Infectious human papillomavirus type 18 pseudovirions1

Article No. mb982113

J. Mol. Biol. (1998) 283, 529±536

COMMUNICATION

Infectious Human Papillomavirus Type 18 Pseudovirions Yves Stauffer1, Kenneth Raj1, Krzysztof Masternak2 and Peter Beard1* 1

Department of Virology, Swiss Institute for Experimental Cancer Research (ISREC) Chemin des Boveresses 155 CH-1066 Epalinges Switzerland 2

Institute of Animal Biology University of Lausanne 1015 Lausanne, Switzerland

Human papillomavirus type 18 (HPV18) capsid proteins L1 and L2, synthesised in mammalian cells using recombinant vaccinia viral expression vectors, are transported to the nucleus and assembled into virus-like particles. When 293T cells, which express SV40 T antigen, were transfected with plasmid DNAs containing an SV40 origin of replication then infected with vaccinia viral vectors encoding L1 and L2, plasmid DNA was encapsidated into the particles. The DNAs ranged in size from 5.4 to 7.9 kb. By encapsidating plasmids containing either the b-galactosidase gene or the puromycin-resistance gene, the pseudovirions were shown to be infectious in that they could transfer b-galactosidase activity or confer resistance to puromycin to a number of cell types, indicating that the uptake and decapsidation of HPV particles are not the main determinants of cell type speci®city of HPV. Episomal HPV16 DNA in a cervical keratinocyte line could also be encapsidated. Further investigation showed that DNA encapsidation is independent of HPV DNA sequences and of T antigen-mediated plasmid DNA replication. Instead, the minor capsid protein, L2, was found to be attached to plasmid mini-chromosomes extracted from these cells, suggesting a role for L2 in encapsidation. Consistent with this, the L1 protein alone was unable to encapsidate DNA, although it was able to form virus-like particles. The results suggest that intracellular episomal DNAs of suitable size can be encapsidated by the HPV18 L1 and L2 proteins without the need of any HPV packaging signal, and reintroduced into cells. # 1998 Academic Press

*Corresponding author

Keywords: virus assembly; human papillomavirus; viral structural proteins; recombinant vaccinia viruses; DNA replication

Human papillomaviruses (HPVs) are a group of small DNA viruses associated with benign and malignant epithelial hyperproliferations of cutaneous or mucosal origin. More than 75 different types of HPV have been isolated, each type exhibiting a particular tropism for certain kinds of epithelia. Among the types that infect the genital mucosa, HPV16, 18 and a few rarer types (HPV 31, 33 and 45) are causative agents of at least 95% of cervical cancers, the second most frequent cancer Abbreviations used: HPV, human papillomavirus; BPV, bovine papillomavirus; VLP, virus-like particle; SV40T, SV40 T antigen; X-gal, 5-bromo-4-chloro-3indolyl-b,D-galactopyranoside; POD, promyelocytic leukaemia oncogenic domain, MOI, multiplicity of infection. E-mail address of the corresponding author: [email protected] 0022±2836/98/430529±08 $30.00/0

in women world-wide. HPV DNA is found in cervical cancer biopsies, about 60% of cases containing HPV16 and 15% containing HPV18 (reviewed by zur Hausen, 1996). Therefore, the study of these high-risk HPVs is important in view of the need for early detection and vaccine development. The 55 nm icosahedral capsid of papillomaviruses is composed of 72 capsomeres arranged in a T ˆ 7 symmetry (Baker et al., 1991; Belnap et al., 1996; Hagensee et al., 1994). The capsomeres are pentamers of the major capsid proteins L1. The minor capsid protein, L2, represents only a small proportion of the total viral capsid protein (molar ratio L1:L2 ˆ 30:1; Roden et al., 1996; Volpers et al., 1994). The position of L2 in the capsid is not exactly known but it has been proposed from three-dimensional reconstruction of the bovine papillomavirus (BPV) capsid that L2 could be # 1998 Academic Press

530 located in the centre of the capsomeres at the 12 pentavalent vertices (Trus et al., 1997). HPVs cannot be easily propagated in cell culture because the viral life cycle depends on the differentiation of epithelial cells (Taichman & LaPorta, 1987). Moreover, very few viral particles of mucosal HPVs are found in vivo in infected tissues. We and others obtained HPV-like particles (VLPs) by producing L1, or L1 and L2 proteins in expression systems based on recombinant baculovirus (Kirnbauer et al., 1992; Rose et al., 1993; Stauffer & Beard, 1995; Volpers et al., 1994) or vaccinia virus (Carter et al., 1994; Zhou et al., 1991) or in yeast (Sasagawa et al., 1995). L1 has been shown to be suf®cient for the formation of VLPs (Hagensee et al., 1993; Kirnbauer et al., 1993), their binding to cells (Roden et al., 1994) and internalisation (Muller et al., 1995). The function of L2 is less clear but it is able to bind to DNA in vitro (Zhou et al., 1994) and reported to be required for encapsidation of BPV1 DNA (Roden et al., 1996). In order to study the requirements for DNA encapsidation and to develop an infectivity assay for HPV18, in the work reported here we ®rst tested the ability of the HPV18 capsid proteins, introduced into cells using recombinant vaccinia viral vectors, to encapsidate papillomavirus DNA. For this we used a human keratinocyte line, W12, carrying episomal HPV16 DNA (Stanley et al., 1989). We then tested the possibility of incorporating plasmid DNAs containing different reporter genes into HPV18 capsids. Encapsidation of DNA with different markers allowed us to show that the pseudovirions are infectious, permitting the development of an HPV18 infectivity assay. Finally, we also studied which factors are important for DNA encapsidation by HPV18 L1 and L2 proteins to occur. Encapsidation of episomal DNA by HPV18 L1 and L2 proteins In HPV-infected cells, the viral genome is normally present in an episomal form. During the viral life cycle, the capsid proteins are synthesised only in the upper part of the differentiating epithelium following viral DNA ampli®cation, and progeny virions made by packaging of the DNA by the capsid proteins (reviewed by zur Hausen, 1996; zur Hausen & de Villiers, 1994). In order to test the ability of HPV18 L1 and L2 proteins to encapsidate HPV DNA in proliferating monolayer tissue culture cells, we infected W12 cells (clone 20850), which carry HPV16 episomes at approximately 1000 copies per cell (Jeon et al., 1995), with recombinant vaccinia virus expressing HPV18 L1 (vvL1) and L2 (vvL2) proteins. The papillomavirus-like particles were then extracted, sedimented through a sucrose cushion and puri®ed by CsCl density-gradient centrifugation. After fractionation of the CsCl gradient, the density of each fraction was determined and aliquots analysed for the presence of HPV16 DNA. As shown in Figure 1, HPV16 DNA accumulated at a density of 1.325 g/

Infectious HPV18 Pseudovirions

ml, corresponding to the density of full HPV virions (zur Hausen & Gissmann, 1980) indicating that DNA encapsidation is independent of cellular differentiation and that the HPV18 coat proteins can encapsidate a heterologous DNA. In order to obtain HPV18 pseudovirions whose infectivity could be easily assayed, we employed a system which would allow the encapsidation of plasmid DNAs carrying a marker gene. To maximise the yield of DNA-containing pseudovirions, we used plasmid DNAs with an SV40 origin of replication which can replicate as episomes after transfection into mammalian cells which express the SV40 T antigen (SV40T; Cereghini & Yaniv, 1984). Cells of the human embryonic kidney line 293T (Pear et al., 1993) constitutively express SV40T and can be easily transfected (generally 50 to 70%) by the calcium phosphate method. Samples of 293T cells were transfected with the pSVb plasmid, containing the SV40 origin of replication and the b-galactosidase gene as a marker, and co-infected with recombinant vaccinia viruses encoding the HPV18 capsid proteins L1 and L2. As shown in Figure 1b, the pSVb DNA also accumulated at the density of full HPV virions. The peak at this position was seen whether or not the sample was treated with DNase I (not shown). Three different forms of the pSVb plasmid DNA could be observed in the peak fraction (Figure 1b, fraction 5). The fastest migrating form corresponds to the supercoiled plasmid, according to its migration relative to the supercoiled DNA size marker. The slowest form is probably the relaxed circular plasmid whereas the intermediate band migrates at a position similar to the linearised plasmid DNA (data not shown). This intermediate band was often observed upon DNA encapsidation. It is not clear if this form is already present in the intracellular packaged DNA or if it appears during the puri®cation of the pseudovirions. To analyse further the CsCl gradient-puri®ed particles, fractions containing DNase I-resistant pSVb DNA banding at the density of full virions were diluted and ultracentrifuged. The resuspended pellet was analysed for the presence of L1 and L2 proteins by immunoblot analysis. HPV18 L1 and L2 capsid proteins were detected in these fractions (Figure 1c). Since different antibodies were used to detect L1 and L2, the band intensities do not correspond to the relative amounts of the two proteins in the particles. When adsorbed to nickel grids, stained with uranyl acetate and examined by electron microscopy, these fractions were found to contain particles with the appearance of full virions (Figure 1d). Infectivity of HPV18 pseudovirions We then wanted to test if the HPV18 pseudovirions were infectious. Lysates from recombinant vaccinia virus-infected cells were treated with 0.5% (v/v) NP40 in order to inactivate the vaccinia virus (Oie, 1985) and HPV18 pseudovirions pelleted

Infectious HPV18 Pseudovirions

531

Figure 1. CsCl density-gradient analysis of HPV18 pseudovirions. The HPV18 L1 and L2 open reading frames were isolated from the original cloned HPV18 genome (a generous gift of H. zur Hausen) and inserted separately into the pHGS-1 vaccinia virus expression vector. Recombinant vaccinia viruses expressing L1 (vvL1) or L2 (vvL2) were then generated as described (Schmutz & Wittek, 1995). W12 cells, carrying about 1000 copies of episomal HPV16 DNA, were cultivated on a 3T3 ®broblast feeder layer in F medium (Flores & Lambert, 1997). Samples of 293T cells, cultured in Dulbecco's modi®ed Eagle's medium (DMEM) containing 10% (v/v) foetal calf serum (FCS) were transfected by the calcium phosphate method with 20 mg of pSVb plasmid DNA (Clonetech) per 15 cm dish. Cells (293T transfectants or W12) were cultured for three days at 37 C prior to co-infection with vvL1 and vvL2 at a multiplicity of infection (MOI) of ten for 24 hours, Cells were lysed with a dounce homogeniser, the nuclei sonicated, and pseudovirions puri®ed by sedimentation at 130,000 g for two hours through a 800 ml cushion of 40% (w/v) sucrose in PBS (containing 0.5 mM MgCl2 and 0.9 mM CaCl2) followed by CsCl density-gradient centrifugation (average density 1.32 g/ml, Beckman SW50 rotor, 45,000 rpm, 220,000 g, for 24 hours). Fractions of the CsCl gradients were analysed for the presence of HPV16 or pSVb DNA by agarose gel electrophoresis, Southern blotting and hybridisation with a speci®c probe. Density curves for each gradient (shown above) were calculated from refractive indices. Supercoiled DNA size markers are shown on the left. a, Encapsidation of HPV16 DNA in W12 cells. b, Encapsidation of pSVb plasmid DNA. c, The pSVb DNA-containing fraction was analysed further for the presence of capsid proteins by immunoblotting using rabbit antisera against HPV18 L1 (diluted 1:3000, lane 1) or HPV 18 L2 (diluted 1:4000, lane 2). d, Virus particles were mounted on carbon-coated nickel grids, stained with 1% (w/v) uranyl acetate and visualised with a Philips CM10 electron microscope. The bar represent 100 nm.

through a 40% (w/v) sucrose cushion. The resuspended pellet was then used as an inoculum to infect different cell types. Infected mouse C127 ®broblasts, stained with 5-bromo-4-chloro-3-indolyl-b,D-galactopyranoside (X-gal), are shown in Figure 2a. In this assay, the blue colour indicates expression of b-galactosidase. Cells of the human epithelial tumour line, HeLa, could also be infected by the pseudovirions (data not shown). In order to quantify the number of infected C127 cells and to con®rm that the infection was dose dependent, increasing amounts of pseudovirions were used to infect 1.5  105 cells. After staining, the blue cells were counted under the microscope. As shown in Figure 2b, the number of coloured cells was found to be proportional to the dose. To further check that the transfer of pSVb into C127 cells was due to infection by the pseudovirions and not free DNA, antiserum against HPV18 L1 protein or preimmune serum was added to pseudovirions prior to infection. The results, shown in Figure 2c, con®rmed that the infection was speci®cally inhibited

by the antiserum against HPV18 L1. In additional controls, up to 5 mg of pSVb plasmid DNA was added to cells under the same conditions as used for infection (Figure 2a, panel 5). This amount of DNA corresponded to about 2500 times the amount of pSVb DNA added to cells as pseudovirions, as measured by hybridisation and phosphorimager quanti®cation. These controls gave no blue cells. We conclude that infectious pseudovirions able to introduce the b-galactosidase gene into cells can be produced after encapsidation of pSVb DNA by HPV18 capsids. The results obtained above provided us with the possibility of developing a more simple assay of infectivity for HPV18. We constructed another SV40-based plasmid, pcDNA I-puro, which carries the puromycin-resistance gene. We then used this plasmid to produce pseudovirions. When 293T cells were infected with these pseudovirions, as in the experiments illustrated by Figure 2, then grown in the presence of puromycin, numerous puromycin-resistant cells were observed, whereas

532

Infectious HPV18 Pseudovirions

uninfected control cells were killed by puromycin (Figure 3). Requirements for plasmid DNA encapsidation by HPV L1 and L2 proteins

Figure 2. Infectivity of HPV18 pseudovirions containing the b-galactosidase gene. About 7  107 293T cells were transfected with pSVb DNA (as for Figure 1), incubated for three days at 37 C and co-infected at a MOI of 0.5 (to allow a longer period of infection) with recombinant vaccinia viruses expressing L1 and L2. After 48 hours, infected cells were sonicated, centrifuged at 5000 g for ten minutes to remove cell debris, and the supernatant treated for 45 minutes at 37 C with 0.5% NP40 to inactivate vaccinia viruses. The HPV18 pseudovirions were then centrifuged through a 40% sucrose cushion (as for Figure 1) and the pellet resuspended in 650 ml of PBS. After centrifugation at 5000 g for ten minutes to remove insoluble material, the supernatant was used as an inoculum for infection. a, Panels 1 to 4, infected C127 cells after 5-bromo-4-chloro-3-indolyl-b,Dgalactopyranoside (X-gal) staining, two days post-infection; panel 5, X-gal-stained C127 cells mock infected (see

In order to determine what the requirements are for DNA encapsidation, we ®rst wanted to know if a DNA packaging signal is necessary. The three DNAs (HPV16 DNA, pSVb and pcDNA I-puro) tested so far in our experiments all contain either the HPV or SV40 origin of replication. A signal for encapsidation has been reported to reside in the vicinity of the origin of DNA replication of SV40 (Dalyot-Herman et al., 1996; Oppenheim et al., 1992). In HPV18, the origin of DNA replication lies within the long control region of the viral genome (Demeret et al., 1995). We therefore tested the plasmid pSVb-18LCR, a pSVb vector into which the HPV18 long control region had been inserted, in DNA encapsidation analyses, as described above. No signi®cant difference in DNA encapsidation ef®ciency was observed whether or not this region was present (data not shown), indicating that the HPV origin of replication neither enhanced nor interfered with encapsidation. We also tested whether a plasmid that does not have the SV40 origin of replication can be encapsidated. The experiment illustrated by Figure 3a was repeated using the plasmid pCIpuro which did not contain the SV40 origin of replication. Puromycin-resistant cells were again obtained, indicating that the SV40 origin is not required for encapsidation to occur (Figure 3b and c). Although a detailed comparison of plasmids with and without the SV40 origin of replication was not done, we observed a somewhat higher yield of infectious particles when the origin was present. We wanted to determine if an interaction between SV40T and the capsid proteins, L1 and L2, played a role in the encapsidation of plasmid DNA in 293T cells. To detect such an interaction, extracts of vvL1- or vvL2-infected 293T cells were immunoprecipitated with antibodies to the two capsid proteins. Although SV40T was abundant in the extract (Figure 4a, lanes 1 and 4), it was not coprecipitated with L1 and L2 (Figure 4a, lanes 2 and 5). Instead, mini-chromosomes obtained from the text) with pSVb plasmid DNA. In panel 1 the magni®cation is lower than in panels 2 to 5. b, Number of blue C127 cells after infection with increasing amounts of virus inoculum. The error bars are based on four determinations. The amount of virus is indicated in microlitres of inoculum; 30 ml gave rise to 695 blue cells, corresponding to a titre of about 2.2  104 infectious pseudovirions per ml of inoculum. c, Fold inhibition of infection after incubation of the pseudovirions, for one hour at 4 C, with preimmune rabbit serum or antiserum (diluted 1:200) against HPV18 L1 produced in bacteria (pATH18L1XX1, described by Jenison et al. (1989), a generous gift from D. Galloway) and puri®ed as described by Jenison et al. (1988).

Infectious HPV18 Pseudovirions

533

Figure 3. Infectivity of HPV18 pseudovirions containing the puromycin-resistance gene. Samples of 293T cells were either transfected with pcDNA I-puro (a plasmid DNA bearing the SV40 origin of the replication and the puromycinresistance gene) or pCI-puro (containing only the puromycin-resistance gene without the SV40 origin of replication). Pseudovirions were prepared as described in the legend to Figure 2 and used to infect cells (293T and HaCaT). Two days later, cells infected by the pseudovirions were selected by the addition of 1.5 mg/ml of puromycin. Several days after the death of all the control cells, the puromycin-resistant cells were stained with 0.5% (w/v) methylene blue. a, Left dish, 293T cells infected with HPV18 pseudovirions containing the pcDNA I-puro plasmid; right dish, non-infected 293T control cells. b, Close-up view of panel a (left dish). c, Close-up view of HaCaT cells infected with HPV18 pseudovirions containing the pCI-puro plasmid after selection with puromycin.

pSVb-transfected 293T cells which had subsequently been infected with vvL2 were ef®ciently immunoprecipitated by antibodies against HPV18 L2 (Figure 4b, lane 2). When this experiment was repeated with L1 in place of L2, or in control immunoprecipitations, only very low amounts of

mini-chromosomes were recovered in the immunoprecipitates (Figure 4b, lanes 1, 3 and 4). In addition, the mini-chromosomes precipitated by anti-L2 did not contain detectable SV40T as judged from a Western blot analysis using anti-T antibodies (Figure 4c). These results show that the L2

Figure 4. a, Immunoprecipitation of L1 and L2 followed by Western blot analysis with antibodies against SV40T. Samples of 293T cells were infected with vvL1 or vvL2 at a MOI of ten. After a period of 36 hours, lysis buffer (10 mM Hepes (pH 7.8), 5 mM KCl, 1 mM EDTA, 0.1% (v/v) NP40 and 0.5 mM DTT) was added to the cells and extracts prepared according to the method described by Tack & Beard (1985). Soluble cell extracts of vvL1-infected 293T cells (lane 1) or vvL2-infected 293T cells (lane 4) were subjected to immunoprecipitation by antibodies against HPV18 L1 (lane 2) or antibodies against HPV18 L2 (lane 5). The cell extracts, immunoprecipitates and supernatants after immunoprecipitation using antibodies against L1 (lane 3) or L2 (lane 6) were separated on a SDS-10% (w/v) polyacrylamide gel, Western blotted and analysed with antibodies against SV40T. b, Southern blot and DNA hybridisation analysis of mini-chromosomes after immunoprecipitation by antibodies against HPV18 L1 or HPV18 L2. Samples of 293T cells were transfected with pSVb plasmid and subsequently infected with either vvL1 or vvL2. After 36 hours, pSVb mini-chromosomes were extracted according to the method of Tack & Beard (1985), and subjected to immunoprecipitation by antibodies against HPV18 L1 (lane 1) or HPV18 L2 (lane 2). To serve as controls, 293T cells were transfected with pSVb plasmid (but not infected with vvL1 or vvL2), treated as above and immunoprecipitated by antibodies against HPV18 L1 (lane 3) and HPV18 L2 (lane 4). The immunoprecipitates were resolved on a 1% (w/v) agarose gel, Southern blot and hybridised with a radiolabelled probe against pSVb. Supercoiled DNA size marker are shown on the left. c, Western blot analysis of pSVb mini-chromosomes immunoprecipitated by antibodies against HPV18 L2. Mini-chromosomes which were precipitated by antibodies against HPV18 L2 (lane 2) and the corresponding control (lane 1; as described for b above) were boiled in SDS loading buffer and subjected to electrophoresis on SDS-10% polyacrylamide gel, followed by Western blotting and detection with antibodies against SV40T. An extract of 293T cells was used as a marker for the position of SV40T on the gel (lane 3).

534 protein, but not L1 nor SV40T, is bound ef®ciently to the plasmid mini-chromosomes. They do not support a role for SV40T in encapsidation, but rather suggest that L2 may be important. This is also indicated by the observation that in our system, the L1 protein alone did not encapsidate DNA although it was able to form virus-like particles. The L2 protein has been shown to bind DNA in vitro (Zhou et al., 1994). Our experiments go further by showing that L2 is able to bind ef®ciently to non-HPV mini-chromosomes in cells (Figure 4). L2 has been reported to be needed for BPV-1 DNA encapsidation by BPV or HPV16 capsids (Roden et al., 1996). A role for L2 has been suggested by the demonstration (Day et al., 1998) that L2 induces the co-localisation of L1 and E2 protein to distinct nuclear structures identi®ed as promyelocytic leukaemia oncogenic domains (PODs). Since E2 binds to sites in the long control region of HPV DNA, this may increase the local concentration of capsid proteins and viral DNA in a de®ned space and so enhance encapsidation. However, a recent study on DNA encapsidation by HPV33 capsids (Unckell et al., 1997), using an SV40-based replication system, detected no increase in encapsidation of a plasmid bearing the HPV33 long control region even in the presence of the HPV33 E2. As SV40 origin-dependent replication is ef®cient and ampli®es episomes to a high copy number (Kirinaka et al., 1994), any effect of E2 and L2 to further concentrate the components of the encapsidation process in PODs might by super¯uous. In natural HPV infection, where the viral DNA copy number could be limiting, E2 might enhance DNA encapsidation by either increasing the viral copy number in association with E1 (Berg & Stenlund, 1997; Piccini et al., 1997; Seo et al., 1993) or by aiding the recruitment of HPV DNA to PODs via L2. Interestingly, SV40T has also been reported to accumulate in the vicinity of PODs (Jiang et al., 1996). Hence, it was suggested (Day et al., 1998) that SV40T and its binding sites in the SV40 origin of replication can functionally replace E2 and its binding sites in HPV DNA, in their role in encapsidation. However, our experiments do not support this hypothesis as we found that the encapsidation of plasmid DNA by HPV L1 and L2 is independent of the SV40 origin of replication. Taken together, our results show that HPV18 L1 and L2 proteins expressed from recombinant vaccinia viral vectors can encapsidate plasmid DNAs which have been transfected into cells. There appears to be no requirement for any speci®c DNA sequences nor is the size of encapsidated DNA critical. Plasmid sizes of 5.4 kb (pcDNA I-puro), 5.5 kb (pCIpuro), 6.9 kb (pSVb) and 7.9 kb (pSVbLCR and HPV16 DNA) can be ef®ciently packaged. The lack of a requirement for the SV40 origin ®ts with the apparent absence of SV40T on the mini-chromosomes bound by the L2 protein. Nevertheless, the SV40 replication origin may be

Infectious HPV18 Pseudovirions

useful to increase the number of plasmids in SV40T-containing cells in order to maximise the production of pseudovirions. They could transfer either the b-galactosidase gene or the puromycinresistance gene to infected cells, suggesting that different kinds of genes could be incorporated into HPV18 capsids. The pseudovirions were able to infect various cell lines of mouse or human origin including C127, HeLa, HaCaT and 293T cells. It is known that HPV virus-like particles can bind to (Roden et al., 1994), and be internalised by (Muller et al., 1995) different cell types. Our observations both con®rm that the cell type speci®city of HPV is not determined by viral uptake and go further by showing that neither is it determined by the process of viral decapsidation in different cell types. The pseudovirions described here should prove to be useful tools to test the ef®ciency of antibodies and other agents to neutralise HPV18 infection in vitro and in vivo. This encapsidation system is ¯exible in that plasmid DNAs carrying a marker gene may be packaged without a requirement for any additional speci®c DNA sequences. From the results above the only requirement for DNA encapsidation is the HPV L1 and L2 capsid proteins. This raises the possibility of designing novel pseudovirions for use as vectors to transfer DNA into different cell types.

Acknowledgements We thank H. zur Hausen, J. Firzlaff and D. Galloway for cloned HPV18 DNAs, P. Lambert for W12 cell clones and R. Wittek for advice on the production of recombinant vaccinia viruses. We are also grateful to B. Hirt and our other colleagues for helpful discussions and to B. Bentele for cell culture. This work was supported by the University of Lausanne, the Swiss Cancer League and the Swiss National Science Foundation.

References Baker, T. S., Newcomb, W. W., Olson, N. H., Cowsert, L. M., Olson, C. & Brown, J. C. (1991). Structures of bovine and human papillomaviruses. Analysis by cryoelectron microscopy and three-dimensional image reconstruction. Biophys. J. 60, 1445± 1456. Belnap, D. M., Olson, N. H., Cladel, N. M., Brown, W. W., Brown, J. C., Kreider, J. W., Christensen, N. D. & Baker, T. S. (1996). Conserved features in papillomavirus and polyomavirus capsids. J. Mol. Biol. 259, 249± 263. Berg, M. & Stenlund, A. (1997). Functional interactions between papillomavirus E1 and E2 proteins. J. Virol. 71, 3853± 3863. Carter, J. J., Hagensee, M. B., Kuang, Lee S., McKnight, B., Koutsky, L. A. & Galloway, D. A. (1994). Use of HPV 1 capsids produced by recombinant vaccinia viruses in an ELISA to detect serum antibodies in people with foot warts. Virology, 199, 284±291. Cereghini, S. & Yaniv, M. (1984). Assembly of transfected DNA into chromatin: structural changes in

Infectious HPV18 Pseudovirions the origin-promoter-enhancer region upon replication. EMBO J. 3, 1243± 1253. Dalyot-Herman, N., Ben-nun-Shaul, O., Gordon-Shaag, A. & Oppenheim, A. (1996). The simian virus 40 packaging signal ses is composed of redundant DNA elements which are partly interchangeable. J. Mol. Biol. 259, 69± 80. Day, P. M., Roden, R. B. S., Lowry, D. R. & Schiller, J. T. (1998). The papillomavirus minor capsid protein, L2, induces localization of the major capsid protein, L1, and the viral transcription/replication protein, E2, to PML oncogenic domains. J. Virol. 72, 142± 150. Demeret, C., Le, Moal M., Yaniv, M. & Thierry, F. (1995). Control of HPV18 DNA replication by cellular and viral transcription factors. Nucl. Acids Res. 23, 4777± 4784. Flores, E. R. & Lambert, P. F. (1997). Evidence for a switch in the mode of human papillomavirus type 16 DNA replication during the viral life cycle. J. Virol. 71, 7167± 7179. Hagensee, M. E., Yaegashi, N. & Galloway, D. A. (1993). Self-assembly of human papillomavirus type 1 capsids by expression of the L1 protein alone or by coexpression of the L1 and L2 capsid proteins. J. Virol. 67, 315± 322. Hagensee, M. E., Olson, N. H., Baker, T. S. & Galloway, D. A. (1994). Three-dimensional structure of vaccinia virus-produced human papillomavirus type 1 capsids. J. Virol. 68, 4503± 4505. Jenison, S. A., Firzlaff, J. M., Langenberg, A. & Galloway, D. A. (1988). Identi®cation of immunoreactive antigens of human papillomavirus type 6b by using Escherichia coli-expressed fusion proteins. J. Virol. 62, 2115± 2123. Jenison, S. A., Yu, X. P., Valentine, J. M. & Galloway, D. A. (1989). Human antibodies react with an epitope of the human papillomavirus type 6b L1 open reading frame which is distinct from the type-common epitope. J. Virol. 63, 809± 818. Jeon, S., Allen-Hoffman, B. L. & Lambert, P. F. (1995). Integration of human papillomavirus type 16 into the human genome correlates with a selective growth advantage of cells. J. Virol. 69, 2989± 2997. Jiang, W. Q., Szekely, L., Klein, G. & Ringertz, N. (1996). Intranuclear redistribution of SV40T, p53, and PML in a conditionally SV40T-immortalized cell line. Expt. Cell Res. 229, (2), 289± 300. Kirinaka, H., Kamihira, M., Iijima, S. & Kobayashi, T. (1994). A new runaway type episomal vector for mammalian cells based on a temperature-sensitive simian virus 40 and inducible erythropoietin production. Appl. Microbiol. Biotech, 41, 591±596. Kirnbauer, R., Booy, F., Cheng, N., Lowy, D. R. & Schiller, J. T. (1992). Papillomavirus L1 major capsid protein self-assembles into virus-like particles that are highly immunogenic. Proc. Natl Acad. Sci. USA, 89, 12180± 12184. Kirnbauer, R., Taub, J., Greenstone, H., Roden, R., Durst, M., Gissmann, L., Lowy, D. R. & Schiller, J. T. (1993). Ef®cient self-assembly of human papillomavirus type 16 L1 and L1-L2 into virus-like particles. J. Virol. 67, 6929± 6936. Muller, M., Gissmann, L., Cristiano, R. J., Sun, X. Y., Frazer, I. H., Jenson, A. B., Alonso, A., Zentgraf, H. & Zhou, J. (1995). Papillomavirus capsid binding and uptake by cells from different tissues and species. J. Virol. 69, 948± 954.

535 Oie, M. (1985). Reversible inactivation and reactivation of vaccinia virus by manipulation of viral lipid composition. Virology, 142, 299± 306. Oppenheim, A., Sandalon, Z., Peleg, A., Shaul, O., Nicolis, S. & Ottolenghi, S. (1992). A cis-acting DNA signal for encapsidation of simian virus 40. J. Virol. 66, 5320± 5328. Pear, W. S., Nolan, G. P., Scott, M. L. & Baltimore, D. (1993). Production of high-titer helper-free retroviruses by transient transfection. Proc. Natl Acad. Sci. USA, 90, 8392± 8396. Piccini, A., Storey, A., Romanos, M. & Banks, L. (1997). Regulation of human papillomavirus type 16 DNA replication by E2, glucocorticoid hormone and epidermal growth factor. J. Gen. Virol. 78, 1963± 1970. Roden, R. B., Kirnbauer, R., Jensen, A. B., Lowy, D. R. & Schiller, J. T. (1994). Interaction of papillomaviruses with the cell surface. J. Virol. 68, 7260± 7266. Roden, R. B., Greenstone, H. L., Kirnbauer, R., Booy, F. P., Jessie, J., Lowy, D. R. & Schiller, J. T. (1996). In vitro generation and type-speci®c neutralization of a human papillomavirus type 16 virion pseudotype. J. Virol. 70, 5875± 5883. Rose, R. C., Bonnez, W., Reichman, R. C. & Garcea, R. L. (1993). Expression of human papillomavirus type 11 L1 proteins in insect cells: in vivo and in vitro assembly of virus-like particles. J. Virol. 67, 1936± 1944. Sasagawa, T., Pushko, P., Steers, G., Gschmeissner, S. E., Hajibagheri, M. A., Finch, J., Crawford, L. & Tommasino, M. (1995). Synthesis and assembly of virus-like particles of human papillomaviruses type 6 and type 16 in ®ssion yeast Schizosaccharomyces pombe. Virology, 206, 126± 135. Schmutz, C. & Wittek, R. (1995). Release of extracellular particles by recombinant vaccinia virus over-expressing the major envelope protein p37 K. J. Gen. Virol. 76, 2963± 2968. Seo, Y. S., Muller, F., Lusky, M., Gibbs, E., Kim, H. Y., Philips, B. & Hurwitz, J. (1993). Bovine papilloma virus (BPV)-encoded E2 protein enhances binding of E1 protein to the BPV replication origin (published erratum appears in Proc. Natl Acad. Sci. USA, (1993) Jun 15;90(12):5878). Proc. Natl Acad. Sci. USA, 9, 2865± 2869. Stanley, M. A., Browne, H. M., Appleby, M. & Minson, A. C. (1989). Properties of a non-tumorigenic human cervical keratinocyte cell line. Int. J. Cancer, 43, 672± 676. Stauffer, Y. & Beard, P. (1995). Revised major capsid protein sequence brings HPV18 more into line with other papillomaviruses. Papillomavirus Rep. 6, 163± 164. Tack, L. & Beard, P. (1985). Both trans-acting factors and chromatin structure are involved in the regulation of transcription from the early and late promoters in simian virus 40 chromosomes. J. Virol. 54, 207± 218. Taichman, L. B. & LaPorta, R. F. (1987). The papillomaviruses. In The Papovaviridae (Salzman, N. P. & Howley, P. M., eds), vol. 2, pp. 109±138, Plenum, New York. Trus, B. L., Roden, R. B., Greenstone, H. L., Vrhel, M., Schiller, J. T. & Booy, F. P. (1997). Novel structural features of bovine papillomavirus capsid revealed Ê resolby a three-dimensional reconstruction to 9 A ution. Nature Struct. Biol. 4, 413± 420.

536

Infectious HPV18 Pseudovirions

Unckell, F., Streeck, R. E. & Sapp, M. (1997). Generation and neutralization of pseudovirions of human papillomavirus type 33. J. Virol. 71, 2934± 2939. Volpers, C., Schirmacher, P., Streeck, R. E. & Sapp, M. (1994). Assembly of the major and the minor capsid protein of human papillomavirus type 33 into virus-like particles and tubular structures in insect cells. Virology, 200, 504± 512. Zhou, J., Sun, X. Y., Stenzel, D. J. & Frazer, I. H. (1991). Expression of vaccinia recombinant HPV16 L1 and L2 ORF proteins in epithelial cells is suf®cient for assembly of HPV virion-like particles. Virology, 185, 251± 257.

Zhou, J., Sun, X.-Y., Louis, K. & Frazer, I. H. (1994). Interaction of human papillomavirus (HPV) type 16 capsid proteins with HPV DNA requires an intact L2 N-terminal sequence. J. Virol. 68, 619± 625. zur Hausen, H. (1996). Papillomavirus infections ± a major cause of human cancers. Biochim. Biophys. Acta, 1288, F55± 78. zur Hausen, H. & de Villiers, E.-M. (1994). Human papillomaviruses. Annu. Rev. Microbiol. 48, 427± 447. zur Hausen, H. & Gissmann, L. (1980). Papillomaviruses. In Viral Oncology (Klein, G., ed.), pp. 433± 445, Raven Press, New York.

Edited by M. Yaniv (Received 5 February 1998; received in revised form 3 August 1998; accepted 4 August 1998)