 5b). The enhanced epithelial response is a promoter-specific effect as the tk promoter shows no such epithelial preference to activation by VP16-E2 (Fig. 6b) and E2 itself (Fig. 7). Previously we have suggested that the E2 protein contributes toward epithelial-specific transcriptional regulation of the BPV-4 LCR (2). Whereas this remains a possibility for the LCR promoter, it is clearly not true for the tk promoter. Transactivators are believed to function, at least in part, through contacting components of the cellular transcription machinery and affecting the formation and/or stability of the preinitiation complex. E2 has previously been shown to bind transcription factor IIB, TATA box-binding protein, Sp1, and AMF-1 (25, 28 -31). It is possible that a combination of interactions with these factors contributes to epithelial specificity or that there are as yet unidentified factors that may be responsible.
The 127-bp BPV-4 promoter region used here contains the TATA box and presumably as yet unidentified core promoter elements but does not have a putative initiator element. Previously it has been suggested that there is an initiator in the BPV-4 promoter just downstream from the TATA box (32). The addition of this putative initiator to the BPV-4 promoter sequence used here resulted in a slight elevation in response to E2 but played no role in the epithelial-specific response to E2. 2 Previous studies have shown that a core promoter must contain at least two elements to be able to respond to E2 and that such elements, the TATA box, the initiator element, or a binding site for an upstream promoter factor, are interchangeable to a large extent (33). Therefore E2 requires the co-operation of at least one additional DNA binding proximal promoter factor, such as Sp1, or other factors that interact with papillomavirus promoters, such as AP-1, Oct-1, NF-1/CTF, or USF for the activation of a minimal TATA box promoter (28,34). Epithelial-specific functions of some of these factors have emerged. NF-1 proteins are the products of a multigene family. Epithelial cells contain proteins derived from the NF1-C gene (NF-1/CTF), but in fibroblasts where the HPV-16 enhancer is inactive, high levels of FIG. 6. a, E2-responsive tk promoter construct. The tk promoter from nucleotide 75-199 containing the TATA box, initiator element, and a GC-rich box was polymerase chain reaction cloned as a BglII-HindIII fragment. Six E2 binding sites were inserted into the BglII site immediately upstream to generate tk 6E2. b, the tk promoter does not show an epithelial preference to activation by VP16-E2. PalK and PalF cells were cotransfected with 1 g of tk 6E2 reporter plasmid and the indicated amounts of pCGVP16-E2 expression vector. pCG was added to make the total amount of DNA transfected 2 g in each case. Results are expressed as fold transactivation over the luciferase activity in the absence of VP16-E2. FIG. 7. HPV-16 E2 activates transcription from the tk promoter in a cell type-independent manner. PalK and PalF cells were cotransfected with tk 6E2 and the indicated amounts of pCMV HPV-16 E2 expression vector. 1 g of reporter construct was used in each assay and the total amount of DNA transfected was made equal by the addition of pCMV. Results are expressed as fold transactivation over the luciferase activity in the absence of E2. the negative regulator derived from the NF1-X gene are expressed (35). Sp1 is one of at least four ubiquitously expressed transcription factors derived from a multigene family (36). Sp3 acts as a repressor probably because of competition with Sp1 for the same binding sites (37). In HPV gene expression Sp1 binds a single site close to the E6 promoter (38 -40). In various epithelial and fibroblast cell lines high levels of Sp1 compared with Sp3 are consistently found where the HPV-16 promoter is active and low levels are found where it is inactive (41). The Sp1 protein can become phosphorylated and glycosylated and might thus be a target of intracellular signaling (42,43). However, none of the consensus binding sites for these factors are present in the BPV-4 promoter. Therefore the sequential action of as yet unidentified proximal promoter factors, either cell type-specific or ubiquitous factors differentially expressed or modified in a cell type-dependent manner, may be involved in determining the enhanced epithelial response of the BPV-4 promoter to upstream activators.
As well as upstream elements, the TATA box itself may be involved in cell type-specific transcriptional regulation. Previous studies have shown that the TATA sequence and core promoter sequences around the TATA box can dictate cell typespecific transcriptional regulation and the response to upstream activators (44 -46). Cell type-specific forms of TFIID or a specific set of non-DNA binding cofactors may be recruited to the BPV-4 LCR promoter allowing E2 and VP16 to utilize a different subset of coactivators in PalK and PalF cells to activate transcription.
Alternatively, it is possible there is a repressor protein(s) that is either specific to, more abundant, or alternatively modified in fibroblasts that inhibits the ability of VP16 and E2 to activate transcription from the viral promoter in this cell type. This may explain why E2 is unable to activate transcription in fibroblasts, but VP16, which has a stronger transactivation domain, is able to partly overcome this negative regulation. VP16-E2 up-regulates transcription from the BPV-4 promoter 180-fold in fibroblasts and 800-fold in keratinocytes (Fig. 4b).
Negative regulation of HPV-16 gene expression has been attributed primarily to the effects of the transcription factor YY1 (47). The importance of cellular negative regulators of viral gene expression has become apparent as studies of primary tumors or metastases were found to contain nonintegrated HPV-16 episomes with mutated YY1 sites (48). Evidence also exists for the presence of a fibroblast repressor on the short arm of chromosome 11. The HPV-16 enhancer-promoter is virtually inactive in normal human diploid fibroblasts but active in human fibroblasts with a deletion in the short arm of one chromosome 11 (del-11 cells). Because the HPV-16 enhancer upstream of the SV40 promoter is active in both normal fibroblasts and del-11 cells, the target for chromosome 11-regulated HPV expression is likely to be located in the HPV-16 early promoter region (nucleotides 57-112) (49). More recently, a novel YY1-independent silencer present in the HPV-16 LCR promoter that can repress the HPV-16 enhancer has been identified (50). Therefore, negative regulators of transcription may play a role in limiting the expression of viral proteins to the host cell type.
Identification of the DNA elements and proteins involved in mediating the enhanced epithelial response of the BPV-4 LCR promoter to upstream activators is currently under investigation. This will lead to a greater understanding of the mechanisms underlying not only mucosal epitheliotropic papillo-mavirus transcriptional control but also cell type-specific transcription in general.