Highly Attenuated Bordetella pertussis Strain BPZE1 as a Potential Live Vehicle for Delivery of Heterologous Vaccine Candidates

Bacterial strains and growth conditions.The bacterial strains used in this study are listed in Table 1. BPSY13.1, BPSY1, and BPSQ5 were derived from B. pertussis BPZE1, a streptomycin-resistant Tohama I derivative producing inactivated pertussis toxin, no dermonecrotic toxin, and background levels of tracheal cytotoxin (40). All B. pertussis strains were grown at 37°C for 72 h on Bordet-Gengou (BG) agar (Difco, Detroit, MI) supplemented with 1% glycerol, 10% defibrinated sheep blood, and 100 μg/ml streptomycin (Sigma Chemical Co., St Louis, MO). Liquid cultures were grown as described previously (37) in Stainer-Scholte medium containing 1 g/liter heptakis(2,6-di-o-methyl)-β-cyclodextrin (Sigma).

All DNA manipulations were carried out with chemically competent Escherichia coli One-Shot TOP10 (Invitrogen). The bacteria were grown at 37°C overnight on Luria-Bertani agar or in Luria-Bertani broth with vigorous shaking. When appropriate, 100 μg/ml ampicillin, 50 μg/ml ampicillin, or 10 μg/ml gentamicin was added to select for antibiotic-resistant strains.

Virus growth and purification.EV71 strain 5865/SIN/00009 (GenBank accession no. AF316321) was propagated in rhabdomyosarcoma (RD) cells using minimum essential medium (Gibco, United States) supplemented with 5% fetal calf serum, 1% sodium pyruvate, and 1.5% sodium bicarbonate. The virus particles were purified as described previously (23). Briefly, infected cells were lysed by subjecting them to freeze-thaw cycles. The virus particles were precipitated in 7% polyethylene glycol 8000 by centrifugation on a 30% sucrose cushion at 25,000 × g for 4 h. The virus titer was expressed as the 50% tissue culture infective dose with RD cells based on typical cytopathic effects (CPE) produced by viral infection. Before injection into mice, the virus suspension was heat inactivated at 56°C for 30 min. The amount of virion protein was quantified by the Bradford method (Bio-Rad Laboratories, United States).

Oligonucleotides, peptides, and antibodies.To circumvent any problems in protein translation due to poor codon usage (28), the original sp70 DNA sequence was optimized to B. pertussis codon usage preference. To generate the FHA-(SP70)₃ construct, the upper and lower DNA strands of optimized sp70 (5′-GATCGGCTACCCGACCTTCGGCGAGCACAAGCAGGAGAAGGACCTGGAGTACGA-3′ and 5′-GATCTCGTACTCCAGGTCCTTCTCCTGCTTGTGCTCGCCGAAGGTCGGGTAGCC-3′) were chemically synthesized and annealed, generating cohesive BglII-compatible ends. To generate the BrkA-(SP70)₃ construct, the upper and lower DNA strands of optimized sp70 (5′-GATCTGTACCCGACCTTCGGCGAGCACAAGCAGGAGAAGGACCTGGAGTACTG-3′ and 5′-GATCCAGTACTCCAGGTCCTTCTCCTGCTTGTGCTCGCCGAAGGTCGGGTACA-3′) were chemically synthesized and annealed, generating cohesive BamHI-compatible ends.

Unconjugated SP70 peptide (22) was chemically synthesized at Mimotopes Pty. Ltd. (Clayton Victoria, Australia).

Rabbit anti-BrkA polyclonal antibodies were a kind gift from Rachel Fernandez (University of British Columbia, Canada).

Construction of recombinant B. pertussis strains.To construct the recombinant B. pertussis BPSY13.1 strain producing the FHA-(SP70)₃ chimera, a 1,620-bp HindIII PCR fragment was amplified from the BPZE1 chromosomal DNA using primers 5′-TTAAGCTTGCGAACGCGCTGCTGTGGG-3′ and 5′-TTAAGCTTCGCATCGGCGCTGCCCAGC-3′ (HindIII sites are underlined) and cloned into HindIII-opened plasmid pBR322 (7), yielding pBRSY0. The PCR fragment corresponded to nucleotides (nt) 5221 to 6840 of the fhaB open reading frame (ORF) and contained its unique BglII site. Three copies of the sp70 DNA sequence were inserted in tandem and sequentially into the unique BglII site of pBRSY0. Insertion of one copy of sp70 DNA into BglII-digested pBRSY0 restored a BglII site only at the 3′ end of the sp70 DNA fragment, allowing insertion of a second sp70 copy and then a third sp70 copy, finally yielding pBRSY3. The 1,755-bp HindIII fragment from pBRSY3 was then cloned into HindIII-opened suicide plasmid pJQmp200rpsL18 (48), yielding pJQSY3. BPZE1 was electroporated with pJQSY3, allowing the fha-(sp70)₃ construct to integrate into the chromosomal DNA by allelic exchange (56) at the fhaB locus.

To construct the recombinant B. pertussis BPSQ5 strain, which express the BrkA-(SP70)₃ chimera, a 1-kb SalI-BamHI PCR fragment and a 945-bp BamHI-HindIII PCR fragment were cloned into pUC19 (60) using the corresponding restriction sites, yielding pUCSY2. Both PCR fragments were amplified from BPZE1 chromosomal DNA. The 1-kb SalI-BamHI PCR fragment was amplified using primers 5′-TTGTCGACGTAGTATCCCTTGGCCGCGC-3′ and 5′-TTGGATCCTGCGCATGCGGCGCGCC-3′ (SalI and BamHI sites are underlined) and encompassed the 5′ end of the brkA ORF (nt 1 to 151), its promoter region, and the first 529 nt of the adjacent brkB ORF, which is transcribed in the opposite direction. A 945-bp BamHI-HindIII PCR fragment was obtained using primers 5′-TTGGATCCACGCCGGCCAGGACGGCAA-3′ and 5′-TTAAGCTTCACGACCCAGGTTCCGCCC-3′ (BamHI and HindIII sites are underlined) and corresponded to nt 789 to 1735 of the brkA ORF. Three copies of the sp70 gene fragment were then inserted in tandem and sequentially into BamHI-digested pUCSY2. Insertion of one copy of sp70 DNA into BamHI-digested pUCSY2 restored a BamHI site only at the 3′ end of the sp70 DNA fragment, which allowed insertion of a second copy and a third copy of sp70, finally yielding pUCSQ2. The 2,125-bp HindIII fragment from pUCSQ2 was finally cloned into HindIII-opened pJQmp200rpsL18, yielding pJQSQ1. BPZE1 was electroporated with pJQSQ1, which allowed the brkA-(sp70)₃ construct to integrate into the chromosomal DNA by allelic exchange at the brkA locus.

To construct the brkA knockout strain BPSY1, the 1,963-bp HindIII fragment from pUCSY2, which contained both the 1-kb SalI-BamHI and 945-bp BamHI-HindIII PCR fragments described above, was inserted into HindIII-opened pJQmp200rpsL18, yielding pJQSY1. BPZE1 was then electroporated with pJQSY1, which allowed allelic exchange at the brkA chromosomal locus and resulted in deletion of nt 151 to 789 in the brkA ORF. BPSY1 therefore produced a truncated BrkA protein consisting of 103 amino acids.

Whole-cell extract preparation and supernatant concentration.Ten milliliters of mid- to late-exponential-phase bacteria in SSAB medium was centrifuged at 7,000 rpm for 15 min at room temperature. The supernatant was concentrated 10-fold using a 30-kDa-cutoff Ultra-4 centrifugal filter device (Amicon) according to the manufacturer's protocol. The bacterial pellet was resuspended in 500 μl of ultrapure water. An equal volume of 2× loading buffer was added before the preparation was heated at 95°C for 10 min. The chromosomal DNA was sheared by passing the suspension 10 times through a 27-gauge needle; this was followed by heating at 95°C for 15 min before 30 μl was loaded onto a sodium dodecyl sulfate-polyacrylamide gel electrophoresis gel for Western blot analysis.

Immunodetection of FHA-(SP70)₃ and BrkA-(SP70)₃ chimeras.Concentrated (10×) culture supernatants or whole-cell extracts of the B. pertussis strains were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis using 8 or 12% polyacrylamide gels. The proteins were electrotransferred onto nitrocellulose membranes and incubated with mouse anti-SP70 polyclonal antibodies (23) diluted 1:100, mouse anti-FHA monoclonal antibodies diluted 1:250 (49), or rabbit anti-BrkA polyclonal antibodies diluted 1:30,000 (45) in Tris-buffered saline containing 0.1% Tween 20 and 2% bovine serum albumin. Alkaline phosphatase-conjugated goat anti-mouse or anti-rabbit immunoglobulin G (IgG) secondary antibodies (Sigma), both diluted 1:3,000, were used for chromogenic detection of the proteins after addition of the alkaline phosphatase substrate (nitroblue tetrazolium and 5-bromo-4-chloro-3-indolylphosphate reagents; Sigma).

FACS.B. pertussis strains grown on BG agar were washed three times with sterile phosphate-buffered saline (PBS) supplemented with 5% glycerol. Fluorescence-activated cell sorting (FACS) was then conducted with the intact B. pertussis cells using a Coulter Epics machine (Beckman Coulter, Palo Alto, CA). Intact bacteria were incubated with rabbit anti-BrkA polyclonal antibodies (45) diluted 1:200 and then with Cy2-conjugated goat anti-rabbit IgG (Jackson Laboratories, West Grove, PA) diluted 1:100. Samples were analyzed with laser excitation at 488 nm, and the data were acquired using the EXPO version 2.0 software (Applied Cytometry Systems, Sheffield, United Kingdom) and analyzed with the WinMDI-2.8 software. The assay was performed two times independently.

Immunofluorescence.Intact B. pertussis cells were prepared as described above for FACS and spotted onto glass slides pretreated with 100 μl of 0.1% poly-l-lysine. The samples were examined with blue light excitation (488 nm) using an epifluorescence microscope (BX40; Olympus, Japan) at a magnification of ×1,000.

i.n. infection.The mice were kept under specific-pathogen-free conditions in individual ventilated cages, and all the experiments were carried out using the guidelines of the National University of Singapore animal study board. For colonization studies, 9-week-old outbred CD1 mice (Biopolis Research Center, Singapore) were each infected intranasally (i.n.) with 5 × 10⁶ CFU of the different B. pertussis strains in 20 μl as described previously (1). At the indicated time points, four mice per group were sacrificed, and their lungs were aseptically removed and homogenized in PBS. Serial dilutions from individual lung homogenates were plated onto BG agar, and the total numbers of CFU per lung were determined after 4 to 5 days of incubation at 37°C. For immunization studies, groups of six 5-week-old BALB/c mice (Biopolis Research Center, Singapore) were infected i.n. twice at a 4-week interval with 5 × 10⁶ CFU of the different B. pertussis strains in 20 μl. An additional group of six mice was inoculated intraperitoneally (i.p.) twice at a 4-week interval with 10 μg of heat-inactivated EV71 in a 50% emulsion of complete and incomplete Freund's adjuvant. At the indicated time points, the mice were bled at the retroorbital sinus.

Antibody detection.The levels of antibodies to SP70 and B. pertussis were measured by an enzyme-linked immunosorbent assay (ELISA). The 96-well microtiter plates (COSTAR; Corning) were coated overnight at 4°C with 50 μl of 0.1 M carbonate buffer (pH 9.6) containing 10 μg/ml of unconjugated SP70 peptide or total B. pertussis BPZE1 cell lysate. After blocking with 2% bovine serum albumin in PBS containing 0.1% Tween 20, 50 μl of serum diluted 1:50 (for anti-SP70 detection) or 1:800 (for anti-B. pertussis detection) was added to the wells. The plates were incubated at 37°C for 1 h, rinsed in PBS-0.1% Tween 20, and incubated at 37°C for 1 h with 50 μl of horseradish peroxidase-conjugated goat anti-mouse IgG(H+L) secondary antibodies (Sigma) at a 1:3,000 dilution. To detect the various IgG subtypes, horseradish peroxidase-conjugated goat anti-mouse IgG1, IgG2a, IgG2b, and IgG3 secondary antibodies (Jackson Laboratories) were used at a 1:5,000 dilution. The reaction was then developed using o-phenylenediamine dihydrochloride substrate (Sigma) at room temperature for 30 min in the dark and stopped by addition of 1 M sulfuric acid. The absorbance at 490 nm was determined with an ELISA plate reader (Tecan Sunrise, United States).

EV71 neutralization assay.The presence of neutralizing antibodies against EV71 was determined by an in vitro microneutralization assay using RD cells, as described previously (23). Mouse serum samples were first incubated at 56°C for 30 min to inactivate complement activity. Briefly, 25 μl of twofold serial dilutions of heat-treated serum was coincubated with equal volumes containing 10³ 50% tissue culture infective doses of virus in a 96-well microtiter plate. Two hours later, 5 × 10⁴ RD cells were added to each well and incubated at 37°C for 48 h. The cells were examined for CPE, and the neutralizing antibody titer was defined as the highest dilution of serum that inhibited virus growth by 100%, thereby preventing CPE. The assay was performed three times independently.

Statistical analysis.The results were analyzed using the unpaired Student t test. Differences were considered significant if the P value was <0.05.

Production of FHA-(SP70)₃ and BrkA-(SP70)₃ chimeras by B. pertussis.Up to 85% of the passenger domain of BrkA can be deleted without affecting the efficacy of translocation of the protein across the outer membrane (45). However, the passenger domain of BrkA has recently been found to possess adjuvant properties and immunogenic activities (9), which may be important when BrkA is used as a carrier for the display of vaccine candidates. We therefore decided to truncate the BrkA protein from amino acid A52 to H263, corresponding to a deletion of 32% of the passenger domain. Three copies of the 15-amino-acid SP70 neutralizing peptide from EV71 were then fused in tandem in the truncated passenger domain of BrkA. Three copies of SP70 were also inserted in tandem into full-length FHA. The chimeric proteins were designated BrkA-(SP70)₃ and FHA-(SP70)₃, respectively. The corresponding DNA constructs were introduced by allelic exchange into the brkA and fhaB chromosomal loci, respectively, of attenuated B. pertussis BPZE1, resulting in strains BPSQ5 and BPSY13.1, respectively.

The production of the FHA-(SP70)₃ and BrkA-(SP70)₃ chimeras by the recombinant strains was analyzed by immunoblotting using anti-SP70 and anti-FHA or anti-BrkA antibodies. A 225-kDa band corresponding to the predicted size of FHA-(SP70)₃ was detected in the culture supernatant of BPSY13.1 using anti-FHA and anti-SP70 antibodies (Fig. 1A and B, respectively). Two bands at 103 and 73 kDa, corresponding to full-length wild-type BrkA and its passenger domain, respectively, were detected in the whole-cell extract of BPZE1 using anti-BrkA antibodies (Fig. 1C). Similarly, two bands at 85 and 55 kDa, corresponding to the predicted sizes of full-length BrkA-(SP70)₃ and its passenger domain, respectively, were detected in the whole-cell extract of BPSQ5 using anti-BrkA and anti-SP70 antibodies (Fig. 1C and D, respectively). No bands were detected in the brkA knockout BPSY1 strain using either antibody (Fig. 1C and D). FHA and BrkA were not detected by anti-SP70 antibodies (Fig. 1B and D, respectively).

• Open in new tab
• Download powerpoint

FIG. 1.

Detection of FHA-(SP70)₃ and BrkA-(SP70)₃ chimeras by immunoblotting. Tenfold-concentrated culture supernatants from BPZE1 and BPSY13.1 cultures (A and B) or whole-cell extracts from BPZE1, BPSQ5, and BPSY1 cultures (C and D) were assayed by immunoblotting using anti-SP70 polyclonal antibodies (B and D), anti-FHA monoclonal antibodies (A), and anti-BrkA polyclonal antibodies (C). Fifty microliters of supernatant or 10 μl of cell extract was loaded.

These data demonstrate that FHA-(SP70)₃ and BrkA-(SP70)₃ were successfully produced by BPSY13.1 and BPSQ5, respectively. Moreover, they indicate that FHA-(SP70)₃ was efficiently secreted by BPSY13.1 into the extracellular milieu. In contrast and as expected, BrkA-(SP70)₃ was not secreted by BPSQ5 at an appreciable level (data not shown).

Cell surface exposure of the BrkA-(SP70)₃ chimera.To assess whether BrkA-(SP70)₃ was exposed at the bacterial surface of BPSQ5, FACS was performed with intact (nonpermeabilized) BPSQ5 cells using anti-BrkA antibodies. The parental BPZE1 and brkA knockout BPSY1 strains were used as positive and negative controls, respectively. As shown in Fig. 2, 100% of the parental BPZE1 cells exhibited surface exposure of BrkA. Similarly, the majority of BPSQ5 cells were found to be positive, and the difference from the parental strain was not statistically significant (P = 0.21) (Fig. 2B). As expected, the BPSY1 strain did not display any significant fluorescence.

• Open in new tab
• Download powerpoint

FIG. 2.

Detection of the BrkA-(SP70)₃ chimera by FACS. Anti-BrkA polyclonal antibodies were coincubated with intact BPSY1, BPZE1, and BPSQ5 cells as indicated. The isotype control was BPZE1 bacteria stained with Cy2-conjugated secondary antibody. (A) Graphs representative of two independent experiments. (B) Average values for the two independent experiments. The results are expressed as means ± standard deviations.

The surface exposure of BrkA-(SP70)₃ was further confirmed by immunofluorescence analysis using anti-BrkA antibodies on intact BPSQ5, BPZE1, and BPSY1 cells. As shown in Fig. 3, BPZE1 and BPSQ5 cells displayed strong and comparable fluorescence signals (Fig. 3G and H, respectively), while no significant fluorescence emission was detected with the BPSY1 strain (Fig. 3F).

• Open in new tab
• Download powerpoint

FIG. 3.

Detection of the BrkA-(SP70)₃ chimera by immunofluorescence microscopy. Anti-BrkA polyclonal antibodies were coincubated with intact BPSY1 (B and F), BPZE1 (C and G), or BPSQ5 (D and H) cells. The isotype control (A and E) was BPZE1 bacteria stained with Cy2-conjugated secondary antibody. Panels A to D show corresponding phase-contrast images.

Due to a high background value, the anti-SP70 immune serum could not be used as the primary antibody in FACS and immunofluorescence studies to confirm the data obtained with the anti-BrkA immune serum (data not shown).

These results demonstrate that BrkA-(SP70)₃ is exposed at the bacterial surface of BPSQ5 at levels comparable to the levels of the wild-type BrkA protein in BPZE1.

Lung colonization by B. pertussis BPSY13.1 and BPSQ5.FHA and BrkA have been shown to play a role in the colonization efficiency of B. pertussis (1, 18). To study whether the recombinant BPSQ5 and BPSY13.1 strains retained the capacity to colonize the murine respiratory tract, mice were infected i.n. with either strain, and the colonization profiles were compared to those of the parental BPZE1 and brkA knockout BPSY1 strains.

BPSY13.1 colonized the lungs as efficiently as the parental BPZE1 strain; a peak of multiplication was observed, followed by progressive clearance of the bacteria from the lungs (Fig. 4A), indicating that the insertion of three copies of SP70 into full-length FHA did not impair the adhesion function of the protein. However, similar to the colonization efficiency of the brkA knockout strain BPSY1, the colonization efficiency of BPSQ5 was found to be slightly but significantly (P < 0.05) reduced 7 and 10 days postinfection compared to that of BPZE1 (Fig. 4B). This observation suggests that the BrkA-(SP70)₃ chimera did not retain the full adhesion function of the wild-type BrkA protein and/or that other functions of BrkA, such as resistance to serum killing (20) and to antimicrobial peptides (21), were impaired in the BrkA-(SP70)₃ chimera, which might account for the reduced colonization ability observed with BPSQ5.

• Open in new tab
• Download powerpoint

FIG. 4.

Lung colonization by the recombinant B. pertussis strains. (A) CD1 mice were infected i.n. with 5 × 10⁶ CFU of B. pertussis BPZE1 (solid squares in panels A and B), BPSY13.1 (open squares in panel A), BPSY1 (solid triangles in panel B), or BPSQ5 (open squares in panel B). The lungs of infected mice were harvested at the indicated time points, and appropriate dilutions of the lung homogenates were plated to determine the total number of CFU per lung. Four mice per group and per time point were assessed individually. Asterisk, P < 0.05 compared to BPZE1.

Systemic anti-SP70 and anti-B. pertussis antibody responses in mice.To examine the abilities of the two recombinant B. pertussis strains to trigger a systemic anti-SP70 antibody response upon nasal administration, groups of six BALB/c mice were infected i.n. twice at a 4-week interval with BPSY13.1, BPSQ5, or BPZE1. As a reference for anti-SP70 antibody production, an additional group of mice was inoculated i.p. with heat-inactivated EV71 using the same immunization schedule. The systemic anti-SP70 and anti-B. pertussis IgG responses were measured by ELISA 2 weeks after the boost.

Both BPSY13.1- and BPSQ5-infected mice developed a strong systemic anti-B. pertussis antibody response comparable to the response observed in the BPZE1-infected mice (Fig. 5A). As expected, no anti-B. pertussis antibody response was seen in the EV71-inoculated animals. However, the EV71-inoculated mice all showed high anti-SP70 antibody responses, while the naïve and BPZE1-infected mice displayed only background absorbance (Fig. 5B). Two of six BPSY13.1-infected mice (mice M5 and M6) produced significant anti-SP70 IgG antibody levels. In contrast, five of six BPSQ5-immunized mice produced significant anti-SP70 antibody levels, and the titers were significantly higher than the titers obtained for the BPSY13.1-immunized group (P < 0.05). However, the antibody titers measured for both groups of mice were found to be significantly lower than the titers measured for the EV71-inoculated group.

• Open in new tab
• Download powerpoint

FIG. 5.

Detection of specific antibody responses. Groups of six mice were infected i.n. twice at a 4-week interval with 5 × 10⁶ CFU of BPZE1, BPSY13.1, or BPSQ5. The EV71 group was inoculated i.p. with 10 μg of inactivated virus using the same immunization schedule. The mice were bled 2 weeks after the boost, and the anti-B. pertussis (A) and anti-SP70 (B) IgG(H+L) titers were determined by ELISA with the individual sera diluted 1/800 and 1/50, respectively, using B. pertussis whole-cell lysate and SP70 peptide as coating antigens, respectively. ⋄, mouse M1; ▪, mouse M2; ▵, mouse M3; ×, mouse M4; ▴, mouse M5; □, mouse M6. The average is indicated by a horizontal line. OD_490nm, optical density at 490 nm.

An anti-SP70 IgG subtype analysis was carried out for the immune sera from all the mouse groups and showed that there was production of significant levels of IgG2a/IgG2b antibodies in the BPSQ5- and BPSY13.1-immunized mice, indicative of a Th1-oriented immune response (Table 2).

The anti-B. pertussis and anti-SP70 antibody responses were monitored over a period of 8 weeks after the boost in the BPSY13.1- and BPSQ5-immunized groups and were found to be as high as the titers measured 2 weeks after the boost (data not shown), demonstrating that the antibody responses triggered by nasal administration of the recombinant strains was sustained.

Neutralizing activity of the immune sera against EV71.To evaluate the functional activities of the sera from the BPSY13.1- and BPSQ5-immunized mice, an in vitro EV71 neutralization assay was used. Serially diluted sera were coincubated with EV71 before infection of RD cells. CPE were determined 48 h later. The sera from the naïve, BPZE1-infected, and EV71-inoculated mice were pooled within groups, while the sera from the BPSY13.1- and BPSQ5-infected mice were analyzed individually.

In contrast to uninfected cells, which had a flattened and spindle-like shape, infected cells appeared to be rounded and swollen with microbodies, as described elsewhere (23; data not shown). As expected, the sera from the naïve and BPZE1-infected mice failed to protect RD cells from viral infection (Table 3). In contrast, the pooled serum from the EV71-inoculated mice provided complete protection to the cells up to a serum dilution of 1:128. For the six BPSY13.1-infected mice, only the serum from mouse M5, corresponding to the highest anti-SP70 IgG titer, displayed significant neutralizing activity against the virus. This serum conferred complete protection to the cells up to a dilution of 1:16. The sera from the five BPSQ5-infected mice, which were found to produce significant anti-SP70 antibodies, showed the ability to neutralize the virus, and complete protection was obtained with serum dilutions ranging from 1:2 to 1:32. Surprisingly, the highest neutralization titer did not correspond to the highest anti-SP70 antibody titer. For example, the sera from mice M1 and M5, which contained significantly different anti-SP70 antibody levels (Fig. 5B), were found to be equally able to neutralize EV71 in vitro.

These results show that nasal administration of BPSQ5 and, to a lesser extent, nasal administration of BPSY13.1 are able to trigger the production of systemic antibodies capable of neutralizing EV71 infection in vitro.