Identification of Broad-Genotype HPV L2 Neutralization Site for Pan-HPV Vaccine Development by a Cross-Neutralizing Antibody

Human Papillomavirus (HPV), a non-enveloped, double-stranded DNA virus, is responsible for 5% of human cancers. The HPV capsid consists of major and minor structural proteins, L1 and L2. L1 proteins form an icosahedral shell with building blocks of the pentameric capsomere, and one L2 molecule extends outward from the central hole of the capsid. Thus, L2 is concealed within L1 and only becomes exposed when the capsid interacts with host cells. The low antigenic variation of L2 means that this protein could offer a target for the development of a pan-HPV vaccine. Toward this goal, here we describe an anti-L2 monoclonal antibody, 14H6, which broadly neutralizes at least 11 types of HPV, covering types 6, 11, 16, 18, 31, 33, 35, 45, 52, 58 and 59, in pseudovirion—based cell neutralization assay. The mAb 14H6 recognizes a minimal linear epitope located on amino acids 21 to 30 of the L2 protein. Alanine scanning mutagenesis and sequence alignment identified several conserved residues (Cys22, Lys23, Thr27, Cys28 and Pro29) that are involved in the 14H6 binding with L2. The epitope was grafted to several scaffolding proteins, including HPV16 L1 virus-like particles, HBV 149 core antigen and CRM197. The resultant chimeric constructs were expressed in Escherichia coli and purified with high efficiency. Immunization with these pan-HPV vaccine candidates elicited high titers of the L2-specific antibody in mice and conferred robust (3-log) titers of cross-genotype neutralization, including against HPV11, 16, 18, 45, 52, 58 and 59. These findings will help in the development of an L2-based, pan-HPV vaccine.


Introduction
In this study, we screened for HPV L2 broad-spectrum neutralizing antibodies from mice immunized with recombinant HPV16-L2 protein. The mAb 14H6, recognizing a linear epitope, was found to neutralize a remarkably broad range of HPV genotypes including HPV6, 11, 16, 18, 31, 33, 45, 52, 58 and 59 PsVs. Further, we grafted the peptide to various scaffold antigens, including CRM197, CRM389, CRMA, HBV core antigen, and HPV16 L1 VLPs, and measured the immunogenicity of L2 broad-spectrum epitope in mice. Overall, we demonstrate that 14H6 identifies a minimally conserved, broad-genotype neutralization site on HPV L2 and its corresponding epitope is immunogenic in mice with the aid of the scaffold proteins. These findings provide insight for an L2-based pan-HPV vaccine design and offer a promising vaccine candidate for a broad-spectrum HPV vaccine.

14H6 mAb, a broad, cross-neutralizing L2 antibody
Previous research has shown that the HPV-L2 protein contains numerous, broad, crossneutralizing epitopes [20,21]. To unravel more of these L2 epitopes, we immunized Balb/c mice with recombinant HPV16 L2 C50 protein, derived using the Escherichia coli expression system, and raised monoclonal antibodies by L2-based ELISA and a pseudovirion-based cell neutralization assay. The immunogen HPV16L2 C50 protein was resolved using SDS-PAGE (Fig 1A), and reacted with immune serum of HPV16L2 ( Fig 1B). Eventually, 17 hybridomas secreting anti-HPV16-L2 monoclonal antibodies (mAbs) were deemed positive and, of those, 4 mAbs were plotted on the log-scaled y-axis against various genotypes of HPV-PsVs on the x-axis. The bar graph was prepared using Prism GraphPad 5.0. All mAb samples were adjusted to an initial concentration of 1.0 mg/ml prior to two-fold serial dilutions. Flow cytometry was used to detect the number of HPVpseudovirus-infected enhanced green fluorescent protein (EGFP)-expressing cells. The collected data were processed to calculate the neutralizing titer and IC 50 for each mAb, as shown in Table 1.

Epitope mapping and key interaction residues for 14H6
We first used western blotting with N-terminal (aa 1-265) segments of HPV L2 to identify 14H6's epitope, and found it located on the L2 N-terminus. In primary screening, an aa stretch (aa 1 to 265) of HPV16-L2 was equally divided into four overlapping 65-mer peptides (PA, PB, PC, and PD) to ascertain where mAb 14H6 interacts. The mAb 14H6 only reacted with PA of HPV L2, corresponding to aa 1 to 65 (Fig 2A, upper panel). In a secondary screening, we synthesized seven overlapping 15-mer peptides spanning aa 1 to 75 (PA1, PA2, PA3, PA4, PA5, PA6, and PA7), and found positive activity with PA3 using ELISA, thus further narrowing down the location of 14H6's epitope to be within aa 21 to 35. Finally, we used progressive, single aa reduction from both ends of PA3, generating twelve fragments (P A3-1, P A3-2, . . ., and P A3-12) and found that peptides beyond aa 21 at the N-terminus or without aa 30 at the Cterminus lost reactivity with mAb 14H6, thus identifying a strip of aa 21 to 30 with sequence of TCKQAGTCPP as 14H6's epitope (Fig 2A, lower panel).
Sequence alignment of this aa stretch among genotypes was performed using MEGA 5.05 and Weblogo programs, and the results showed that aa 24 and 25 are highly varied among HPV subtypes (Fig 2B and 2C). Furthermore, alanine scanning mutagenesis and subsequent ELISA demonstrated that mutation of C22A, K23A, T27A, C28A or P29A abrogated the binding between 14H6 mAb and L2, and mutation of T21A or P30A decreased the binding (Fig 2D), indicating that these residues are involved in the interaction between L2 and mAb 14H6. Campos et al. [32] has reported that C22 and C28 are essential for papillomavirus infectivity, which is consistent with our findings. In addition, Rubio et al. [21] found that aa 30 (serine for HPV31 and HPV59, but proline for other types) plays an essential role in the interaction of HPV31 L2 and mAb K4L2 (epitope of aa 20-38), with mAb K4L2 presenting a lower neutralization for HPV31 PsV as compared with the others tested. Similarly, our mAb 14H6 also discriminates between P30 and S30 of 14H6 peptide reactivity for HPV types ( Fig 2E). Overall, the conservation of most of the residues from 14H6's epitope among HPV genotypes may imply that the epitope has a broad neutralization capacity, which should be exploited for the development of a pan-HPV vaccine immunogen. Chimeric constructs of 14H6's epitope grafted to HPV16-L1, HBc149 and CRM197/389/A scaffold protein Using PA3-11 as a core region, we next designed six HPV16-L2 14H6 peptides, 14H6a-f, for insertion into the scaffold proteins (Fig 3). Centering on aa 21 to 30, each peptide became progressively longer with the addition of an aa to each end; it was hypothesized that such extensions might allow for the better exposure of the peptide epitope on the scaffold proteins. It was previously shown that there is a limited insertion capacity for maintaining VLP assembly in the HPV16-L1-α4 loop [33]. Therefore, prior to generating our constructs, we used Accelrys Discovery Studio software (San Diego, CA) to visualize and predict the spatial position of the heterologous insert on the homology-modeled structures of the 14H6 chimeras (Fig 3). 14H6a-f fragments were grafted into position at aa 130 in the DE loop and aa 426 in the α4 loop of HPV16 L1; the major immune-dominant region (MIR) (79/83) of HBc149; and directly following the C-terminus of CRM197, CRM389, and CRMA. These built structural models suggested that the inserted peptides would be fully exposed under thermodynamically favorable condition and, therefore, would probably present L2-specific antigenicity and immunogenicity upon immunization. All of the 14H6a-f chimeras were expressed in E. coli and purified to considerably highly purity, as shown in SDS-PAGE and 14H6-binding western blotting (Fig 4A). The chimeras with potential VLP forms were visualized in negative stain transmission electron microscopy  and CRMA, respectively. The moieties in red depict the flexible aa stretches flanking both sides of the L2 inserts in the HBcore149 and following the Cterminus of CRM197, CRM389 and CRMA, which are included for optimal surface display. Homology structural modeling of the recombinant VLPs and proteins was implemented using MODELER module of Accelrys Discovery Studio 2.5. All illustrative models were prepared using PyMol [51].  (TEM), which revealed that all six HPV16-L1-DE-14H6a-f proteins were able to form uniform VLPs ( Fig 4B). Particles with an average diameter of 50 nm were observed, similar to that of wild-type VLPs. However, for the α4 loop chimeras, HPV16-L1-α4-14H6a-f, we noticed a declining trend in particle size and conformity concomitant with an increase in peptide length ( Fig 4B). HPV16L1-α4-14H6a, and-b formed VLPs with a mean diameter of 50 nm, whereas HPV16L1-α4-14H6c VLPs were notably smaller with an approximate mean diameter of 30 nm. Large irregular aggregates of pentameric capsomeres were observed for HPV16L1-α4-14H6d, -e, and-f, all of which failed to form uniform VLPs ( Fig 4B). On the other hand, the morphologies of the six HBc149-14H6a-f chimeras were similar to that of wild-type, with healthy globular particles of a mean diameter of 30 nm ( Fig 4B). Therefore, the HPV16-L1-α4 loop, the HPV16-L1-DE-loop and MIR of HBc149 can harbor at least 20 aa while still maintaining a self-assembling capacity.
Finally, the 14H6-specific antigenicities of the purified chimeras were evaluated by direct ELISA. In general, the fusion proteins and VLPs exhibited similar EC50s to those of the positive controls, with the exception of the HPV16L1-α4-14H6a-f VLPs ( Fig 4C). The reactivity of HPV16L1-α4-14H6a-f VLPs, most significantly, VLPs of HPV16L1-α4-14H6c, were notably reduced, as indicated by the increased EC50; this was perhaps due to the structural instability of the recombinant complexes as perceived in the TEM.
Immunogenicity of 14H6a-f Chimera BALB/c mice were vaccinated three times at two-week intervals with 5 μg of the 14H6a-f recombinants absorbed with aluminum hydroxide adjuvant to confirm an anti-HPV16-L2 IgG response. The antibodies were then titered via ELISA. The anti-HPV16-L2 IgG was detectable as early as two weeks after the first immunization (data not shown). By the end of week 8, IgG responses were at maximum levels. With the exception of CRM197 and CRM389 recombinants, at least one of the 14H6 recombinants for each vector elicited a 10 5 IgG response (Fig 5).
To determine and compare the immunogenicity and broad neutralization activities of each recombinant 14H6a-f protein and VLP against HPV types 6, 11, 16, 18, 31, 45, 52, 58 and 59, Total L2-specific IgG titer in mice elicited by 14H6a-f chimeras. Chimeric purified proteins of 14H6a-f and scaffolding proteins were inoculated into mice to investigate the specific immunogenicity of 14H6's epitope. Prototype scaffold proteins and HPV16 L2 served as negative and positive controls, respectively. The antibody level was detected by HPV16 L2-coating ELISA. Significantly, all of the chimeras induced L2-specific responses that were comparable with the HPV16 L2 control. In the case of the CRM chimeras, the A domain fused with 14H6c-f, thereby eliciting a markedly higher antibody titer, similar to that of the other scaffolds (HPV VLP or HBc antigen).
doi:10.1371/journal.pone.0123944.g005 subsequent neutralization assays were performed using the antisera from week 8 (Fig 6). Significant cross-neutralizing activities were detected for the 14H6 peptide displaying HBc149 VLPs, HPV16-DE-L1 VLPs and CRMAs. Inoculation of HBc-149-14H6a-f VLPs prompted immunity against HPV16, 18, 45 and 52 PsVs. The neutralizing geometric mean titer (GMT) was exclusively high against HPV18 PsV at 10 3 , which was equivalent to that against HPV16 PsV (Fig 6C). A comparative study demonstrated stable and robust neutralizing activities of the HPV16-DE-14H6a-f VLP-induced antibodies as compared with that of the HPV16-α4-14H6a-f VLPs, which is concordant with the results from the ELISA and TEM assays (Fig 6A  and 6B). Moreover, the E.coli-derived wild-type HPV16 L1 protein and its 14H6 L2-chimeras manifest their excellent L1-specific immunogenicity as being comparable with HPV16/18/6/11 quadrivalent vaccine (Gardasil). However, the HPV16 L2-specific titer is 2-log lower otherwise broader-genotype than L1-specific (see HPV16 Ã assay of chimeras vs. HPV16 assay of HPV16 L1 VLP), which makes the difference of neutralization titers between L1 alone and L1+L2 undiscernible ( Fig 6A). Interestingly, we observed a difference in the titers of the crossneutralizing antibodies produced from CRM197/389/A-14H6a-f (Fig 6D, 6E and 6F): CRMA provoked a greater and broader cross-type neutralizing reaction compared with CRM197 and CRM389, implying that the catalytic domain A of DT alone was superior for presenting the heterologous L2 peptides. Taken together, we show that effective positioning of 14H6 peptides on various scaffold vectors, particularly CRMA, upon immunization elicits the production of broad, cross-neutralizing antibodies against at least HPV16, 11,18,45,52,58,59 PsVs. These

Discussion
In this study, we obtained an anti-HPV16-L2 neutralizing antibody, 14H6, that also crossneutralizes HPV6, 11, 16, 18, 31, 33, 35, 45, 52, 58, and 59 PsVs. It has hitherto one of the widest ranges of cross-neutralization and displays a remarkable neutralizing GMT for each targeted HPV ( Table 2). The 14H6 epitope was then mapped through the use of overlapping synthetic peptides, and found to recognize a linear stretch at L2 aa 21 to 30 near its N-terminus. Previous studies have revealed that N-terminal regions of L2 confer multiple, key, functional roles that are essential for virus attachment and viral genome assembly. These regions include DNA binding domains (aa 1-12) [34], cyclophilin B and β-actin binding domains (aa 25-45) [35], and a furin cleavage site (aa 9-12) [18]. During virus adhesion, furin proteases on the cell surface cleave the L2 N-terminus to expose aa 13 to 31 [36]. It was proposed that, since this exposed sequence triggers a critical interaction with a putative L2 receptor to allow virus entry, it must be highly conserved.
Several HPV16-L2-based cross-neutralizing antibodies have been identified, and these epitopes either envelope or overlap with that of 14H6 (  [20,21]. Whereas mAb RG-1 displayed a neutralization titer of 1280 against HPV16 and 18 PsVs, K4L220-38 had a neutralizing titer ranging from 625 to 3125 against HPV16, 18, 45, 58, 57, and 27 PsVs. Comparatively, our mAb 14H6 appears to exhibit a much higher and wider cross-neutralizing activity, even though it shares a similar neutralization profile with these previously described mAbs; albeit, the neutralization against HPV31 and HPV6 are lower. Immunization of L2 oligopeptides have previously shown to induce relatively low neutralizing antibody titers [37]; this is despite efforts to modify the aa sequence through the use of tandem repeats of an L2 peptide on bacterial thioredoxin or concatenating L2 peptides of different papillomavirus types, among others (Table 2). Immunologically, the unique display of certain peptides, such as in the form of VLPs, intra-molecular adjuvants, and as oligomeric proteins, have been able to enhance the immunogenicity of the intended epitope polypeptides; in VLPs or oligomers, the immunogens can carry multivalent peptides to cover variants among a target sequence. For more than 20 years, HPV VLPs and HBc149 VLPs have been frequently used as epitope-display particle vectors [38,39]. In our laboratory, we have frequently used insertions of diverse short peptides into one of the two constrained surface-exposed loops-the DE-loop (130/131) or the α4 loop (426/427)-on the HPV16-L1 protein, and found that such insertions do not affect the L1's ability to self-assemble into immunogenic VLPs (data not shown). A Cterminal truncations of HBc149 has also been widely used as immune-enhancing carrier proteins [40]. Upon the formation of VLPs, an abundance of peptide inserts are presented on the surface as foreign antigens [41].
Our data and previous studies are in agreement that the DE loop of HPV L1 is better at peptide display than the α4 loop, since it is able to accommodate inserts with more than 30 aa while maintaining high immunogenicity [14,33]. This assertion was supported by findings from atomic modeling of the bovine papillomavirus (BPV), which showed that, although the α4 loop is exposed on the surface of VLPs, it also serves an important function in VLP assembly. Upon insertion of a peptide with more than 16 aa, the α4 loop will distort, and this may lead to assembly failure of the neighboring pentameric capsomeres; indeed, discrete aggregates and pentamers were observed in this study under TEM. Our study has also demonstrated that insertion of 14H6a-f into the MIR of HBc149 did not affect the assembly of the HBcore VLPs, and was able to elicit a remarkable cross-neutralizing response against HPV16, 18, 45, and 52. In addition, the 14H6 CRMA chimeras exhibited a wider spectrum and an equivalent crossneutralizing potency of six HPV types (HPV11, 16, 18, 45, 52 and 59) as compared with the 14H6 HPV16-L1-DE and HBcore VLP chimeras. Furthermore, fusion of CRMA with the 14H6 epitope was more effective in terms of peptide display than CRM197 and CRM389, suggesting that the catalytic domain A of CRM197 may favor peptide folding and be potent in immunogenic enhancement. It should be mentioned here that this is the first time CRMA has been exploited and shown to be a more effective display vector than the other moieties and intact CRM197. In our immunogenicity analysis using multiple HPV types for neutralization, none of the 14H6 chimeras could induce marked cross-neutralization titers in anti-14H6 sera, despite CRMA-14H6 showing neutralization for six HPV types (Fig 6). However, in the neutralization assay, mAb 14H6 was shown to exhibit high and wide-genotype neutralizing capacity (Fig 1C). This discrepancy reflects the persistent challenge associated with evoking a host immune response in an immunodominant manner against a linear epitope. During vaccination, different scaffold proteins may contribute to changes in peptide folding during recombinant expression and subsequent immune presentation. In terms of sequence conservation for the 14H6 epitope, there is still variation among the HPV genotypes. In a word, our findings suggest an L2-based pan-HPV vaccine should cover extensive L2-epitope chimeric constructions to elicit broader HPV type protection coverage, at least including scaffold protein fusion design, peptide length variation and HPV L2 sequence diversity. In future experiments, the 14H6 epitope peptide should be instead designed using key residues from a common motif, TCKxxxTCPP, in order to cover more of the HPV genotypes. This modified peptide should then be grafted to CRMA and/or HPV VLPs to confer considerable immunogenicity. This would indeed help in the generation of an L2-based, pan-HPV vaccine, with future L2-based vaccine clinical trials revealing the neutralization mechanism of L2 relative to that of the known L1.

Ethics Statement
In this study, the experimental animals (BALB/c mice) were purchased from Shanghai Institutes for Biological Sciences (Shanghai, China), and fed in the animal facility of Xiamen University Laboratory Animal Center (XMULAC). The manipulation and vaccination on the animals were strictly referred to the guideline and compliant with the regulation, which was provided by XMULAC. Prior to the implementation, the experiment schemes and protocols were reviewed by Xiamen University Institutional Committee for the Care and Use of Laboratory Animals and approved by Xiamen University Laboratory Animal Management Ethics Committee. During the experiments, all animals were well-fed and monitored twice per day. The manipulation, including vaccination, blood collection and surgery, was handled by welltrained professional in terms of the essential principle of gentle and painless. Finally, experimental mice were injected intraperitoneally with nembutal sodium for euthanasia.

Murine monoclonal antibodies (mAbs)
BALB/c mice were immunized subcutaneously three times at an interval of two weeks with HPV16C50L2 (20 μg/animal) absorbed with Freund's Complete Adjuvant. The anti-HPV16L2 murine mAbs 14H6, 15E5, 15E4, and 6H8 were then produced using the hybridoma technology previously described [45]. Protein A affinity chromatography was employed to purify the anti-HPV mAbs IgGs and the purified mAbs were subsequently diluted to 1.0 mg/ml in PBS and stored at −20°C.

Homology modeling
Using protein BLAST search (NCBI, www.ncbi.nlm.nih.gov/blast/), three crystal structures extracted from Protein Data Bank (PDB no. 1DZL, 1QGT and 4AEO matching HPV16L1, HBc149 and CRM, respectively) were selected as templates for homology modeling of the corresponding monomers. After sequence alignment with the respective 14H6 chimeric sequences, initial 3D models were generated using the Homology module of Discovery Studio 2.5 program (Accelrys). BPV (PDB no.3IYJ) and HBc149 (PDB no. 1QGT) VLP templates were used to generate the complete HPV16L1-14H6 and HBc149-14H6 VLP models. Stepwise minimizations were subsequently applied to the 14H6 insertion site of each model to produce thermodynamically favored conformations.

Expression and purification of 14H6 chimeras
The genes encoding HPV16 L1-14H6, HBc149-14H6 and CRM-14H6 chimeras were cloned into pTO-T7 vector [42], and transformed to E. coli ER2566 for chimeric expression. The transformated bacteria were cultured in LB medium at 37°C overnight and the protein expression was initiated by the addition of isopropyl-β-D-thiogalactoside (IPTG, final concentration of 10μM) and carried out with further incubation at 25°C for 6 h. Cells were harvested by centrifugation and resuspended with cell lysis solution (20 mM Tris, pH7.4, 300 mM NaCl, and 5 mM EDTA). The 14H6 chimeras were released from the cells by sonication and separated from the cell debris by centrifugation. As for HPV16 L1-14H6 and CRM -14H6 chimeras, the interest proteins in lysate supernatants were precipitated by ammonium sulfate of 25% saturation while incubating at 4°C for 2h. The precipitates were resuspended in 20mM PB pH 8.0, 500mM NaCl. Then, HPV16 L1-14H6 proteins with the dilution of equal volume 20mM PB, pH 8.0, 40mM dithiothreitol (DTT), were further purified by cation ion exchange chromatography (SP Sepharose 6FF resin, GE Healthcare, Uppsala, Sweden) and eluted in 800 mM NaCl fraction; and CRM -14H6 proteins with final NaCl concentration of 2.0 M were further purified by hydrophobic interaction chromatography (Phenyl Sepharose 6FF resin, GE Healthcare) and eluted with 200mM NaCl. For HBc149-14H6 chimeras, the lysate supernatant were treated with heating at 56°C for 30min to eliminate most contaminants, then the resultant supernatant was precipitated by 40% saturated ammonium sulfate. Precipitates were resuspended with PBS (pH 7.4) and further purified with a DEAE-5PW column (TOSOH Bioscience, Tokyo, JP). Finally, the purified HPV16 L1-14H6 chimeras self-assembled into virus-like particles with the removal of reductant DTT.

SDS-PAGE and western blotting
The 14H6 chimeras were analyzed by SDS-PAGE using methods outlined by Laemmli with slight modifications. Protein samples were mixed with equal volumes of 6× loading buffer (composed of 100 mM Tris-HCl, pH 6.8, 200 mM BME, 4% SDS, 0.2% Bromophenol blue and 20% glycerol). Sample mixtures were heated at 80°C for 10 min and subsequently loaded onto 10% acrylamide gels. For western blotting analyses, separated samples of the 14H6 chimeras were transferred from the SDS gels to nitrocellulose membranes. The membranes were blocked with 5% skim milk, soaked in 1:500 diluted HPV16 L2-specific mice sera, and incubated at room temperature for 1 h. Subsequent washing was performed using 0.2% Tween-20 in phosphate-buffered saline, pH 7.4. Alkaline phosphatase-conjugated secondary antibody (Dako, Denmark) was then added to capture the bound primary antibody. A mixture of nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate was then added to allow color development and visualization of the target protein bands.

Transmission electron microscopy (TEM)
Approximately 15 μl of negatively stained HPV16L1-and HBc149-vectored 14H6 chimeric VLPs at 200 μg/ml were absorbed onto carbon-coated copper grids, blotted dry, and stained with freshly filtered 2% phosphotungstic acid (pH 6.4). Grids were examined under the FEI Tecnai T12 TEM at an accelerating voltage of 120 kV and then photographed at a nominal 25,000× magnification.

Direct binding ELISA
14H6 chimeric proteins (300 ng/well) were coated into the wells of 96-well microplates. After plate blocking, 100 μl of 2-fold serially diluted anti-HPV16L2 mAbs 14H6 (initiated from 1 μg/ml in PBS) were added to the wells. HRP-conjugated goat anti-mouse Ig antibody (diluted 1:5000 in HS-PBS, Abcam; Cambridge, UK) was used as a secondary antibody. The absorbance at 450 nm with reference to 620 nm was record using an automated ELISA reader (TECAN, Männedorf, Switzerland). The cut-off value was set as absorbance of 450 (Δ620)nm = 0.2 to define the positive titer.

Animal immunization
Four-to-six weeks old female BALB/c mice were divided into forty-two groups (n = 4 in each group) and each group of mice was vaccinated subcutaneously three times, each separated by an interval of two weeks. 5 μg of immunogens, (14H6 chimeric proteins and negative controls including HPV16C50L2, HPV16 VLP, HBc149, CRM197/389/A) absorbed with aluminum hydroxide adjuvant and 1/8 human dose of HPV16/18/6/11 quadrivalent vaccine (Gardasil) were injected subcutaneously on day 0, 14 and 28. Serum samples were collected individually from each mouse at the 14th day after each injection. Serum samples on the fifty-sixth day were used to analyze the neutralizing titers by following neutralization assay. The neutralization titers of 4 mice per group were plotted in averaged value with standard error vs. various types of applied HPV pseudoviron.

Detection of sera neutralizing antibodies
293FT cells were incubated at 37°C in the wells of a 96-well plate at a density of 1.5×10 4 cells per well for 6 h. Sera were subjected to a 2-fold dilution. PsVs were diluted to 2×10 5 TCID50/ μl. Sixty μl of the PsV diluent and 60 μl of the serially diluted sera were mixed and incubated at 4°C for 1 h. The negative control was prepared by mixing 60 μl of the PsV diluent with 60 μl of the culture medium. Then, 100 μl of the above mixtures were added designated wells and incubated at 37°C for 72 h. Cells were then treated with trypsin and analyzed by flow cytometry. The endpoint titers were calculated as the log 10 of the highest sera dilution with a percent infection inhibition higher than 50%. Every sample was run at least three times, and the values presented here are calculated as the mean value of all repeats.