Papillomavirus pseudovirions that encapsidate a marker plasmid instead of the viral genome are widely used to study HPV biology and infection, circumventing the difficulties to obtain biochemical quantities of native virions 1224. Using such HPV16 and HPV33 pseudovirions, we first determined the neutralizing potential of various HPV-specific antibodies (Fig. 1). Three days post infection with HPV pseudovirions, infection was monitored by the number of cells with green nuclear fluorescence, caused by transmission of a GFP marker gene to the nucleus via the HPV vector. Pseudovirus infection in the presence of the HPV33-specific neutralizing monoclonal antibodies (mAbs H33.B6, H33.J3, and H33.E12) was abolished only with pseudovirions of the respective type. Moreover, we used the recently described mAb H16.56E, generated after immunization with HPV16 VLPs, and also observed type-specific neutralization, demonstrating the validity of this surrogate system for use in testing papillomavirus neutralizing antibodies (Fig. 1A). Binding of these antibodies to conformationally intact HPV VLPs bound to Heparin-BSA-coated Elisa plates confirmed the selective specificity of antibodies H33.J3, H33.B6 and H33.E12 for HPV33 (Fig. 1B). Subsequent experiments were performed to characterize and fine map the epitopes of these HPV33-specific antibodies.

For various HPV types it has been reported that type-specific monoclonal antibodies primarily reside in surface-exposed hyper variable loops. Our experimental approach for defining residues involved in neutralization of HPV33 by mAbs H33.B6, H33.J3 and H33.E12 was therefore based on the exchange of type-specific loop sequences between the closely related papillomavirus types 16 and 33. Poorly conserved regions in HPV major capsid proteins L1 were identified by sequence alignment and localized by RasMol, based on the coordinates of HPV16 (Pdb file 1DZL). As shown in Fig. 2, 30 divergent amino acids between HPV33 and HPV16 were localized in 4 surface-exposed hyper variable loops, named BC (aa 51–59), DE (aa 132–140), FG (260–291), and HI (346–358), according to the HPV16 L1-structure reported by Chen et al. 17. In HPV33 L1, the FG loop was found to consist of two separate hyper variable regions, designated in this paper as FGa (260–270) and FGb (282–291) (Fig. 2).

To further characterize the epitopes of HPV33-specific antibodies, hybrid virus-like particles were designed in which type-specific sequences in the major capsid protein L1 of HPV33 were replaced by corresponding amino acids of HPV16, eliminating the putative epitopes. Vice versa, HPV33-specific sequences were introduced into HPV16 L1 for ectopic expression. Ten different hybrid L1 proteins (HPV33:16BC; HPV33:16DE; HPV33:16FGa; HPV33:16FGb; HPV33:16HI; HPV16:33BC; HPV16:33DE; HPV16:33DE/FGa, HPV16:33DE/FGb, and HPV16:33HI) were constructed and expressed in HUTK^--143B cells. Western blot analysis revealed that all hybrid proteins were expressed at similar levels (data not shown). Binding of monoclonal antibodies to hybrid L1 protein was first tested by immunofluorescence under non-denaturing conditions (Fig. 3). Reactivity of H33.J3 with hybrid particles was lost by exchanging the BC loop but was retained after replacement of the other loops (Fig. 3A). This suggests that the BC loop is the binding site for H33.J3 and that exchange of neighboring surface loops results in conformationally intact L1 assemblies. HPV16 L1 hybrid particles became reactive with this antibody when the HPV33 BC loop, but not the DE, FG, and HI loops, were ectopically expressed on HPV16 (Fig 3B). Reactivity of H16:56E with HPV16:33BC was retained, suggesting that this antibody recognizes a different epitope and, in addition, that this hybrid L1 protein also forms conformationally intact assemblies.

The epitope recognized by the H33.B6 antibody was shown to be more complex, as exchange of loops DE or FGb resulted in the loss of reactivity. Vice versa, introduction of both HPV33 loops into HPV16 L1 transferred reactivity of H33.B6 to the HPV16:33DE/FGb hybrid (Fig. 3B). Surprisingly, exchange of the DE loop alone was sufficient to render HPV16:33DE reactive with this antibody. However, the concomitant exchange of the DE and FGa loops abrogated the binding of H33.B6 with HPV16:33DE/FGa. Therefore, without being part of the epitope, the FGa loop has significant influence on the conformation of the DE loop and thus contributes to the conformation recognized by H33.B6.

The monoclonal antibody H33.E12 binding site also displays a high level of complexity. Individual swapping of loops DE, FGa, and FGb results in the loss of binding to HPV33 hybrid L1 proteins (Fig. 3A), whereas the exchange of the BC and HI loops had no effect. In contrast to H33.B6, transfer of individual or two combined HPV33 loops onto HPV16 did not result in the reconstruction of the epitope. Unfortunately, we were not successful in the construction of hybrid 16L1 protein carrying all three HPV33 loops required for binding of H33.E12. Using our HPV16:33 chimeric particles, we could also show that the FGa loop is an important part of the H16.56E epitope, since only HPV33:16FGa particles were recognized by this antibody. Vice versa, the fact that all HPV16:33 chimeras were still recognized by this antibody demonstrates that the H16.56E binding site is not a one-loop epitope but rather formed by discontiguous sequences of the L1 protein.

To confirm the validity of our immunofluorescence approach for measuring conformation-dependent antibody binding, we generated and purified hybrid HPV33:16BC VLPs, using recombinant vaccinia viruses and HPV16:33BC after transfection of codon-optimized L1. Reactivity of the monoclonal antibodies with VLPs was measured in a heparin-BSA ELISA (Fig. 4). Swap of the BC loop resulted in the loss of reactivity of hybrid HPV33:16BC with H33.J3 and a gain of reactivity with H16:33BC. Binding of H33.B6 and H16.56E were not affected by this exchange and solely dependent on the backbone (33L1 for HPV33:16BC and 16L1 for HPV16:33BC) of the chimeric L1 molecules.

To exemplarily demonstrate that the transfer of HPV33-specific epitopes is functional, hybrid pseudovirions HPV16:33BC were generated that contain the HPV33 BC loop in the context of HPV16, following a published protocol 24. The mutant was cotransfected with the HPV16 wtL2 expression plasmid and a GFP-expressing marker plasmid to be packaged. The mutant protein efficiently assembled with the L2 protein and the marker plasmid into pseudoviruses that were used in subsequent neutralization assays. As shown in Fig. 5, HPV16:33BC and wt HPV33, but not wt HPV16 pseudovirions, were efficiently neutralized by H33.J3. Hybrid viruses were not neutralized by H33.B6 and H33.E12. These data clearly demonstrate the functional expression of the heterotypic epitope on HPV16.

The osteosarcoma cell line HuTK^-143B 31 was grown at 37°C in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal calf serum and antibiotics. The human embryonic kidney cell line 293TT 24 was maintained in DMEM/10% FCS with 1% Glutamax I and 1% non-essential amino acids (Invitrogen). Three conformation-dependent, neutralizing mouse monoclonal antibodies, H33.B6 (IgG2a), H33.E12 (IgG2a) and H33.J3 (IgG2b), respectively, with specificity for HPV33 were kindly provided by N. D. Christensen, Hershey, PA. The HPV16-neutralizing mAb H16.56E, was generated by immunization of mice with HPV16 VLPs, and used as previously reported 3233.

Type-specific amino acids in hypervariable loops of the HPV33- and HPV16 L1 capsid proteins were identified by CLUSTAL amino acid sequence alignment 34. For generation of HPV33:16 hybrid virus-like-particles, various loop sequences of the HPV33 L1 capsid protein (BC, DE, FGa, FGb, HI; Fig. 2) were exchanged by the corresponding amino acids of HPV16 by introducing codon-modified sequences from p16L1h 35 into pTM33L1 12. HPV16:33 hybrids were generated reciprocally, using the codon-modified pUF3hu16L1 vector and codon-modified loop sequences of HPV33 L1. Overlap extension PCR 36 was used to introduce multiple substitutions simultaneously. Pairs of PAGE-purified mutagenesis primers with 100 % complementarity (Table 1) were purchased from Invitrogen and PCR was carried out using puReTaq Ready-to-go PCR-beads (Amersham Biosciences). In a first step two separate PCR reactions were prepared to generate fragments in forward and reverse orientations, both carrying the desired mutations. Thereby, the reverse mutagenesis primer was used together with an outer forward primer, the forward mutagenesis primer in combination with an outer reverse primer. L1 expression plasmids were used as template and PCR was performed for 40 cycles with denaturation at 95°C for 45 seconds, annealing at 42°C for 1 min and elongation at 72°C for 2 min. PCR fragments generated by these PCRs were purified by agarose gel electrophoresis, followed by Jetsorp gel extraction prior to their use in subsequent reactions. Because of an average overlap of 60 bp between appropriate fragments, these sequences were hybridized by pre-extension PCR 37, in which the 3'overlap of each strand acts as a primer for the extension of the complementary strand. This was done by 2 cycles with denaturation at 95°C for 5 min and annealing at 72°C for 2 min. Resulting products were PCR-amplified by addition of the outer primers of step 1 (conditions: denaturation at 95°C, 45 sec; annealing at 50–56°C, 1 min; elongation at 72°C, 2 min; 35 cycles). Subsequently, the gel-purified mutant L1 amplimers (sized between 800–1900 bp) were cloned into singular restriction sites in the transfer vectors pUF3hu16L1 or pTM33L1 to generate the HPV16/HPV33 or HPV33/HPV16 hybrid L1-constructs. Ligation mixtures were transfected into chemically competent cells of E. coli (DH5α). Colonies containing the desired mutations were identified by their newly introduced restriction sites or directly by sequencing. If only one of the two fragments could be generated in the first PCR round, the purified fragment was used in a following PCR as a megaprimer. The fragment was added in excess over the plasmid template and combined with a counter-directed common primer, using the following conditions for a total of 35 cycles: denaturation at 95°C for 45 sec, annealing at 65°C for 1 min, elongation at 72°C for 2 min. Generation of HPV16:33-hybrids with double loop exchanges occurred successively. One loop was introduced by the approach described above. To introduce the second loop, a forward primer was generated using the hybrid L1 as a template. Subsequently, the fragment served as a megaprimer to amplify the complete expression plasmid with high-fidelity Pwo DNA polymerase for 18 cycles (denaturation for 30 sec at 95°C, annealing for 1 min at 50°C, elongation for 14 min at 72°C). The PCR product was then digested with DpnI to eliminate methylated template DNA and the remaining mutant plasmids were expressed in E. coli.

HuTK^- cells were grown on glass coverslips overnight, infected with the vaccinia helper virus VTF7-3 for 1 h (MOI of 5) and subsequently transfected using Lipofectamin plus (Invitrogen) and 1 μg transfer plasmid pTM1 carrying wt or mutated HPV33L1 sequences under the control of a T7-promotor. Expression of the pUF3 vector-based wt or hybrid HPV16 L1-constructs occurred by lipofection without any helper viruses. After an incubation period of 10 – 24 h at 37°C cells were fixed with 2 % paraformaldehyde for 20 min at room temperature, permeabilized with 0.1 % Nonident P-40 for 15 min and subsequently blocked in 5 % goat serum dissolved in PBS. Incubations with primary mAbs and secondary Cy2-conjugated Affinipure goat anti-mouse IgG (Jackson Immunoresearch Products) were carried out for 1 h at 37°C. Thereafter, coverslips were washed with PBS several times, stained with 0.2 μg/ml Bis-benzimide trihydrochloride (Hoechst 33342; Sigma) and mounted onto slides by using Fluoprep mounting medium (BioMérieux). Pictures were taken using a Zeiss Axiovert 200 M microscope and a Zeiss Axiocam digital camera. The appropriate Axiovision Software 3.0 was used for merging pictures.

HPV33-VLPs and pseudovirions were produced in HuTK^- cells by infection with recombinant vaccinia viruses vac33L1, vac33L2 and helper virus VTF7-3, as described previously 1238. For generation of pseudovirions, cells were transfected 24 h prior to infection with a marker plasmid encoding a dimeric green fluorescent protein (GFP), resulting in HPV particles containing the GFP reporter DNA. Forty-four hours post infection VLPs/PsV were extracted from nuclei by sonication in hypotonic buffer supplemented with 0.5% NP-40 and purified by buoyant caesium chloride density gradients. HPV16 pseudovirions were prepared as described previously 24 by co-transfection of 293TT cells with pUF3hu16L1 wt or pUF3hu16/33L1-hybrid plasmids, together with pUF3hu16L2 wt and the pEGFPGFP marker plasmid. Subsequent to incubation at 37°C for 48 h cells were lysed and pseudovirions were purified on an OptiPrep gradient. Thereby, lysis of cells was achieved by adding the non-ionic detergent Brij58 (Sigma) at a final concentration of 0.5 % in DPBS supplemented with 9.5 mM MgCl₂. Lysates were digested over night at 37°C with 2 U of Benzonase (Sigma) to complete virus maturation 39. Subsequently the lysate was mixed with a 0.17 volume of 5 M NaCl, clarified by centrifugation at 1500 × g for 10 min, loaded on top of an OptiPrep step gradient (27%/33%/39% OptiPrep in DPBS-800 mM NaCl) and centrifuged for 4h at 234.000 × g. After centrifugation, 250 μl-fractions were collected by bottom puncture of the tubes and 1 μl of each fraction was tested in a pseudovirus infection assay.

Human embryonic kidney 293TT cells were grown overnight in 24-well plates and infected with 1 μl of HPV pseudovirions (PsV) in a total volume of 500 μl DMEM. Cells were grown at 37°C for 72 h and infectious events were monitored by counting cells with green nuclear fluorescence. To perform virus neutralization assays, PsV were bound to cells for 1 h at 4°C, unbound virions were removed and various dilutions of HPV-specific neutralizing antibodies were added to cells in a total volume of 250 μl DMEM. After 1 h at 37°C the culture medium was replaced and incubation was continued for 72 h.

VLP-ELISAs were used to study the interaction of conformationally intact VLPs with heparin and performed as previously described 2940. Briefly, polysorb microtiter plates (NUNC, Wiesbaden, Germany) were coated overnight with 100 ng of heparin-BSA/well in phosphate-buffered saline (PBS), washed and subsequently blocked with BSA (50 μg/ml) for 30 minutes. Plates were again washed, 100 μl VLPs (1 μg/ml) were added and incubated for 1 h at 37°C. Unbound particles were eliminated by washing. HPV type-specific antibodies H16.56E, H33.B6, H33.J3 and H33.E12 were added for 1 h at 37°C at the indicated concentrations (1:100 – 1:5000). After washing three times with PBS-Tween 20 (PBS-T), 100 μl horseradish peroxidase-coupled secondary antibodies (goat anti-mouse IgG; 1:10.000 in PBS-T) obtained from Jackson Immunochemicals were added and incubated for additional 30 min at 37°C. Plates were washed and developed with ready to use trimethyl benzidine (KPL). The reaction was stopped after 10 min at 37°C with 100 μl 1N HCl. Absorbance was measured at 450 nm using a Multiscan EX (Thermo Life Sciences).

The RasMol program is a molecular graphics visualisation tool for macromolecular structures 41. Localization of amino acids in loops structures of capsid protein L1 from HPV16 or HPV33 was based on the atomic coordinates of the HPV16 major capsid protein L1 17 and visualized using the PDB file 1DZL in the RasMol program.