Vaccine 33 (2015) 2335–2341
Contents lists available at ScienceDirect
Vaccine journal homepage: www.elsevier.com/locate/vaccine
High-yield production of recombinant virus-like particles of enterovirus 71 in Pichia pastoris and their protective efficacy against oral viral challenge in mice Chao Zhang, Zhiqiang Ku, Qingwei Liu, Xiaoli Wang, Tan Chen, Xiaohua Ye, Dapeng Li, Xia Jin, Zhong Huang ∗ Vaccine Research Center, Key Laboratory of Molecular Virology & Immunology, Institut Pasteur of Shanghai, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 320 Yueyang Road, Shanghai 200031, China
a r t i c l e
i n f o
Article history: Received 14 December 2014 Received in revised form 23 February 2015 Accepted 12 March 2015 Available online 25 March 2015 Keywords: Enterovirus 71 Virus-like particle Vaccine Pichia pastoris Oral infection
a b s t r a c t Enterovirus 71 (EV71) is one of the major causative pathogens of hand, foot and mouth disease (HFMD), which is highly prevalent in the Asia-Pacific regions. Severe HFMD cases with neurological complications and even death are often associated with EV71 infections. However, no licensed EV71 vaccine is currently available. Recombinant virus-like particles (VLPs) of EV71 have been produced and shown to be a promising vaccine candidate in preclinical studies. However, the performance of current recombinant expression systems for EV71 VLP production remains unsatisfactory with regard to VLP yield and manufacturing procedure, and thus hinders further product development. In this study, we evaluated the expression of EV71 VLPs in Pichia pastoris and determined their protective efficacy in mouse models of EV71 infections. We showed that EV71 VLPs could be produced at high levels up to 4.9% of total soluble protein in transgenic P. pastoris yeast co-expressing P1 and 3CD proteins of EV71. The resulting yeastproduced VLPs potently induced neutralizing antibodies against homologous and heterologous EV71 strains in mice. More importantly, maternal immunization with VLPs protected neonatal mice in both intraperitoneal and oral challenge experiments. Collectively, these results demonstrated the success of simple, high-yield production of EV71 VLPs in transgenic P. pastoris, thus lifting the major roadblock in commercial development of VLP-based EV71 vaccines. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Enterovirus 71 (EV71) is the major causative agent of hand, foot and mouth disease (HFMD) [1–4]. EV71 transmission naturally occurs through the oral–fecal route. Infection with EV71 has been often associated with severe HFMD cases with neurological complications, such as brainstem encephalitis, polio-like paralysis, and pulmonary edema [1–4]. However, no commercial vaccine is currently available for preventing EV71 infections [5–7]. Virus-like particles (VLPs) are safe and highly immunogenic and therefore constitute a state-of-the-art platform for vaccine development [8,9]. Indeed, EV71-derived VLPs have been produced in baculovirus/insect cell expression system and shown to induce neutralizing antibodies in mice [10,11] and monkeys [12]. More importantly, insect cell-produced VLPs conferred protection in
mice inoculated intraperitoneally (i.p.) with lethal dose of EV71 [10,13]. Recently, it has been shown that EV71 VLPs could be produced in transgenic Saccharomyces cerevisiae [14]. However, the VLP expression in S. cerevisiae was at low levels of approximately 0.25 mg per liter yeast culture [14] and thus may hinder further product development. In the present study, we evaluated the expression of EV71 VLPs in another yeast species, Pichia pastoris, and determined the protective efficacy of these VLPs in mouse models of EV71 infections. Our results demonstrated that EV71 VLPs could be produced in P. pastoris at high levels. Moreover, P. pastoris-produced VLPs were found to elicit protective immunity against both i.p. and oral viral challenges in mice. 2. Materials and methods 2.1. Cells and viruses
∗ Corresponding author. Tel.:+86 21 54923067; fax: +86 21 54923044. E-mail addresses: [email protected], [email protected] (Z. Huang). http://dx.doi.org/10.1016/j.vaccine.2015.03.034 0264-410X/© 2015 Elsevier Ltd. All rights reserved.
RD and Vero cells were grown as described previously [15]. PichiaPinkTM yeast strains (Invitrogen, USA) were grown according
2336
C. Zhang et al. / Vaccine 33 (2015) 2335–2341
to the manufacturer’s instruction. EV71 strains used in this study, including EV71/BrCr, EV71/G081, EV71/G082, EV71/FY091, EV71/FY09-2 and EV71/SZ98, were described previously [13]. A mouse adapted EV71 strain termed EV71/MAV-W was prepared as follows: the parental EV71/G082 virus was subjected to four rounds of in vivo adaption in one-day-old ICR mice followed by in vitro recovery in Vero cells. Then, the resultant virus was propagated once in Vero cells, yielding the EV71/MAV-W virus stock. Two CA16 strains, CA16/SZ05 and CA16/G08, were described previously [16]. Virus titers were determined as previously described [15,17]. 2.2. Antibodies and peptide Rabbit polyclonal antibodies for detection of EV71 capsid subunit proteins were described previously [18]. Rabbit polyclonal antibody against EV71 VLP was generated in house from a rabbit immunized with insect cell-derived EV71 VLP. SP70 peptide (-YPTFGEHKQEKDLEY-) was synthesized by GL Biochem (Shanghai, China). 2.3. Vector construction and Pichia pastoris transformation The P1 gene of the EV71/G082 strain was codon-optimized, synthesized, and cloned into the backbone vector pPink-HC (Invitrogen), resulting in plasmid YE001. Similarly, the 3CD gene was engineered into pPink-HC to make plasmid YE002. The plasmid YE003 was generated by inserting the 3CD expression cassette into YE001 from the Bgl II site. For P. pastoris transformation, plasmids were linearized by Afl II digestion and then used to transform PichiaPinkTM Strain 1 (Invitrogen) by electroporation. The P. pastoris transformation and subsequent selection of transformants were performed according to the manufacturer’s instruction.
at 37 ◦ C for 1 h. Color development and absorbance measurement were carried out as described previously [11]. Western blot assays were performed as described previously [11].
2.6. Preparation of VLP and control antigen for immunization To prepare the VLP antigen, the selected yeast strains were grown and then induced for antigen expression with methanol according to the manufacturer’s instruction (Invitrogen). The induced yeast cells were pelleted and suspended in 0.15 M phosphate buffered saline (PBS) buffer. The cells were then lysed using a high pressure cell disrupter (JNBIO, China) at 1800 Bar, and the resultant lysates were clarified by centrifugation at 12,000 × g for 15 min. Then, NaCl and PEG 8000 were added into the lysates to reach final concentrations of 200 mM and 10% (W/V), respectively, and the mixtures were stirred gently at 4 ◦ C overnight to allow protein precipitation. Next day, the mixtures were subjected to centrifugation at 12,000 × g for 15 min, and the resultant pellets were resuspended in PBS buffer, followed by centrifugation at 12,000 × g for 15 min to remove insoluble fractions. The clarified supernatants were then subjected to sucrose cushion and sucrose gradient ultracentrifugation as previously described [11]. The final VLP preparation was quantified by ELISA as described above. The empty vector pPink-HC transformed yeast cells were processed in an identical manner to generate the negative control antigen for immunization experiments.
2.7. Electron microscopy Electron microscopy was performed as described previously [11].
2.4. Screening of P. pastoris transformants for VLP expression To identify high-expression P. pastoris strains, small-scale expression experiments were performed according to the manufacturer’s instruction (Invitrogen). Briefly, transformed yeast colonies were individually inoculated into 5 ml BMGY medium and grown with shaking at 30 ◦ C for 24 h. The yeast cells were pelleted by centrifugation and then resuspended in 1 ml BMMY containing 0.5% methanol followed by culture at 30 ◦ C for 48 h. At last, yeast cells from each culture were harvested, resuspended in 150 l breaking buffer, mixed with equal volume of acid-washed glass beads (Sigma), and then subjected to repeated vortex. The broken cell/bead mixtures were clarified by centrifugation at 12,000 rpm for 10 min, and the resulting supernatants were analyzed for VLP expression by ELISA and Western blotting as described below. Total soluble protein (TSP) in the supernatants was measured by Bradford assay. 2.5. ELISA and Western blot assays For antigen quantification, ELISA was performed with insect cell-derived purified EV71 VLPs [13] as the reference standard. Briefly, wells of 96-well microtiter plates were coated with 50 l of rabbit anti-EV71 VLP sera (diluted 1:10,000 in PBS) at 4 ◦ C overnight; then the plates were washed three times with PBST buffer after each of the following steps. Consecutively, 200 l/well of 5% milk in PBST was added for blocking and incubated at 37 ◦ C for 1 h; 50 l/well of the protein lysates or VLP reference standard serially diluted in PBST plus 1% milk was added and incubated at 37 ◦ C for 2 h; 50 l/well of anti-EV71 monoclonal antibody D5 [15] was added and incubated at 37 ◦ C for 1 h; then 50 l/well of HRP-Conjugated anti-mouse IgG (Sigma) was added and incubated
2.8. Mouse immunization and virus challenge The animal studies were approved by the Institutional Animal Care and Use Committee at the Institut Pasteur of Shanghai. ICR mice used in this study were purchased from Shanghai Laboratory Anima Center (SLAC). Prior to immunization, antigens were mixed with Alhydrogel® (Invivogen, USA) adjuvant to make the experimental vaccines. A single injection dose contained 1 g antigen (EV71 VLPs or similarly prepared negative control antigen) and 0.5 mg aluminum hydroxide in a final volume of 100 l. This vaccine dosage was chosen because it was previously shown that two doses (1 g/dose) of insect cell-expressed EV71 VLPs were sufficient to induce high-titer neutralizing antibody response in mice [13]. Groups of 6 female ICR mice (6–8 week) were injected intraperitoneally (i.p.) with the experimental vaccines at week 0 and 2. Blood samples were collected at week 4 for antibody measurement. For passive immunization/protection experiments, groups of female ICR mice were immunized with VLPs or control antigen as described above and allowed to mate two weeks after the last immunization. In the i.p. challenge assay, the 5-day-old neonatal mice born to immunized dams were i.p. injected with 4.2 × 104 TCID50 of EV71/MAV-W. In the oral challenge assay, the 1-dayold neonatal mice born to immunized dams were orally inoculated with 1.26 × 105 TCID50 of EV71/MAV-W. The EV71/MAV-W doses used in the challenge assays were determined in our preliminary studies (data not shown). The challenged mice were monitored daily for survival and clinical score for a period of 14 days. Clinical scores were graded as follows: 0, healthy; 1, reduced mobility; 2, limb weakness; 3, paralysis; 4, death.
C. Zhang et al. / Vaccine 33 (2015) 2335–2341
2.9. Antibody measurement and neutralization assay The virus-specific antibodies in immunized mouse sera were measured by ELISA assay with 200 ng/well of EV71 SP70 peptide as the coating antigen as previously described [13]. Neutralization assays were performed as previously described [11].
3. Results 3.1. High-yield expression of EV71 VLPs in Pichia pastoris Three vectors were constructed for the expression of EV71 VLPs in Pichia pastoris: YE001 and YE002 contained a single P1 and a single 3CD expression cassette, respectively, whereas YE003 possessed both P1 and 3CD cassettes (Fig. 1A). We first co-transformed the PichiaPinkTM yeast with both YE001 and YE002. A yeast clone transformed with the empty vector pPink-HC served as the negative control. EV71 antigen in the lysates of the resulting transformants was quantified by ELISA and total soluble protein (TSP) was measured by Bradford assay. For screening, relative EV71 antigen level for a given sample was expressed as a percentage of TSP. As shown in Fig. 1B, no EV71 protein was detected for the control sample (empty vector-transformed yeast); whereas 2 out of 15 clones produced EV71 antigen at levels above 0.5% TSP, with the highest one being 1.45%. We next transformed the same parent yeast strain with YE003. Screening of 15 YE003-transformed clones showed that most of them produced EV71 antigen at levels above 0.5% TSP and the highest expression reached 4.9% (Fig. 1B). These data demonstrate that YE003 is a better vector than the YE001/YE002 combination at achieving high EV71 antigen expression levels. Cleavage of P1 protein into VP0, VP1, and VP3 subunit proteins is required for VLP assembly. To determine whether the expressed P1 protein was processed by 3CD, lysate from the YE003-transformed and the empty vector-transformed yeast clones was subjected to Western blotting with VP1-specific antisera as the detection
2337
antibody and insect cell-derived EV71 VLPs as the reference standard. No specific band was observed for the empty vector transformant; whereas the YE003-transformed yeast produced a ∼35 kDa band as did the insect cell-derived EV71 VLPs (Fig. 1C), indicating that the P1 polyprotein was properly processed to yield the capsid subunit proteins. To evaluate VLP assembly, yeast lysates were subjected to sucrose gradient ultracentrifugation. The resulting fractions were analyzed by ELISA. No significant reactivity was observed in the control yeast samples (Fig. 2A). In contrast, the profile of the lysate of YE003-transformed yeast is similar to that of insect cell-derived VLPs, with the peak reactivity detected in the fractions #8 (Fig. 2A), suggesting the formation of yeast-derived VLPs. Moreover, Western blot analysis of the yeast fractions showed that VP0, VP1 and VP3 proteins co-sedimented (Fig. 2B), indicating that the VLPs were made of these three capsid subunit proteins. Furthermore, examination of the yeast-derived VLP preparation by electron microscopy revealed the presence of spherical particles with a diameter of 30 nm (Fig. 2C).
3.2. Induction of high-titer neutralizing antibodies by yeast-derived VLPs For immunization studies, VLP was purified from YE003transformed yeast and the control antigen was prepared in the same manner from an empty vector (pPink-HC) transformant. The yeast-derived VLP showed a banding pattern identical to that of the insect cell-produced VLP reference standard in SDS-PAGE (Fig. 3A). To evaluate their immunogenicity, the yeast-derived VLP or the control antigen was used to i.p. immunize mice at weeks 0 and 2. Serum samples were collected at weeks 2 and 4 and analyzed for EV71-specific antibodies by ELISA with the SP70 peptide as the capture antigen. As shown in Fig. 3B, 4 of 6 mice in the VLP group became positive for EV71-specific antibody after the first dose and all of them were seroconverted after the booster injection;
Fig. 1. Co-expression of P1 and 3CD of EV71 in Pichia pastoris. (A) Diagrams of the constructs used in this study. TRP2-L and TRP2-R, the up- and down-stream parts of the TRP region; PAOX1 , AOX1 promoter; CYC1 TT, CYC1 transcription termination region; ADE2, expression cassette encoding phosphoribosylaminoimidazole carboxylase, used as the selection marker. (B) EV71 antigen expression levels in different yeast clones. The lysates from different clones were measured for EV71 antigen by ELISA and for total soluble protein (TSP) by Bradford assay. Lysate from an empty vector-transformed yeast clone was used as the negative control (ctr). Data are mean ± SD of triplicate wells. (C) Western blot analysis with an anti-VP1 polyclonal antibody. Lane M, marker; lane 1, empty vector-transformed yeast; lane 2, YE003-transformed yeast; lane 3, insect cell-produced EV71 VLPs.
2338
C. Zhang et al. / Vaccine 33 (2015) 2335–2341
Fig. 2. Characterization of VLP assembly. (A–B) Characterization of yeast-derived VLPs by sucrose gradient analyses. Yeast-derived antigens and insect cell-produced VLP were subjected to sucrose gradient ultracentrifugation. Twelve fractions were taken from top to bottom for each sample. The gradient fractions were then analyzed by (A) ELISA and (B) Western blotting with anti-VP0, anti-VP1, or anti-VP3 polyclonal antibodies, respectively. (C) Electron microscopy of yeast-derived VLPs. Bar = 100 nm.
in contrast, none of mice injected with the control antigen had detectable anti-EV71 antibodies even after the second immunization. Analysis of the same serum samples by ELISA with insect cell-derived VLP as the capture antigen revealed a reactivity profile (Fig. 3C) similar to that with the SP70 peptide as the capture antigen (Fig. 3B). These results indicate that the VLP, but not the control yeast antigen, elicited EV71-specific antibodies. The ability of individual antisera to inhibit in vitro EV71 infection was assessed by the micro-neutralization assay. The antisera from the control antigen-immunized mice had no neutralization effect even at the lowest dilution (1:16) and was therefore assigned a titer of 8 for computation of geometric mean titer (GMT); in contrast, the anti-VLP sera potently neutralized the homologous strain EV71/G082 with a GMT of 8192 (Fig. 3D). To determine their cross-neutralization ability, the individual antisera collected at week 4 were pooled for each group and
Fig. 3. Yeast-derived VLPs elicited high-titer neutralizing antibodies in mice. (A) SDSPAGE analysis of the antigens used for mouse immunization. Lane M, protein ladder; lane 1, control antigen prepared from empty vector-transformed yeast; lane 2, VLPs prepared from YE003-transformed yeast; lane 3, insect cell-produced VLPs. (B) EV71-specific antibody responses in mice following immunizations. Groups of six ICR mice were injected i.p. at weeks 0 and 2 with 1 g/dose of yeast-derived VLPs or the control antigen. Sera collected at weeks 2 and 4 were analyzed for EV71-specific antibody by ELISA with the SP70 peptide as capture antigen. The antisera were diluted 1:20 and used in ELISA. Each symbol represents a mouse, and the solid line indicates the geometric mean value of the group. (C) EV71-specific antibody responses analyzed by ELISA with insect cell-derived VLPs as the capture antigen. The antisera were diluted 1:1,000 and used in ELISA. Each symbol represents a mouse, and the solid line indicates the geometric mean value of the group. (D) Neutralization titers of antisera determined by the micro-neutralization assay. The antisera from mice received the empty vector control antigen did not show any neutralization activity at 1:16 dilution (the lowest dilution tested) and were therefore assigned a titer of 8 for GMT computation. Each symbol represents a mouse, and the solid line indicates the geometric mean value of the group.
C. Zhang et al. / Vaccine 33 (2015) 2335–2341
Fig. 4. Maternal immunization with VLPs protected newborn mice against i.p. viral challenge. Neonatal mice (5-day-old) born to dams immunized with VLP or the control antigen were i.p. injected with 4.2 × 104 TCID50 of EV71/MAV-W. Then the challenged mice were monitored daily for (A) survival and (B) clinical score for a period of 14 days. Clinical scores were graded as follows: 0, healthy; 1, reduced mobility; 2, limb weakness; 3, paralysis; 4, death. The numbers of mice in each group were indicated in the bracket.
then tested for neutralization against a panel of enteroviruses. As shown in Table 1, the pooled antisera from the control antigenimmunized mice did not exhibit neutralization against any of the virus strains tested; in contrast, the pooled anti-VLP sera potently neutralized all EV71 strains but showed no neutralization effect on the two CA16 strains even at the highest concentration tested (1:32). These results demonstrated that the neutralization activity of the anti-VLP antibodies is EV71-specific and intratypically cross-reactive.
2339
Fig. 5. Maternal immunization with VLPs protected newborn mice against oral infection. Neonatal mice (1-day-old) born to dams immunized with VLP or the control antigen were orally administrated with 1.26 × 105 TCID50 of EV71/MAV-W. Then the inoculated mice were monitored daily for (A) survival and (B) clinical score for a period of 14 days. Clinical scores were graded as follows: 0, healthy; 1, reduced mobility; 2, limb weakness; 3, paralysis; 4, death. The numbers of mice in each group were indicated in the bracket.
3.4. Maternal immunization with yeast-derived VLPs protected neonatal mice against oral infection We further examined whether VLP immunization could protect mice from EV71 oral infection. One-day-old mice born to immunized dams were orally inoculated with the EV71/MAV-W, and then observed daily for survival and clinical signs. As shown in Fig. 5, the control mice gradually developed clinical signs, including reduced mobility and limb weakness and paralysis, and eventually died with a final mortality rate of 41.7%; in contrast, all mice born to the VLP immunized dam survived without notable clinical signs.
3.3. Maternal immunization with yeast-derived VLPs protected neonatal mice against i.p. viral challenge
4. Discussion
To evaluate the protective efficacy, female ICR mice that had been immunized with VLPs or the control antigen were allowed to mate with naïve male mice, and the neonatal mice born to immunized dams were inoculated i.p. with the mouse-adapted EV71 strain, EV71/MAV-W. The challenged mice were subsequently observed on a daily basis for survival and clinical signs for a period of 14 days. The neonatal mice born to the control antigen-immunized dams started to show clinical signs at 3 days post-infection (dpi) and eventually died within 11 dpi (Fig. 4). In contrast, the survival rate for the VLP group was 92.9% and all the surviving mice were free of significant clinical signs (Fig. 4). These results indicate that yeast-derived VLP can protect mice from lethal viral challenge.
Recently, three inactivated EV71 vaccine candidates have shown protective efficacy in phase 3 clinical trials [19–21]. However, the production of inactivated EV71 was at relatively low levels in cultures. For example, it was reported that about 1.25 mg/L of inactivated EV71 could be produced in 40-l pilot-scale production runs [22]. It is therefore challenging to produce large quantity of inactivated EV71 vaccines for worldwide or even nationwide mass immunization of the at-risk populations. As an alternative EV71 vaccine platform, VLPs have been produced in recombinant expression systems and they have been shown to efficiently protect against lethal challenge in animal models [10,13,14]. In particular, S. cerevisiae yeast, which has been successfully used to produce VLP-based HPV vaccines, was tested for EV71 VLP expression. However, the EV71 VLP yield in
2340
C. Zhang et al. / Vaccine 33 (2015) 2335–2341
Table 1 Neutralization capacity of antisera against a panel of EV71 and CA16 strains. Pooled antisera against
Neutralization titer against
EV71/ BrCr Control antigen VLP
EV71/ G081
EV71/ G082
EV71/ FY09-1
EV71/ FY09-2
EV71/ SZ98
EV71/ MAV-W
CA16/ SZ05
CA16/ G08
Contents lists available at ScienceDirect
Vaccine journal homepage: www.elsevier.com/locate/vaccine
High-yield production of recombinant virus-like particles of enterovirus 71 in Pichia pastoris and their protective efficacy against oral viral challenge in mice Chao Zhang, Zhiqiang Ku, Qingwei Liu, Xiaoli Wang, Tan Chen, Xiaohua Ye, Dapeng Li, Xia Jin, Zhong Huang ∗ Vaccine Research Center, Key Laboratory of Molecular Virology & Immunology, Institut Pasteur of Shanghai, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 320 Yueyang Road, Shanghai 200031, China
a r t i c l e
i n f o
Article history: Received 14 December 2014 Received in revised form 23 February 2015 Accepted 12 March 2015 Available online 25 March 2015 Keywords: Enterovirus 71 Virus-like particle Vaccine Pichia pastoris Oral infection
a b s t r a c t Enterovirus 71 (EV71) is one of the major causative pathogens of hand, foot and mouth disease (HFMD), which is highly prevalent in the Asia-Pacific regions. Severe HFMD cases with neurological complications and even death are often associated with EV71 infections. However, no licensed EV71 vaccine is currently available. Recombinant virus-like particles (VLPs) of EV71 have been produced and shown to be a promising vaccine candidate in preclinical studies. However, the performance of current recombinant expression systems for EV71 VLP production remains unsatisfactory with regard to VLP yield and manufacturing procedure, and thus hinders further product development. In this study, we evaluated the expression of EV71 VLPs in Pichia pastoris and determined their protective efficacy in mouse models of EV71 infections. We showed that EV71 VLPs could be produced at high levels up to 4.9% of total soluble protein in transgenic P. pastoris yeast co-expressing P1 and 3CD proteins of EV71. The resulting yeastproduced VLPs potently induced neutralizing antibodies against homologous and heterologous EV71 strains in mice. More importantly, maternal immunization with VLPs protected neonatal mice in both intraperitoneal and oral challenge experiments. Collectively, these results demonstrated the success of simple, high-yield production of EV71 VLPs in transgenic P. pastoris, thus lifting the major roadblock in commercial development of VLP-based EV71 vaccines. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Enterovirus 71 (EV71) is the major causative agent of hand, foot and mouth disease (HFMD) [1–4]. EV71 transmission naturally occurs through the oral–fecal route. Infection with EV71 has been often associated with severe HFMD cases with neurological complications, such as brainstem encephalitis, polio-like paralysis, and pulmonary edema [1–4]. However, no commercial vaccine is currently available for preventing EV71 infections [5–7]. Virus-like particles (VLPs) are safe and highly immunogenic and therefore constitute a state-of-the-art platform for vaccine development [8,9]. Indeed, EV71-derived VLPs have been produced in baculovirus/insect cell expression system and shown to induce neutralizing antibodies in mice [10,11] and monkeys [12]. More importantly, insect cell-produced VLPs conferred protection in
mice inoculated intraperitoneally (i.p.) with lethal dose of EV71 [10,13]. Recently, it has been shown that EV71 VLPs could be produced in transgenic Saccharomyces cerevisiae [14]. However, the VLP expression in S. cerevisiae was at low levels of approximately 0.25 mg per liter yeast culture [14] and thus may hinder further product development. In the present study, we evaluated the expression of EV71 VLPs in another yeast species, Pichia pastoris, and determined the protective efficacy of these VLPs in mouse models of EV71 infections. Our results demonstrated that EV71 VLPs could be produced in P. pastoris at high levels. Moreover, P. pastoris-produced VLPs were found to elicit protective immunity against both i.p. and oral viral challenges in mice. 2. Materials and methods 2.1. Cells and viruses
∗ Corresponding author. Tel.:+86 21 54923067; fax: +86 21 54923044. E-mail addresses: [email protected], [email protected] (Z. Huang). http://dx.doi.org/10.1016/j.vaccine.2015.03.034 0264-410X/© 2015 Elsevier Ltd. All rights reserved.
RD and Vero cells were grown as described previously [15]. PichiaPinkTM yeast strains (Invitrogen, USA) were grown according
2336
C. Zhang et al. / Vaccine 33 (2015) 2335–2341
to the manufacturer’s instruction. EV71 strains used in this study, including EV71/BrCr, EV71/G081, EV71/G082, EV71/FY091, EV71/FY09-2 and EV71/SZ98, were described previously [13]. A mouse adapted EV71 strain termed EV71/MAV-W was prepared as follows: the parental EV71/G082 virus was subjected to four rounds of in vivo adaption in one-day-old ICR mice followed by in vitro recovery in Vero cells. Then, the resultant virus was propagated once in Vero cells, yielding the EV71/MAV-W virus stock. Two CA16 strains, CA16/SZ05 and CA16/G08, were described previously [16]. Virus titers were determined as previously described [15,17]. 2.2. Antibodies and peptide Rabbit polyclonal antibodies for detection of EV71 capsid subunit proteins were described previously [18]. Rabbit polyclonal antibody against EV71 VLP was generated in house from a rabbit immunized with insect cell-derived EV71 VLP. SP70 peptide (-YPTFGEHKQEKDLEY-) was synthesized by GL Biochem (Shanghai, China). 2.3. Vector construction and Pichia pastoris transformation The P1 gene of the EV71/G082 strain was codon-optimized, synthesized, and cloned into the backbone vector pPink-HC (Invitrogen), resulting in plasmid YE001. Similarly, the 3CD gene was engineered into pPink-HC to make plasmid YE002. The plasmid YE003 was generated by inserting the 3CD expression cassette into YE001 from the Bgl II site. For P. pastoris transformation, plasmids were linearized by Afl II digestion and then used to transform PichiaPinkTM Strain 1 (Invitrogen) by electroporation. The P. pastoris transformation and subsequent selection of transformants were performed according to the manufacturer’s instruction.
at 37 ◦ C for 1 h. Color development and absorbance measurement were carried out as described previously [11]. Western blot assays were performed as described previously [11].
2.6. Preparation of VLP and control antigen for immunization To prepare the VLP antigen, the selected yeast strains were grown and then induced for antigen expression with methanol according to the manufacturer’s instruction (Invitrogen). The induced yeast cells were pelleted and suspended in 0.15 M phosphate buffered saline (PBS) buffer. The cells were then lysed using a high pressure cell disrupter (JNBIO, China) at 1800 Bar, and the resultant lysates were clarified by centrifugation at 12,000 × g for 15 min. Then, NaCl and PEG 8000 were added into the lysates to reach final concentrations of 200 mM and 10% (W/V), respectively, and the mixtures were stirred gently at 4 ◦ C overnight to allow protein precipitation. Next day, the mixtures were subjected to centrifugation at 12,000 × g for 15 min, and the resultant pellets were resuspended in PBS buffer, followed by centrifugation at 12,000 × g for 15 min to remove insoluble fractions. The clarified supernatants were then subjected to sucrose cushion and sucrose gradient ultracentrifugation as previously described [11]. The final VLP preparation was quantified by ELISA as described above. The empty vector pPink-HC transformed yeast cells were processed in an identical manner to generate the negative control antigen for immunization experiments.
2.7. Electron microscopy Electron microscopy was performed as described previously [11].
2.4. Screening of P. pastoris transformants for VLP expression To identify high-expression P. pastoris strains, small-scale expression experiments were performed according to the manufacturer’s instruction (Invitrogen). Briefly, transformed yeast colonies were individually inoculated into 5 ml BMGY medium and grown with shaking at 30 ◦ C for 24 h. The yeast cells were pelleted by centrifugation and then resuspended in 1 ml BMMY containing 0.5% methanol followed by culture at 30 ◦ C for 48 h. At last, yeast cells from each culture were harvested, resuspended in 150 l breaking buffer, mixed with equal volume of acid-washed glass beads (Sigma), and then subjected to repeated vortex. The broken cell/bead mixtures were clarified by centrifugation at 12,000 rpm for 10 min, and the resulting supernatants were analyzed for VLP expression by ELISA and Western blotting as described below. Total soluble protein (TSP) in the supernatants was measured by Bradford assay. 2.5. ELISA and Western blot assays For antigen quantification, ELISA was performed with insect cell-derived purified EV71 VLPs [13] as the reference standard. Briefly, wells of 96-well microtiter plates were coated with 50 l of rabbit anti-EV71 VLP sera (diluted 1:10,000 in PBS) at 4 ◦ C overnight; then the plates were washed three times with PBST buffer after each of the following steps. Consecutively, 200 l/well of 5% milk in PBST was added for blocking and incubated at 37 ◦ C for 1 h; 50 l/well of the protein lysates or VLP reference standard serially diluted in PBST plus 1% milk was added and incubated at 37 ◦ C for 2 h; 50 l/well of anti-EV71 monoclonal antibody D5 [15] was added and incubated at 37 ◦ C for 1 h; then 50 l/well of HRP-Conjugated anti-mouse IgG (Sigma) was added and incubated
2.8. Mouse immunization and virus challenge The animal studies were approved by the Institutional Animal Care and Use Committee at the Institut Pasteur of Shanghai. ICR mice used in this study were purchased from Shanghai Laboratory Anima Center (SLAC). Prior to immunization, antigens were mixed with Alhydrogel® (Invivogen, USA) adjuvant to make the experimental vaccines. A single injection dose contained 1 g antigen (EV71 VLPs or similarly prepared negative control antigen) and 0.5 mg aluminum hydroxide in a final volume of 100 l. This vaccine dosage was chosen because it was previously shown that two doses (1 g/dose) of insect cell-expressed EV71 VLPs were sufficient to induce high-titer neutralizing antibody response in mice [13]. Groups of 6 female ICR mice (6–8 week) were injected intraperitoneally (i.p.) with the experimental vaccines at week 0 and 2. Blood samples were collected at week 4 for antibody measurement. For passive immunization/protection experiments, groups of female ICR mice were immunized with VLPs or control antigen as described above and allowed to mate two weeks after the last immunization. In the i.p. challenge assay, the 5-day-old neonatal mice born to immunized dams were i.p. injected with 4.2 × 104 TCID50 of EV71/MAV-W. In the oral challenge assay, the 1-dayold neonatal mice born to immunized dams were orally inoculated with 1.26 × 105 TCID50 of EV71/MAV-W. The EV71/MAV-W doses used in the challenge assays were determined in our preliminary studies (data not shown). The challenged mice were monitored daily for survival and clinical score for a period of 14 days. Clinical scores were graded as follows: 0, healthy; 1, reduced mobility; 2, limb weakness; 3, paralysis; 4, death.
C. Zhang et al. / Vaccine 33 (2015) 2335–2341
2.9. Antibody measurement and neutralization assay The virus-specific antibodies in immunized mouse sera were measured by ELISA assay with 200 ng/well of EV71 SP70 peptide as the coating antigen as previously described [13]. Neutralization assays were performed as previously described [11].
3. Results 3.1. High-yield expression of EV71 VLPs in Pichia pastoris Three vectors were constructed for the expression of EV71 VLPs in Pichia pastoris: YE001 and YE002 contained a single P1 and a single 3CD expression cassette, respectively, whereas YE003 possessed both P1 and 3CD cassettes (Fig. 1A). We first co-transformed the PichiaPinkTM yeast with both YE001 and YE002. A yeast clone transformed with the empty vector pPink-HC served as the negative control. EV71 antigen in the lysates of the resulting transformants was quantified by ELISA and total soluble protein (TSP) was measured by Bradford assay. For screening, relative EV71 antigen level for a given sample was expressed as a percentage of TSP. As shown in Fig. 1B, no EV71 protein was detected for the control sample (empty vector-transformed yeast); whereas 2 out of 15 clones produced EV71 antigen at levels above 0.5% TSP, with the highest one being 1.45%. We next transformed the same parent yeast strain with YE003. Screening of 15 YE003-transformed clones showed that most of them produced EV71 antigen at levels above 0.5% TSP and the highest expression reached 4.9% (Fig. 1B). These data demonstrate that YE003 is a better vector than the YE001/YE002 combination at achieving high EV71 antigen expression levels. Cleavage of P1 protein into VP0, VP1, and VP3 subunit proteins is required for VLP assembly. To determine whether the expressed P1 protein was processed by 3CD, lysate from the YE003-transformed and the empty vector-transformed yeast clones was subjected to Western blotting with VP1-specific antisera as the detection
2337
antibody and insect cell-derived EV71 VLPs as the reference standard. No specific band was observed for the empty vector transformant; whereas the YE003-transformed yeast produced a ∼35 kDa band as did the insect cell-derived EV71 VLPs (Fig. 1C), indicating that the P1 polyprotein was properly processed to yield the capsid subunit proteins. To evaluate VLP assembly, yeast lysates were subjected to sucrose gradient ultracentrifugation. The resulting fractions were analyzed by ELISA. No significant reactivity was observed in the control yeast samples (Fig. 2A). In contrast, the profile of the lysate of YE003-transformed yeast is similar to that of insect cell-derived VLPs, with the peak reactivity detected in the fractions #8 (Fig. 2A), suggesting the formation of yeast-derived VLPs. Moreover, Western blot analysis of the yeast fractions showed that VP0, VP1 and VP3 proteins co-sedimented (Fig. 2B), indicating that the VLPs were made of these three capsid subunit proteins. Furthermore, examination of the yeast-derived VLP preparation by electron microscopy revealed the presence of spherical particles with a diameter of 30 nm (Fig. 2C).
3.2. Induction of high-titer neutralizing antibodies by yeast-derived VLPs For immunization studies, VLP was purified from YE003transformed yeast and the control antigen was prepared in the same manner from an empty vector (pPink-HC) transformant. The yeast-derived VLP showed a banding pattern identical to that of the insect cell-produced VLP reference standard in SDS-PAGE (Fig. 3A). To evaluate their immunogenicity, the yeast-derived VLP or the control antigen was used to i.p. immunize mice at weeks 0 and 2. Serum samples were collected at weeks 2 and 4 and analyzed for EV71-specific antibodies by ELISA with the SP70 peptide as the capture antigen. As shown in Fig. 3B, 4 of 6 mice in the VLP group became positive for EV71-specific antibody after the first dose and all of them were seroconverted after the booster injection;
Fig. 1. Co-expression of P1 and 3CD of EV71 in Pichia pastoris. (A) Diagrams of the constructs used in this study. TRP2-L and TRP2-R, the up- and down-stream parts of the TRP region; PAOX1 , AOX1 promoter; CYC1 TT, CYC1 transcription termination region; ADE2, expression cassette encoding phosphoribosylaminoimidazole carboxylase, used as the selection marker. (B) EV71 antigen expression levels in different yeast clones. The lysates from different clones were measured for EV71 antigen by ELISA and for total soluble protein (TSP) by Bradford assay. Lysate from an empty vector-transformed yeast clone was used as the negative control (ctr). Data are mean ± SD of triplicate wells. (C) Western blot analysis with an anti-VP1 polyclonal antibody. Lane M, marker; lane 1, empty vector-transformed yeast; lane 2, YE003-transformed yeast; lane 3, insect cell-produced EV71 VLPs.
2338
C. Zhang et al. / Vaccine 33 (2015) 2335–2341
Fig. 2. Characterization of VLP assembly. (A–B) Characterization of yeast-derived VLPs by sucrose gradient analyses. Yeast-derived antigens and insect cell-produced VLP were subjected to sucrose gradient ultracentrifugation. Twelve fractions were taken from top to bottom for each sample. The gradient fractions were then analyzed by (A) ELISA and (B) Western blotting with anti-VP0, anti-VP1, or anti-VP3 polyclonal antibodies, respectively. (C) Electron microscopy of yeast-derived VLPs. Bar = 100 nm.
in contrast, none of mice injected with the control antigen had detectable anti-EV71 antibodies even after the second immunization. Analysis of the same serum samples by ELISA with insect cell-derived VLP as the capture antigen revealed a reactivity profile (Fig. 3C) similar to that with the SP70 peptide as the capture antigen (Fig. 3B). These results indicate that the VLP, but not the control yeast antigen, elicited EV71-specific antibodies. The ability of individual antisera to inhibit in vitro EV71 infection was assessed by the micro-neutralization assay. The antisera from the control antigen-immunized mice had no neutralization effect even at the lowest dilution (1:16) and was therefore assigned a titer of 8 for computation of geometric mean titer (GMT); in contrast, the anti-VLP sera potently neutralized the homologous strain EV71/G082 with a GMT of 8192 (Fig. 3D). To determine their cross-neutralization ability, the individual antisera collected at week 4 were pooled for each group and
Fig. 3. Yeast-derived VLPs elicited high-titer neutralizing antibodies in mice. (A) SDSPAGE analysis of the antigens used for mouse immunization. Lane M, protein ladder; lane 1, control antigen prepared from empty vector-transformed yeast; lane 2, VLPs prepared from YE003-transformed yeast; lane 3, insect cell-produced VLPs. (B) EV71-specific antibody responses in mice following immunizations. Groups of six ICR mice were injected i.p. at weeks 0 and 2 with 1 g/dose of yeast-derived VLPs or the control antigen. Sera collected at weeks 2 and 4 were analyzed for EV71-specific antibody by ELISA with the SP70 peptide as capture antigen. The antisera were diluted 1:20 and used in ELISA. Each symbol represents a mouse, and the solid line indicates the geometric mean value of the group. (C) EV71-specific antibody responses analyzed by ELISA with insect cell-derived VLPs as the capture antigen. The antisera were diluted 1:1,000 and used in ELISA. Each symbol represents a mouse, and the solid line indicates the geometric mean value of the group. (D) Neutralization titers of antisera determined by the micro-neutralization assay. The antisera from mice received the empty vector control antigen did not show any neutralization activity at 1:16 dilution (the lowest dilution tested) and were therefore assigned a titer of 8 for GMT computation. Each symbol represents a mouse, and the solid line indicates the geometric mean value of the group.
C. Zhang et al. / Vaccine 33 (2015) 2335–2341
Fig. 4. Maternal immunization with VLPs protected newborn mice against i.p. viral challenge. Neonatal mice (5-day-old) born to dams immunized with VLP or the control antigen were i.p. injected with 4.2 × 104 TCID50 of EV71/MAV-W. Then the challenged mice were monitored daily for (A) survival and (B) clinical score for a period of 14 days. Clinical scores were graded as follows: 0, healthy; 1, reduced mobility; 2, limb weakness; 3, paralysis; 4, death. The numbers of mice in each group were indicated in the bracket.
then tested for neutralization against a panel of enteroviruses. As shown in Table 1, the pooled antisera from the control antigenimmunized mice did not exhibit neutralization against any of the virus strains tested; in contrast, the pooled anti-VLP sera potently neutralized all EV71 strains but showed no neutralization effect on the two CA16 strains even at the highest concentration tested (1:32). These results demonstrated that the neutralization activity of the anti-VLP antibodies is EV71-specific and intratypically cross-reactive.
2339
Fig. 5. Maternal immunization with VLPs protected newborn mice against oral infection. Neonatal mice (1-day-old) born to dams immunized with VLP or the control antigen were orally administrated with 1.26 × 105 TCID50 of EV71/MAV-W. Then the inoculated mice were monitored daily for (A) survival and (B) clinical score for a period of 14 days. Clinical scores were graded as follows: 0, healthy; 1, reduced mobility; 2, limb weakness; 3, paralysis; 4, death. The numbers of mice in each group were indicated in the bracket.
3.4. Maternal immunization with yeast-derived VLPs protected neonatal mice against oral infection We further examined whether VLP immunization could protect mice from EV71 oral infection. One-day-old mice born to immunized dams were orally inoculated with the EV71/MAV-W, and then observed daily for survival and clinical signs. As shown in Fig. 5, the control mice gradually developed clinical signs, including reduced mobility and limb weakness and paralysis, and eventually died with a final mortality rate of 41.7%; in contrast, all mice born to the VLP immunized dam survived without notable clinical signs.
3.3. Maternal immunization with yeast-derived VLPs protected neonatal mice against i.p. viral challenge
4. Discussion
To evaluate the protective efficacy, female ICR mice that had been immunized with VLPs or the control antigen were allowed to mate with naïve male mice, and the neonatal mice born to immunized dams were inoculated i.p. with the mouse-adapted EV71 strain, EV71/MAV-W. The challenged mice were subsequently observed on a daily basis for survival and clinical signs for a period of 14 days. The neonatal mice born to the control antigen-immunized dams started to show clinical signs at 3 days post-infection (dpi) and eventually died within 11 dpi (Fig. 4). In contrast, the survival rate for the VLP group was 92.9% and all the surviving mice were free of significant clinical signs (Fig. 4). These results indicate that yeast-derived VLP can protect mice from lethal viral challenge.
Recently, three inactivated EV71 vaccine candidates have shown protective efficacy in phase 3 clinical trials [19–21]. However, the production of inactivated EV71 was at relatively low levels in cultures. For example, it was reported that about 1.25 mg/L of inactivated EV71 could be produced in 40-l pilot-scale production runs [22]. It is therefore challenging to produce large quantity of inactivated EV71 vaccines for worldwide or even nationwide mass immunization of the at-risk populations. As an alternative EV71 vaccine platform, VLPs have been produced in recombinant expression systems and they have been shown to efficiently protect against lethal challenge in animal models [10,13,14]. In particular, S. cerevisiae yeast, which has been successfully used to produce VLP-based HPV vaccines, was tested for EV71 VLP expression. However, the EV71 VLP yield in
2340
C. Zhang et al. / Vaccine 33 (2015) 2335–2341
Table 1 Neutralization capacity of antisera against a panel of EV71 and CA16 strains. Pooled antisera against
Neutralization titer against
EV71/ BrCr Control antigen VLP
EV71/ G081
EV71/ G082
EV71/ FY09-1
EV71/ FY09-2
EV71/ SZ98
EV71/ MAV-W
CA16/ SZ05
CA16/ G08
