Vaccine 32 (2014) 4296–4303
Contents lists available at ScienceDirect
Vaccine journal homepage: www.elsevier.com/locate/vaccine
A virus-like particle based bivalent vaccine confers dual protection against enterovirus 71 and coxsackievirus A16 infections in mice Zhiqiang Ku a , Qingwei Liu a , Xiaohua Ye a , Yicun Cai a , Xiaoli Wang a , Jinping Shi a , Dapeng Li a , Xia Jin a , Wenqi An b , Zhong Huang a,∗ a Vaccine Research Center, Key Laboratory of Molecular Virology & Immunology, Institut Pasteur of Shanghai, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai 200031, China b Hualan Biological Bacterin Company, Xinxiang 453003, Henan, China
a r t i c l e
i n f o
Article history: Received 26 February 2014 Received in revised form 22 May 2014 Accepted 6 June 2014 Available online 17 June 2014 Keywords: Enterovirus 71 Coxsackievirus A16 Virus-like particle Vaccine Neutralizing antibody
a b s t r a c t Enterovirus 71(EV71) and coxsackievirus A16 (CA16) are responsible for hand, foot and mouth disease which has been prevalent in Asia-Pacific regions, causing significant morbidity and mortality in young children. Co-circulation of and co-infection by both viruses underscores the importance and urgency of developing vaccines against both viruses simultaneously. Here we report the immunogenicity and protective efficacy of a bivalent combination vaccine comprised of EV71 and CA16 virus-like particles (VLPs). We show that monovalent EV71- or CA16-VLPs-elicited serum antibodies exhibited potent neutralization effect on the homotypic virus but little or no effect on the heterotypic one, whereas the antisera against the bivalent vaccine formulation were able to efficiently neutralize both EV71 and CA16, indicating there is no immunological interference between the two antigens with respect to their ability to induce virusspecific neutralizing antibodies. Passive immunization with monovalent VLP vaccines protected mice against a homotypic virus challenge but not heterotypic infection. Surprisingly, antibody-dependent enhancement (ADE) of disease was observed in mice passively transferred with mono-specific anti-CA16 VLP sera and subsequently challenged with EV71. In contrast, the bivalent VLP vaccine conferred full protection against lethal challenge by either EV71 or CA16, thus eliminating the potential of ADE. Taken together, our results demonstrate for the first time that the bivalent VLP approach represents a safe and efficacious vaccine strategy for both EV71 and CA16. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction Hand foot and mouth disease (HFMD) is a common infectious disease which mostly occurs in children under the age of five, causing severe illness and deaths [1]. However, there is no licensed vaccine against HFMD yet. Enterovirus 71(EV71) and coxsakievirus A16 (CA16) are responsible for the majority of the reported HFMD cases [2–5]. Both viruses are members of the enterovirus genus of the picornaviridae family [6,7]. EV71 infection is often associated with severe HFMD cases with neurological involvement or even deaths [1,6,8,9], and therefore has been the sole target for HFMD vaccine development [10–12]. Individuals infected with CA16 usually present as a self-limiting disease with mild symptoms [13,14]. However, accumulating evidences indicate that CA16 infections can also
∗ Corresponding author. Tel.: +86 21 54923067; fax: +86 21 54923044. E-mail address: [email protected] (Z. Huang). http://dx.doi.org/10.1016/j.vaccine.2014.06.025 0264-410X/© 2014 Elsevier Ltd. All rights reserved.
cause severe neurological complications and deaths [15–20]. Cocirculation of EV71 and CA16 has resulted in clinical cases of co-infection by the two viruses [13,14], in which the disease was exaggerated [21]. Therefore, to ensure a broad and effective protection against HFMD, both EV71 and CA16 should be targeted for vaccine development. EV71 vaccine development has progressed rapidly with three inactivated whole-virus vaccine candidates having completed phase 3 clinical trial [22]. However, there is a concern that individuals immunized with a single EV71 vaccine may still be infected by CA16 and develop HFMD, and thus may mislead the public into questioning the HFMD vaccine effectiveness [23]. A few CA16 vaccine candidates have been shown to induce protective immunity in animal models [24–27]. However, a bivalent vaccine that combines antigens derived from these two viruses for dual protection has not been developed until recently [28]. Recombinant virus-like particles (VLPs) have proven a successful strategy for development of new-generation vaccine (reviewed in [29,30]). We and others have shown that recombinant EV71
Z. Ku et al. / Vaccine 32 (2014) 4296–4303
VLPs were able to elicit protective immunity in animal models [31–33]. Recently, we also generated recombinant VLPs of CA16 and demonstrated their ability to induce high-titer CA16-neutralizing antibodies that protected mice against lethal challenge [25]. In this follow-up study, we determined, for the first time, the immunogenicity and protective efficacy in mice of a bivalent vaccine comprised of EV71- and CA16-VLPs. 2. Materials and methods
4297
HindIII, and then ligated into pIExBac-1 at the same sites, to make pIExBac-(CA16)P1. Similarly, the 3CD gene was amplified with primers (forward 5 -AATCCATGGGACCGAGCTTAGACTTCG3 and reverse 5 -GACAAGCTTCTAAAATAATTCGAGCCA-3 ), and cloned into NcoI/HindIII digested pIExBac-1, to make pIExBac(CA16)3CD. Generation of recombinant baculoviruses and infection of Sf9 cells were carried out as described previously [32]. The procedure for the expression and purification of CA16-VLPs was the same as that for EV71-VLPs described above.
2.1. Cells and viruses 2.5. SDS-PAGE and Western blotting RD, Vero and Neuro-2A cells were grown as described previously [34]. The EV71 strains used in this study included EV71/BrCr (purchased from ATCC), EV71/G082 [32], EV71/FY09-1, EV71/FY09-2, EV71/SZ98, and a mouse-adapted EV71 strain which is termed EV71/MAV-N. To prepare EV71/MAV-N, the parental strain EV71/G082 was subjected to two rounds of in vivo adaptation in one-day-old ICR mice followed by in vitro recovery in RD cells. Then, the resultant virus was passaged once in Neuro2A cells, yielding the EV71/MAV-N virus stock. Two CA16 strains, CA16/SZ05 and CA16/G08, have been described previously [25]. A coxsackievirus A10 (CA10) strain M.K. (Kowalik) was obtained from ATCC. All viruses were titrated for the 50% tissue culture infectious dose (TCID50 ) as described previously [35]. 2.2. Recombinant capsid subunit proteins and peptides Escherichia coli (E.coli)-expressed VP1 protein of EV71 was described previously [36]. E.coli-expressed VP1 protein of CA16 were described previously [34]. EV71-derived SP70 peptide (YPTFGEHKQEKDLEY) [37] and CA16-derived PEP71 peptide (FGEHLQANDLDYGQC) [38] were synthesized by GL Biochem (Shanghai, China). 2.3. Polyclonal and monoclonal antibodies Polyclonal antibodies against VP0, VP1, or VP3 proteins of CA16 were described previously [34]. Polyclonal antibodies against VP0 or VP1 proteins of EV71 were described previously [36,39]. AntiEV71 monoclonal antibody D5 was described previously [40]. The anti-CA16 monoclonal antibody 9B5 was generated in house from a mouse immunized with inactivated CA16. 2.4. VLP production and purification EV71-VLPs were produced as described previously [32]. Briefly, Sf9 insect cells co-infected with two baculoviruses expressing P1 and 3CD of EV71/G082 (genotype C4), respectively, were harvested at 72 h post infection, and lysed with Tris–NaCl buffer (50 mM Tris, 150 mM NaCl) containing 1% NP-40. The lysates were layered onto 20% sucrose cushion and subjected to ultracentrifugation at 25,000 rpm for 6 h. The resultant pellets were resuspended in Tris–NaCl buffer and then layered onto 10–50% sucrose gradients for ultracentrifugation at 39,000 rpm for 3 h. Twelve fractions were harvested from top to bottom and assayed for VLP as described previously. The VLP rich-fractions were pooled and subjected to size exclusion purification using a superdex-200 column (GE Healthcare, NJ, USA). Purified VLPs were quantified using a Bio-Rad Protein Assay kit (Bio-Rad, Hercules, CA, USA) according to the manufacturer’s instructions. The same approach was used to produce CA16-VLPs. To construct the baculovirus vectors, the P1 gene of CA16/SZ05 was amplified from the pFBD-CVP1/3CD plasmid [25] with primers (forward 5 -AATCCATGGGGTCACAAGTCTCC-3 and reverse 5 -GAC AAGCTTCTACAATGTTGTTATCTTGTC-3 ), digested with NcoI and
SDS-PAGE and Western blotting were carried out as described previously [25,32]. Briefly, purified EV71 and CA16 VLPs were separated on 12% polyacrylamide gels and proteins were stained with Coomassie blue R250 or transferred onto PVDF membranes for Western blot. The membranes were probed with a capsid subunit protein-specific polyclonal antibody followed by a corresponding HRP-conjugated second antibody.
2.6. Mouse immunization and virus challenge For immunization, vaccine antigens were formulated with a commercial adjuvant Imject Alum (Pierce, Rockford, IL, USA) which contains 40 mg/ml of aluminum hydroxide. Briefly, antigens were diluted with PBS to a desired concentration in a volume of 50 l and then mixed with Imject Alum at a volumetric ratio of 1:1 according to the manufacturer’s instructions. In one immunization experiment, groups of 10 female ICR mice (7 weeks old) were injected intraperitoneally (i.p.) with the antigen/alum mixtures containing PBS, 1 g of EV71-VLP, 1 g of CA16-VLP, or 1 g of EV71-VLP plus 1 g of CA16-VLP, respectively, at weeks 0 and 2. At week 3 (one week after the second immunization), three female mice from each group were randomly selected. Each of them was then housed separately and paired with an unimmunized naïve male mouse for mating. The remaining immunized female mice (seven animals per group) were bled for antibody measurement at week 4 and terminated at week 5. In another immunization experiment to determine the kinetics of antibody production, groups of five female ICR mice (6 weeks old) were injected i.p. with the indicated vaccine preparations at weeks 0 and 2. Blood samples were collected at 1, 3, 5, 7 and 9 weeks after the last immunization and assayed for neutralization titers. The in vivo protective efficacy of the VLP vaccines was evaluated by two assays. In one assay, groups of adult female mice were immunized with the VLP vaccines as described above and were then allowed to mate one week after the last immunization. The neonatal mice born to the VLP immunized dams were challenged within one day after birth by i.p. injection with either CA16/G08 (4 × 107 TCID50 ) or EV71/MAV-N (3.3 × 107 TCID50 ). The challenged mice were monitored daily for survival and clinical score for a period of 15 days. Clinical scores were graded as follows: 0, healthy; 1, reduced mobility; 2, limb weakness; 3, paralysis; 4, death. In another assay, groups of one-day-old ICR mice were i.p. injected with 50 l of antisera. One day later, the mice were inoculated i.p. with either CA16/G08 (8 × 107 TCID50 ) or EV71/MAV-N (6.6 × 107 TCID50 ). The challenged mice were monitored daily for survival and clinical score for a period of 16 days as described above. The animal studies were approved by the Institutional Animal Care and Use Committee at the Institut Pasteur of Shanghai and
4298
Z. Ku et al. / Vaccine 32 (2014) 4296–4303
the animals were cared for in accordance with the institutional guidelines.
of E.coli-expressed CA16 VP1 protein [34], 200 ng/well of the EV71 SP70 peptide (YPTFGEHKQEKDLEY) [41], or 200 ng/well of the CA16 PEP71 peptide (FGEHLQANDLDYGQC) [38], respectively.
2.7. Antibody measurement 2.8. Neutralization assay Virus-specific antibodies in immunized mouse sera were measured by indirect ELISA as described previously [32] with the following modification: 96-well plates were coated with 100 ng/well of E.coli-expressed EV71 VP1 proteins [36], 100 ng/well
Serum samples were analyzed for their neutralization capacity against a panel of EV71, CA16 and CA10 strains by in vitro microneutralization test as describled previously [32].
Fig. 1. Antigenic characterization of EV71-VLPs and CA16-VLPs. (A) SDS-PAGE and Western blot analyses of EV71-VLPs and CA16-VLPs. Purified VLPs were separated on 12% polyacrylamide gels and proteins were stained with Coomassie blue or transferred onto PVDF membranes for Western blot analysis using the indicated polyclonal antibodies against capsid subunit proteins. Lane M, protein marker; lane E, EV71-VLPs; lane C, CA16-VLPs. (B) ELISA analysis of the VLPs with polyclonal antibodies against capsid subunit proteins. The ELISA plate was coated with 10 ng of BSA, EV71-VLPs, or CA16-VLPs. Data are means ± SD of the OD450 readings of triplicate wells. (C to D) ELISA analysis of the VLPs with (C) the anti-EV71 monoclonal antibody D5 and (D) the anti-CA16 monoclonal antibody 9B5. Data are means ± SD of the OD450 readings of triplicate wells.
Z. Ku et al. / Vaccine 32 (2014) 4296–4303
2.9. Statistics All statistical analysis was performed with GraphPad Prism version 5. Kaplan–Meier survival curves were compared by the log rank test. Other experiments were analyzed by the Student’s twotailed t-test. 3. Results 3.1. Antigenic characterization of EV71- and CA16-VLPs EV71- and CA16-VLPs were produced in insect cells using baculovirus vectors. SDS-PAGE analysis of purified EV71- and CA16VLPs shows three protein bands representing VP0, VP1 and VP3 subunits for each preparation (Fig. 1A). Western blotting with a set of subunit protein-specific polyclonal antibodies confirms the identity of the protein bands and also reveals different degree of cross reactivity between heterologous antigen/antibody pairs. In particular, anti-(CA16)VP0 and anti-(CA16)VP3 antibodies reacted strongly with VP0 and VP3 proteins of EV71, whereas the crossreactivity of anti-(EV71)VP0, anti-(EV71)VP1, and anti-(CA16)VP1 with the counterpart antigens was weak (Fig. 1A). Interestingly, the inferred molecular weight of the VP1 proteins of EV71 and CA16 was somewhat different (Fig. 1A and B) although the total numbers of amino acid residues of these two proteins are the same. A similar observation was also made for the VP3 proteins (Fig. 1A).
4299
The VLPs were also tested for reactivity with polyclonal and monoclonal antibodies by ELISA. As shown in Fig. 1B, the control antigen BSA did not show significant reactivity with any antisera tested; whereas both EV71- and CA16-VLPs reacted with all antisera to different extent, in agreement with the above Western blotting results. Detection with monoclonal antibodies revealed that EV71-VLPs reacted in a dose-dependent manner with the EV71-neutralizing antibody D5 but not the anti-CA16 monoclonal antibody 9B5 (Fig. 1C and D), whereas CA16-VLPs were only recognized by 9B5 (Fig. 1C and D), thus confirming the identity of the two kinds of VLPs. 3.2. Antibody response following VLP immunization To determine the immunogenicity, three groups of mice were immunized i.p. with EV71-VLP, CA16-VLP, or bivalent VLP (designated Bi-VLP) consisting of 1:1 mixed EV71-VLP and CA16-VLP, respectively, at weeks 0 and 2. Another group of mice was injected with PBS as a control. At two weeks after the last immunization, sera were collected from all immunized mice for antibody measurement. All sera were diluted 1:100 and then analyzed by ELISA with different capture antigens, including the recombinant EV71 VP1 protein [36], the recombinant CA16 VP1 protein [34], the EV71derived SP70 peptide [41], and the CA16-derived PEP71 peptide [38]. Sera from the PBS control group did not show significant reactivity with the four coating antigens (Fig. 2A–D). For the EV71-VLP immunized mouse sera, strong reactivity was detected with the
Fig. 2. Antibody responses following VLP immunization in mice. Four groups of ICR mice were i.p immunized with PBS, EV71-VLP, CA16-VLP, or the bivalent VLP (designated BiVLP) consisting of both EV71- and CA16-VLP, respectively, as described in Section 2. Sera from seven mice each group were collected at two weeks after the last immunization. All sera were diluted 1:100 and then analyzed for antibody response by ELISA with the indicated coating antigens: (A) the recombinant VP1 protein of EV71, (B) the recombinant VP1 protein of CA16, (C) the SP70 peptide of EV71, or (D) the PEP71 peptide of CA16. Each symbol represents one mouse and the line indicates the geometric mean value of the group.
4300
Z. Ku et al. / Vaccine 32 (2014) 4296–4303
Fig. 3. Neutralization ability of the mouse anti-sera. (A and B) Serum samples from individual mice were collected at two weeks after the final immunization and tested for neutralization against (A) EV71/G082 or (B) CA16/SZ05 virus. The antisera of the PBS group did not show any neutralization for both viruses at a 1:16 dilution (the lowest dilution tested) and were therefore assigned a titer of 8 for geometric mean titer (GMT) computation. The data are representative results of two independent neutralization experiments. Each symbol represents one mouse, and the line indicates the GMT of the group. Statistical significance was determined by the Student’s t test and is indicated as follows: n.s., P > 0.05; * P < 0.05; ** P < 0.005. (C and D) Kinetics of neutralizing antibody titers. Four groups of five ICR mice were i.p immunized with PBS, EV71-VLP, CA16-VLP and Bi-VLP antigens as described in Section 2. Blood samples were collected every two weeks after the last immunization and individual sera were pooled for each group. The pooled sera were assayed for neutralization of (C) EV71/G082 and (D) CA16/SZ05. Representative results of two independent neutralization experiments are shown.
VP1 protein and SP70 peptide of EV71 (Fig. 2A and C), and weak or no reactivity with the corresponding CA16 antigens (Fig. 2B and D). Similarly, antisera of the CA16-VLP group reacted strongly with the CA16-derived VP1 protein and PEP71 peptide, and weakly with the EV71 counterparts (Fig. 2A–D). In contrast, the bivalent VLP immunized sera exhibited potent reactivity with all coating antigens (Fig. 2A–D).
to A genogroup, but had no effect on CA16; anti-CA16 VLP sera strongly neutralized homologous and heterologous CA16 strains and marginally neutralized all EV71 strains; in contrast, the bivalent VLP-immunized sera were able to neutralize all EV71 and CA16 strains at high titers. None of the antisera could neutralize CA10 (another enterovirus causing HFMD), suggesting even the bivalent VLP can not provide protection against CA10.
3.3. Neutralization capacity of VLP-immunized mouse sera
3.4. Durability of VLP-induced neutralizing antibody responses
Sera from the PBS or VLP immunized mice were evaluated for their capacity to neutralize EV71 and CA16 in vitro. As shown in Fig. 3, the anti-PBS sera did not show neutralization at 1:16 (the lowest dilution tested) and therefore assigned a titer of 8 for geometric mean titer (GMT) computation; the EV71-VLP immunized sera exhibited potent neutralization effects on EV71 but not CA16; the CA16-VLP immunized sera strongly neutralized CA16 (GMT = 927) and also weakly cross-neutralized EV71 (GMT = 32); in contrast, antisera from the bivalent vaccine group were able to potently neutralize both EV71 (GMT = 1680) and CA16 (GMT = 1131). To determine the neutralization breadth, antisera collected at three weeks after the last immunization were pooled for each group, and tested for neutralization against a virus panel. As shown in Table 1, anti-EV71 VLP sera exhibited potent neutralization against all EV71 strains, including the BrCr stain belonging
We next investigated whether VLP-induced neutralizing antibody responses are long-lasting. Mice were immunized twice at a 2-week interval with PBS, EV71-VLP, CA16-VLP, or Bi-VLP, and blood samples were collected at 1, 3, 5, 7, or 9 weeks after the last immunization. Sera from each individual immunized mouse were pooled for each group at each time point. The pooled sera were then tested for neutralization against EV71 and CA16 viruses. As shown in Fig. 3C and D, all VLP formulations induced robust neutralizing antibody responses in mouse sera since week 1, and the neutralization specificity and potency of the pooled antisera were consistent with the results from single mouse sera (Fig. 3A and B). More importantly, the neutralization titers against either EV71 or CA16 remained at similar levels throughout the whole 9-week test period for all groups. These results indicated that the bivalent VLP vaccine can induce a balanced and durable neutralizing antibody response.
Z. Ku et al. / Vaccine 32 (2014) 4296–4303
4301
Table 1 Neutralization capacity of anti-sera against a panel of EV71 and CA16 viruses. Neutralization titer against
Pooled antisera
Anti-PBS Anti-EV71 VLP Anti-CV16 VLP Anti-Bi VLP
EV71/BrCr
EV71/G082
EV71/FY09-1
EV71/FY09-2
EV71/SZ98
EV71/MAV-N
CA16/G08
CA16/SZ05
CA10
Contents lists available at ScienceDirect
Vaccine journal homepage: www.elsevier.com/locate/vaccine
A virus-like particle based bivalent vaccine confers dual protection against enterovirus 71 and coxsackievirus A16 infections in mice Zhiqiang Ku a , Qingwei Liu a , Xiaohua Ye a , Yicun Cai a , Xiaoli Wang a , Jinping Shi a , Dapeng Li a , Xia Jin a , Wenqi An b , Zhong Huang a,∗ a Vaccine Research Center, Key Laboratory of Molecular Virology & Immunology, Institut Pasteur of Shanghai, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai 200031, China b Hualan Biological Bacterin Company, Xinxiang 453003, Henan, China
a r t i c l e
i n f o
Article history: Received 26 February 2014 Received in revised form 22 May 2014 Accepted 6 June 2014 Available online 17 June 2014 Keywords: Enterovirus 71 Coxsackievirus A16 Virus-like particle Vaccine Neutralizing antibody
a b s t r a c t Enterovirus 71(EV71) and coxsackievirus A16 (CA16) are responsible for hand, foot and mouth disease which has been prevalent in Asia-Pacific regions, causing significant morbidity and mortality in young children. Co-circulation of and co-infection by both viruses underscores the importance and urgency of developing vaccines against both viruses simultaneously. Here we report the immunogenicity and protective efficacy of a bivalent combination vaccine comprised of EV71 and CA16 virus-like particles (VLPs). We show that monovalent EV71- or CA16-VLPs-elicited serum antibodies exhibited potent neutralization effect on the homotypic virus but little or no effect on the heterotypic one, whereas the antisera against the bivalent vaccine formulation were able to efficiently neutralize both EV71 and CA16, indicating there is no immunological interference between the two antigens with respect to their ability to induce virusspecific neutralizing antibodies. Passive immunization with monovalent VLP vaccines protected mice against a homotypic virus challenge but not heterotypic infection. Surprisingly, antibody-dependent enhancement (ADE) of disease was observed in mice passively transferred with mono-specific anti-CA16 VLP sera and subsequently challenged with EV71. In contrast, the bivalent VLP vaccine conferred full protection against lethal challenge by either EV71 or CA16, thus eliminating the potential of ADE. Taken together, our results demonstrate for the first time that the bivalent VLP approach represents a safe and efficacious vaccine strategy for both EV71 and CA16. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction Hand foot and mouth disease (HFMD) is a common infectious disease which mostly occurs in children under the age of five, causing severe illness and deaths [1]. However, there is no licensed vaccine against HFMD yet. Enterovirus 71(EV71) and coxsakievirus A16 (CA16) are responsible for the majority of the reported HFMD cases [2–5]. Both viruses are members of the enterovirus genus of the picornaviridae family [6,7]. EV71 infection is often associated with severe HFMD cases with neurological involvement or even deaths [1,6,8,9], and therefore has been the sole target for HFMD vaccine development [10–12]. Individuals infected with CA16 usually present as a self-limiting disease with mild symptoms [13,14]. However, accumulating evidences indicate that CA16 infections can also
∗ Corresponding author. Tel.: +86 21 54923067; fax: +86 21 54923044. E-mail address: [email protected] (Z. Huang). http://dx.doi.org/10.1016/j.vaccine.2014.06.025 0264-410X/© 2014 Elsevier Ltd. All rights reserved.
cause severe neurological complications and deaths [15–20]. Cocirculation of EV71 and CA16 has resulted in clinical cases of co-infection by the two viruses [13,14], in which the disease was exaggerated [21]. Therefore, to ensure a broad and effective protection against HFMD, both EV71 and CA16 should be targeted for vaccine development. EV71 vaccine development has progressed rapidly with three inactivated whole-virus vaccine candidates having completed phase 3 clinical trial [22]. However, there is a concern that individuals immunized with a single EV71 vaccine may still be infected by CA16 and develop HFMD, and thus may mislead the public into questioning the HFMD vaccine effectiveness [23]. A few CA16 vaccine candidates have been shown to induce protective immunity in animal models [24–27]. However, a bivalent vaccine that combines antigens derived from these two viruses for dual protection has not been developed until recently [28]. Recombinant virus-like particles (VLPs) have proven a successful strategy for development of new-generation vaccine (reviewed in [29,30]). We and others have shown that recombinant EV71
Z. Ku et al. / Vaccine 32 (2014) 4296–4303
VLPs were able to elicit protective immunity in animal models [31–33]. Recently, we also generated recombinant VLPs of CA16 and demonstrated their ability to induce high-titer CA16-neutralizing antibodies that protected mice against lethal challenge [25]. In this follow-up study, we determined, for the first time, the immunogenicity and protective efficacy in mice of a bivalent vaccine comprised of EV71- and CA16-VLPs. 2. Materials and methods
4297
HindIII, and then ligated into pIExBac-1 at the same sites, to make pIExBac-(CA16)P1. Similarly, the 3CD gene was amplified with primers (forward 5 -AATCCATGGGACCGAGCTTAGACTTCG3 and reverse 5 -GACAAGCTTCTAAAATAATTCGAGCCA-3 ), and cloned into NcoI/HindIII digested pIExBac-1, to make pIExBac(CA16)3CD. Generation of recombinant baculoviruses and infection of Sf9 cells were carried out as described previously [32]. The procedure for the expression and purification of CA16-VLPs was the same as that for EV71-VLPs described above.
2.1. Cells and viruses 2.5. SDS-PAGE and Western blotting RD, Vero and Neuro-2A cells were grown as described previously [34]. The EV71 strains used in this study included EV71/BrCr (purchased from ATCC), EV71/G082 [32], EV71/FY09-1, EV71/FY09-2, EV71/SZ98, and a mouse-adapted EV71 strain which is termed EV71/MAV-N. To prepare EV71/MAV-N, the parental strain EV71/G082 was subjected to two rounds of in vivo adaptation in one-day-old ICR mice followed by in vitro recovery in RD cells. Then, the resultant virus was passaged once in Neuro2A cells, yielding the EV71/MAV-N virus stock. Two CA16 strains, CA16/SZ05 and CA16/G08, have been described previously [25]. A coxsackievirus A10 (CA10) strain M.K. (Kowalik) was obtained from ATCC. All viruses were titrated for the 50% tissue culture infectious dose (TCID50 ) as described previously [35]. 2.2. Recombinant capsid subunit proteins and peptides Escherichia coli (E.coli)-expressed VP1 protein of EV71 was described previously [36]. E.coli-expressed VP1 protein of CA16 were described previously [34]. EV71-derived SP70 peptide (YPTFGEHKQEKDLEY) [37] and CA16-derived PEP71 peptide (FGEHLQANDLDYGQC) [38] were synthesized by GL Biochem (Shanghai, China). 2.3. Polyclonal and monoclonal antibodies Polyclonal antibodies against VP0, VP1, or VP3 proteins of CA16 were described previously [34]. Polyclonal antibodies against VP0 or VP1 proteins of EV71 were described previously [36,39]. AntiEV71 monoclonal antibody D5 was described previously [40]. The anti-CA16 monoclonal antibody 9B5 was generated in house from a mouse immunized with inactivated CA16. 2.4. VLP production and purification EV71-VLPs were produced as described previously [32]. Briefly, Sf9 insect cells co-infected with two baculoviruses expressing P1 and 3CD of EV71/G082 (genotype C4), respectively, were harvested at 72 h post infection, and lysed with Tris–NaCl buffer (50 mM Tris, 150 mM NaCl) containing 1% NP-40. The lysates were layered onto 20% sucrose cushion and subjected to ultracentrifugation at 25,000 rpm for 6 h. The resultant pellets were resuspended in Tris–NaCl buffer and then layered onto 10–50% sucrose gradients for ultracentrifugation at 39,000 rpm for 3 h. Twelve fractions were harvested from top to bottom and assayed for VLP as described previously. The VLP rich-fractions were pooled and subjected to size exclusion purification using a superdex-200 column (GE Healthcare, NJ, USA). Purified VLPs were quantified using a Bio-Rad Protein Assay kit (Bio-Rad, Hercules, CA, USA) according to the manufacturer’s instructions. The same approach was used to produce CA16-VLPs. To construct the baculovirus vectors, the P1 gene of CA16/SZ05 was amplified from the pFBD-CVP1/3CD plasmid [25] with primers (forward 5 -AATCCATGGGGTCACAAGTCTCC-3 and reverse 5 -GAC AAGCTTCTACAATGTTGTTATCTTGTC-3 ), digested with NcoI and
SDS-PAGE and Western blotting were carried out as described previously [25,32]. Briefly, purified EV71 and CA16 VLPs were separated on 12% polyacrylamide gels and proteins were stained with Coomassie blue R250 or transferred onto PVDF membranes for Western blot. The membranes were probed with a capsid subunit protein-specific polyclonal antibody followed by a corresponding HRP-conjugated second antibody.
2.6. Mouse immunization and virus challenge For immunization, vaccine antigens were formulated with a commercial adjuvant Imject Alum (Pierce, Rockford, IL, USA) which contains 40 mg/ml of aluminum hydroxide. Briefly, antigens were diluted with PBS to a desired concentration in a volume of 50 l and then mixed with Imject Alum at a volumetric ratio of 1:1 according to the manufacturer’s instructions. In one immunization experiment, groups of 10 female ICR mice (7 weeks old) were injected intraperitoneally (i.p.) with the antigen/alum mixtures containing PBS, 1 g of EV71-VLP, 1 g of CA16-VLP, or 1 g of EV71-VLP plus 1 g of CA16-VLP, respectively, at weeks 0 and 2. At week 3 (one week after the second immunization), three female mice from each group were randomly selected. Each of them was then housed separately and paired with an unimmunized naïve male mouse for mating. The remaining immunized female mice (seven animals per group) were bled for antibody measurement at week 4 and terminated at week 5. In another immunization experiment to determine the kinetics of antibody production, groups of five female ICR mice (6 weeks old) were injected i.p. with the indicated vaccine preparations at weeks 0 and 2. Blood samples were collected at 1, 3, 5, 7 and 9 weeks after the last immunization and assayed for neutralization titers. The in vivo protective efficacy of the VLP vaccines was evaluated by two assays. In one assay, groups of adult female mice were immunized with the VLP vaccines as described above and were then allowed to mate one week after the last immunization. The neonatal mice born to the VLP immunized dams were challenged within one day after birth by i.p. injection with either CA16/G08 (4 × 107 TCID50 ) or EV71/MAV-N (3.3 × 107 TCID50 ). The challenged mice were monitored daily for survival and clinical score for a period of 15 days. Clinical scores were graded as follows: 0, healthy; 1, reduced mobility; 2, limb weakness; 3, paralysis; 4, death. In another assay, groups of one-day-old ICR mice were i.p. injected with 50 l of antisera. One day later, the mice were inoculated i.p. with either CA16/G08 (8 × 107 TCID50 ) or EV71/MAV-N (6.6 × 107 TCID50 ). The challenged mice were monitored daily for survival and clinical score for a period of 16 days as described above. The animal studies were approved by the Institutional Animal Care and Use Committee at the Institut Pasteur of Shanghai and
4298
Z. Ku et al. / Vaccine 32 (2014) 4296–4303
the animals were cared for in accordance with the institutional guidelines.
of E.coli-expressed CA16 VP1 protein [34], 200 ng/well of the EV71 SP70 peptide (YPTFGEHKQEKDLEY) [41], or 200 ng/well of the CA16 PEP71 peptide (FGEHLQANDLDYGQC) [38], respectively.
2.7. Antibody measurement 2.8. Neutralization assay Virus-specific antibodies in immunized mouse sera were measured by indirect ELISA as described previously [32] with the following modification: 96-well plates were coated with 100 ng/well of E.coli-expressed EV71 VP1 proteins [36], 100 ng/well
Serum samples were analyzed for their neutralization capacity against a panel of EV71, CA16 and CA10 strains by in vitro microneutralization test as describled previously [32].
Fig. 1. Antigenic characterization of EV71-VLPs and CA16-VLPs. (A) SDS-PAGE and Western blot analyses of EV71-VLPs and CA16-VLPs. Purified VLPs were separated on 12% polyacrylamide gels and proteins were stained with Coomassie blue or transferred onto PVDF membranes for Western blot analysis using the indicated polyclonal antibodies against capsid subunit proteins. Lane M, protein marker; lane E, EV71-VLPs; lane C, CA16-VLPs. (B) ELISA analysis of the VLPs with polyclonal antibodies against capsid subunit proteins. The ELISA plate was coated with 10 ng of BSA, EV71-VLPs, or CA16-VLPs. Data are means ± SD of the OD450 readings of triplicate wells. (C to D) ELISA analysis of the VLPs with (C) the anti-EV71 monoclonal antibody D5 and (D) the anti-CA16 monoclonal antibody 9B5. Data are means ± SD of the OD450 readings of triplicate wells.
Z. Ku et al. / Vaccine 32 (2014) 4296–4303
2.9. Statistics All statistical analysis was performed with GraphPad Prism version 5. Kaplan–Meier survival curves were compared by the log rank test. Other experiments were analyzed by the Student’s twotailed t-test. 3. Results 3.1. Antigenic characterization of EV71- and CA16-VLPs EV71- and CA16-VLPs were produced in insect cells using baculovirus vectors. SDS-PAGE analysis of purified EV71- and CA16VLPs shows three protein bands representing VP0, VP1 and VP3 subunits for each preparation (Fig. 1A). Western blotting with a set of subunit protein-specific polyclonal antibodies confirms the identity of the protein bands and also reveals different degree of cross reactivity between heterologous antigen/antibody pairs. In particular, anti-(CA16)VP0 and anti-(CA16)VP3 antibodies reacted strongly with VP0 and VP3 proteins of EV71, whereas the crossreactivity of anti-(EV71)VP0, anti-(EV71)VP1, and anti-(CA16)VP1 with the counterpart antigens was weak (Fig. 1A). Interestingly, the inferred molecular weight of the VP1 proteins of EV71 and CA16 was somewhat different (Fig. 1A and B) although the total numbers of amino acid residues of these two proteins are the same. A similar observation was also made for the VP3 proteins (Fig. 1A).
4299
The VLPs were also tested for reactivity with polyclonal and monoclonal antibodies by ELISA. As shown in Fig. 1B, the control antigen BSA did not show significant reactivity with any antisera tested; whereas both EV71- and CA16-VLPs reacted with all antisera to different extent, in agreement with the above Western blotting results. Detection with monoclonal antibodies revealed that EV71-VLPs reacted in a dose-dependent manner with the EV71-neutralizing antibody D5 but not the anti-CA16 monoclonal antibody 9B5 (Fig. 1C and D), whereas CA16-VLPs were only recognized by 9B5 (Fig. 1C and D), thus confirming the identity of the two kinds of VLPs. 3.2. Antibody response following VLP immunization To determine the immunogenicity, three groups of mice were immunized i.p. with EV71-VLP, CA16-VLP, or bivalent VLP (designated Bi-VLP) consisting of 1:1 mixed EV71-VLP and CA16-VLP, respectively, at weeks 0 and 2. Another group of mice was injected with PBS as a control. At two weeks after the last immunization, sera were collected from all immunized mice for antibody measurement. All sera were diluted 1:100 and then analyzed by ELISA with different capture antigens, including the recombinant EV71 VP1 protein [36], the recombinant CA16 VP1 protein [34], the EV71derived SP70 peptide [41], and the CA16-derived PEP71 peptide [38]. Sera from the PBS control group did not show significant reactivity with the four coating antigens (Fig. 2A–D). For the EV71-VLP immunized mouse sera, strong reactivity was detected with the
Fig. 2. Antibody responses following VLP immunization in mice. Four groups of ICR mice were i.p immunized with PBS, EV71-VLP, CA16-VLP, or the bivalent VLP (designated BiVLP) consisting of both EV71- and CA16-VLP, respectively, as described in Section 2. Sera from seven mice each group were collected at two weeks after the last immunization. All sera were diluted 1:100 and then analyzed for antibody response by ELISA with the indicated coating antigens: (A) the recombinant VP1 protein of EV71, (B) the recombinant VP1 protein of CA16, (C) the SP70 peptide of EV71, or (D) the PEP71 peptide of CA16. Each symbol represents one mouse and the line indicates the geometric mean value of the group.
4300
Z. Ku et al. / Vaccine 32 (2014) 4296–4303
Fig. 3. Neutralization ability of the mouse anti-sera. (A and B) Serum samples from individual mice were collected at two weeks after the final immunization and tested for neutralization against (A) EV71/G082 or (B) CA16/SZ05 virus. The antisera of the PBS group did not show any neutralization for both viruses at a 1:16 dilution (the lowest dilution tested) and were therefore assigned a titer of 8 for geometric mean titer (GMT) computation. The data are representative results of two independent neutralization experiments. Each symbol represents one mouse, and the line indicates the GMT of the group. Statistical significance was determined by the Student’s t test and is indicated as follows: n.s., P > 0.05; * P < 0.05; ** P < 0.005. (C and D) Kinetics of neutralizing antibody titers. Four groups of five ICR mice were i.p immunized with PBS, EV71-VLP, CA16-VLP and Bi-VLP antigens as described in Section 2. Blood samples were collected every two weeks after the last immunization and individual sera were pooled for each group. The pooled sera were assayed for neutralization of (C) EV71/G082 and (D) CA16/SZ05. Representative results of two independent neutralization experiments are shown.
VP1 protein and SP70 peptide of EV71 (Fig. 2A and C), and weak or no reactivity with the corresponding CA16 antigens (Fig. 2B and D). Similarly, antisera of the CA16-VLP group reacted strongly with the CA16-derived VP1 protein and PEP71 peptide, and weakly with the EV71 counterparts (Fig. 2A–D). In contrast, the bivalent VLP immunized sera exhibited potent reactivity with all coating antigens (Fig. 2A–D).
to A genogroup, but had no effect on CA16; anti-CA16 VLP sera strongly neutralized homologous and heterologous CA16 strains and marginally neutralized all EV71 strains; in contrast, the bivalent VLP-immunized sera were able to neutralize all EV71 and CA16 strains at high titers. None of the antisera could neutralize CA10 (another enterovirus causing HFMD), suggesting even the bivalent VLP can not provide protection against CA10.
3.3. Neutralization capacity of VLP-immunized mouse sera
3.4. Durability of VLP-induced neutralizing antibody responses
Sera from the PBS or VLP immunized mice were evaluated for their capacity to neutralize EV71 and CA16 in vitro. As shown in Fig. 3, the anti-PBS sera did not show neutralization at 1:16 (the lowest dilution tested) and therefore assigned a titer of 8 for geometric mean titer (GMT) computation; the EV71-VLP immunized sera exhibited potent neutralization effects on EV71 but not CA16; the CA16-VLP immunized sera strongly neutralized CA16 (GMT = 927) and also weakly cross-neutralized EV71 (GMT = 32); in contrast, antisera from the bivalent vaccine group were able to potently neutralize both EV71 (GMT = 1680) and CA16 (GMT = 1131). To determine the neutralization breadth, antisera collected at three weeks after the last immunization were pooled for each group, and tested for neutralization against a virus panel. As shown in Table 1, anti-EV71 VLP sera exhibited potent neutralization against all EV71 strains, including the BrCr stain belonging
We next investigated whether VLP-induced neutralizing antibody responses are long-lasting. Mice were immunized twice at a 2-week interval with PBS, EV71-VLP, CA16-VLP, or Bi-VLP, and blood samples were collected at 1, 3, 5, 7, or 9 weeks after the last immunization. Sera from each individual immunized mouse were pooled for each group at each time point. The pooled sera were then tested for neutralization against EV71 and CA16 viruses. As shown in Fig. 3C and D, all VLP formulations induced robust neutralizing antibody responses in mouse sera since week 1, and the neutralization specificity and potency of the pooled antisera were consistent with the results from single mouse sera (Fig. 3A and B). More importantly, the neutralization titers against either EV71 or CA16 remained at similar levels throughout the whole 9-week test period for all groups. These results indicated that the bivalent VLP vaccine can induce a balanced and durable neutralizing antibody response.
Z. Ku et al. / Vaccine 32 (2014) 4296–4303
4301
Table 1 Neutralization capacity of anti-sera against a panel of EV71 and CA16 viruses. Neutralization titer against
Pooled antisera
Anti-PBS Anti-EV71 VLP Anti-CV16 VLP Anti-Bi VLP
EV71/BrCr
EV71/G082
EV71/FY09-1
EV71/FY09-2
EV71/SZ98
EV71/MAV-N
CA16/G08
CA16/SZ05
CA10
