Characterization of the enterovirus 71 P1 polyprotein expressed in Pichia pastor as a candidate vaccine

Abstract

Human enterovirus 71 (EV71) plays an important role in hand, foot, and mouth disease (HFMD), which recently caused the death of hundreds of children in the Asia-Pacific region. However, there are no specific treatments available for EV71 infections; thus, a safe and effective vaccine is needed urgently. In this study, we developed an effective and economical method for producing EV71 polyprotein (P1 protein) in Pichia pastoris. Furthermore, we evaluated the potential of P1 protein as a candidate vaccine against EV71 virus. The data revealed that P1 protein induced persistent high cross-neutralization antibodies for different EV71 subtypes, and elicited significant splenocyte proliferation. The high levels of interleukin-10(IL-10) and interferon-gamma (IFN-γ) showed that P1 protein induced Th1 and Th2 immune responses. Interestingly, vaccinating female mice with the P1 protein conferred cross-protection against different EV71 subtypes to their neonatal offspring.Compared with heat-inactivated EV71, the P1 protein elicited improved humoral and cellular immune responses and showed good cross-protection with different EV71 subtypes. Therefore, the EV71-P1 protein produced by P. pastoris is a promising candidate vaccine against EV71.

Introduction

Enterovirus 71 (EV71), first isolated in California, USA in 1969,¹ is the most frequently occurring pathogen that causes hand, foot, and mouth disease (HFMD). Symptoms of HFMD include neurological dysfunction such as aseptic meningitis and poliomyelitis-like paralysis.² EV71 is reportedly responsible for HFMD outbreaks worldwide.^3-5 Moreover, large outbreaks have occurred frequently in Asia since 2007, particularly in mainland China⁶ where more than 1.5 million cases occur annually. Of these, more than 90% of the critical cases and deaths are caused by EV71. Children younger than 5 y old are particularly susceptible to EV71,⁷ and the associated neurological disease is often fatal.⁸ As there are no effective clinical treatments for this virus, EV71 infections are regarded as an important public health problem that can potentially cause serious illness and death in young children.⁹

Enterovirus 71(EV71) is a small, non-enveloped RNA virus of the genus Enterovirus in the Picornaviridae family. EV71 has a single-stranded positive-sense RNA genome of approximately 7500 nucleotides,¹⁰ which has a single open reading frame (ORF) flanked by 5′and 3′untranslated regions (UTRs). The ORF is translated into a large polyprotein that is cleaved into P1, P2, and P3 regions. The P1 region has been expressed and purified in an earlier sdudy.²⁷ It encodes four capsid proteins: VP1, VP2, VP3, and VP4. The P2 and P3 regions encode non-structural proteins (such as proteases 2A, 2B, and 3CD), which are responsible for viral replication and virulence. These proteins assemble spontaneously and form crystalline virus-like particles.¹¹

Since the development of an EV71 vaccine is critical for the prevention and control of HFMD, there are various types of EV71 vaccines, including a peptide vaccine, DNA vaccine, attenuated and inactivated vaccine, and virus-like particles.^12-16 Enterovirus vaccine development also suggests that other inactivated vaccines are feasible; however, considering the financial and industrial requirements, a protein engineering approach may be more feasible.Previous studies have reported that humoral immunity^17,18 and cell immunity^19,20 are crucial for protection against EV71 infection. Therefore, vaccine design has so far focused on the capsid protein,which includes most of the antigenic determinants. The VP1 recombinant protein was reported to induce mixed humoral and cellular immunity in mice.^14,21-24 Recently, it was found that combined peptides in the VP2 protein domain could be effective in protecting against EV71,²⁵ and the VP4 N terminus can induce cross-protection against EV71.²⁶ These studies suggest that P1 protein is an ideal EV71 vaccine candidate.

In this study, we assessed the potential of the P1 protein as a vaccine against EV71 infection. The data revealed that P1 stimulated strong humoral and cellular immune responses, and conferred cross-protection against different EV71 subtypes. These suggest that the EV71-P1 protein produced by P. pastoris is a potential EV71 vaccine.

Materials and Methods

Cell lines and virus strain 71

RD cells (rhabdomyosarcoma) and C4 strain of EV71 (FY0805, HQ882182) were provided by the Yelinbai (Wuhan University, China), and the A strain of EV71(BrCr) was purchased from the China Center for Type Culture Collection (Wuhan, China).Both EV71 viruses were propagated in RD using Dulbecco’s modified Eagle medium (DMEM, Sigma) containing 5% fetal bovine serum (FBS,Gibco)at 37 °C and then were collected from the supernatant after 5 d.

Animal care

All animals were purchased from the Experimental Animal Center of the HuBeiProvince Disease Control Center (Wuhan, China) and housed at the Experimental Animal Center of JianghanUniversity(Wuhan, China).

P1 protein expression and purification

The EV71-P1 gene sequence was modified from the EV71 C4-subtype strain (GU196833). The transmembrane domain in the P1 protein was predicted by the softwares including TMpred, TMHMM and DNASTART. After deletion of the transmembrane domain, the preferential condon optimization of p.pastoris was processed then synthesized into pPIC9k vector by Shanghai Generay biotech Company. EV71-P1- pPIC9k was transferred to the P. pastoris strain GS115, and then the EV71-P1 protein was expressed and secreted into the medium. The soluble P1 protein was purified by DEAE-SFF column chromatography.²⁷

Immunization and serum sample collection

Six-week-old rabbits were used in the immunization experiments. For use as vaccine, the purified EV71 C4 subtype was inactivated by heating at 56^◦C for 30 min.Each rabbit was immunized with the purified P1 protein and heat-inactivated EV71 virus, respectively. All samples were diluted in PBS and mixed with complete Freund’s adjuvant (for primary injection; Sigma) or incomplete Freund’s adjuvant (for booster injection; Sigma) at a volumetric ratio of 1:1. Each rabbit received the same dose of booster injection after 15 and 28 d. Blood samples were collected from each rabbit every week after injection.

Total anti-EV71 IgG assays

The total anti-EV71 IgG in the rabbit serum samples was determined by performing an enzyme-linked immune sorbent assay (ELISA) using the heat-inactivated C4 subtype of EV71 as the coating antigen. BSA-blocked plates were incubated overnight for different times with diluted rabbit sera at4 °C. The plates were then stained with tetramethylbenzidine (Boster, AR1104) and measured at OD₄₅₀nm.

Neutralization assay with different EV71 subtypes

The neutralization titers were determined based on a TCID50 reduction assay using RD cells. After heat-inactivation at 56 °C for 30 min, 50 µL of 2-fold serially diluted rabbit sera were mixed with an equal volume of 100 TCID50 EV71 in a 96-well plate, and incubated at 37 °C for 1 h. Then, 1.0 × 10⁴ RD cells in 100µL of DMEM with 10% FBS were added to the mixture, and the cytopathic effects (CPE) were measured after 5 d of culture. The neutralization titer is the highest serum dilution that revealed no CPE. The experiment was repeated five times, and the average neutralization titer was recorded.

Different subtypes of lethal EV71 challenge

Considering EV71 infection causes no apparent clinical symptoms in adult BALB/c mice, viral challenge was performed using newborn mice. Eight-week-old female BALB/c mice received three injections of P1 protein, heat-inactivated EV71, or PBS at 8,10, and 12 wk, and then the mice were mated. The neonatal mice were challenged with low (10 LD50) or high (50000 LD50) doses of C (C4) and A (BrCr) stains of the EV71 virus (100µL/mouse) intraperitoneally (i.p.). The mice were observed daily for mortality until 4 wk after infection.

In vivo protection against lethal EV71 infection

Sera were collected from rabbits (2 µg P1 protein group and 2 µg heat-inactivated group) with the top titers. The heat-inactivated (56 °C, 30 min) rabbit sera and live EV71 viruses (10–50 LD50 per mouse) were incubated at 37 °C for 1 h, and then the mixture was injected (i.p) into neonatal BALB/c mice. Mice were observed daily for 4 wk after infection.

Spleen lymphocyte proliferation and cytokine production

Two-week-old female BALB/c mice were injected with P1 protein, heat-inactivated EV71, or PBS at 2, 6, and 8 wk, as described above. The spleens were aseptically isolated from male mice (n = 5 each group) at day 14 after the third immunization, and the lymphocytes were separated using a mouse lymphocyte separation kit. Lymphocytes were seeded at a density of 1 × 10⁶ cells per well in a 96-well plate, and stimulated with P1 protein, heat-inactivated EV71, ConA (Sigma, USA) or PBS (negative control) for 72 h, and then MTT assay was performed.To analyze cytokine production, the sera of mice were collected after the third injection, and the levels of IFN-γ and IL-10 were measured using commercially available ELISA kits..

Statistical analysis

All data was analyzed using independent t tests and expressed as the mean of 3−5 independent experiments.

Results

Expression and characterization of P1 protein

The expressed and purified P1 protein was characterized by SDS-PAGE and western blotting analysis.Analysis revealed that the purified P1 protein corresponded to the P1 protein that we reported previously (Fig. 1).²⁷ The yield of P1 protein was about 25 mg/L in the shaker flask with a yeast concentration of 180 g/L, and the optimization of conditions for large-scale production of P1 protein is in progress.

Figure 1. The identify of the EV71-P1 protein. (A) The SDS-PAGE analysis of P1 protein. The concentration of separatin gel was 10% for SDS-PAGE. M, market; 1, the purified P1 protein. (B) western blot analysis of P1 protein. The concentration of separating gel was 10% for western blot. 1, the purified P1 protein; 2, the pPIC9k-GS115 supernatant.

Humoral response elicited by P1 protein in rabbits

To determine the antibody titers, serum samples were collected every week after injection. The anti-EV71 antibody results (Fig. 2) showed both P1 protein and heat-inactivated EV71 viruses elicited anti-EV71 antibodies. Antibodies in the P1 protein groups (2µg/rabbit and 20µg/rabbit, respectively) were detected at week 4, and reached their maximum (1:6000) at week 14. Additionally, high-level titers were maintained for at least 24 wk. Compared with P1 protein, heat-inactivated EV71 induced lower titers. Figure 2 also revealed that the immune efficiency of heat-inactivated EV71 viruses was proportional to the dose, whereas no obvious difference was observed in P1 groups with different doses. However, the trend of the anti-EV71 antibody titers was similar in all experimental groups.

Figure 2. The mean of anti-EV71 antibody titers. Serum samples were collected on the days indicated on the x-axis. Group (2 μg P1 protein) was immunized with 2 µgP1 protein; group (20 μg P1 protein) was immunized with 20 µg P1 protein; group (2 μg inactivated) was immunized with 2 µg. Heat-inactivated EV71; group (20 μg inactivated) was immunized with 20 µg. Heat- inactivated EV71 alone respectively on day 0 and day 14 and day 28 (indicated by *)

Neutralization assays were used to determine whether the antibodies neutralized live EV71. Data revealed that the serum of rabbits injected with P1 protein could neutralize EV71 and the development neutralization titers obtained showed a similar trend to that induced by inactivated EV71 (Fig. 3). The highest neutralization titer was detected at week 14, and persisted for at least 24 wk, as well as the neutralization ability against A and C subtypes were comparable. Compared with P1 protein, the inactivated EV71 elicited significantly lower titers for A and C type EV71.

Figure 3. The neutralizing ability assay of anti-EV71 antibody in vitro. The sera were collected from the immunized rabbits at different time points were serially diluted, mixed with A(BrCr) (Fig. 3A) and C(C4) (Fig. 3B) subtype EV71 virus and then were used to infect RD cells. After 5 d, CPE was observed and the maximun dilution that gave no CPE was determined as the neutralization titer. The rabbits were injected with 2 µg P1 protein and heat- inactivated EV71 respectively on day 0, day 14, and day 28 (indicated by *). (A) The neutralizing ability assay of anti-EV71 antibody against A (BrCr) EV71 in vitro. (B) The neutralizing ability assay of anti-EV71 antibody against C(C4) EV71 in vitro.

Maternal P1 antibodies protected newborn mice against lethal EV71 subtypes A and C

The protection conferred by the maternal anti-P1 antibodies was further evaluated using an immunization study. One-day-old neonatal mice were immunized with lethal doses of A and C subtype EV71. These animals were chosen because EV71 infections only develop clinical symptoms in one-day old BALB/c mice. Neonatal mice infected with EV71 showed obvious nervous system symptoms such as listlessness, lethargy, limb paralysis, and even death. However, the protection efficiency of both the P1 protein and the inactivated EV71 groups was 100% at 5000 LD50 (Table 1A and Table 1B) against different subtypes. Increasing the EV71 dose to 10000 LD50, the protection efficiency retained 100% protection against the C subtype in all vaccine groups. In contrast, the protection against a 10000 LD50 dose of the EV71 A subtype was reduced to 90% in the P1 protein group, and 80% in the inactivated EV71 group. This suggests that antibodies transmitted from pregnant mice protect suckling mice from EV71 infections efficiently. Furthermore, the similar cross-protection was conferred by the P1 protein and the inactivated EV71 vaccine.

Table 1A. The protection conferred by the maternal antibodies against the lethal EV71 challenge in suckling mice. The female mice were immunized with the purified P1 protein; heat- inactivated EV71 at weeks 2,4,6 and allowed to conceive after the 3rd injections. The neonatal mice were born and then challenged i.p. with 10-50000 times LD50 dosage of A(BrCr ) or C(C4) subtype EV71 when one day old. The protection conferred by the maternal antibodies against the EV71 A(BrCr) subtype lethal EV71 challenge.

virus dosage	P1group protection	Inactiveted EV71 group
5000LD50	100%(10/10)	100%(11/11)
10000LD50	90%(9/10)	80%(8/10)
25000LD50	50%(6/12)	50%(5/10)
50000LD50	10%(1/10)	9%(1/11)

Table 1B . The protection conferred by the maternal antibodies against the EV71 C(C4) subtype lethal EV71 challenge.

virus dosage	P1group protection	inactiveted EV71 group
5000LD50	100%(12/12)	100%(11/11)
10000LD50	100%(10/10)	100%(12/12)
25000LD50	63%(6/11)	60%(6/10)
50000LD50	31%(4/13)	33%(4/12)

Protection of the anti-P1 serum against lethal EV71 challenge in suckling mice

The protection efficiency of the anti-P1 serum was then evaluated in a passive immunization study (Table 2A and Table 2B). Ninety and 100% survival rates were observed after 4 wk in the mice that received a mixture of live EV71 and anti-inactivated EV71 serum, and a mixture of live EV71 and anti-P1 serum, respectively. The results showed both anti-P1 serum and anti-inactivated EV71 serum could neutralize the EV71 and increase the survival rate of EV71-infected suckling mice significantly compared with rabbit serum before vaccination.

Table 2A. The protection conferred by antisera against lethal EV71 challenge in suckling mice. The sera were collected from rabbits (2 μg P1 group and heat-inactivated group) respectly. 100 uL of serum with top titer could neutralize the 10 LD50 virus base on the LD50 and neutralize ability assay results. The heat-inactivated (56oC, 30 min) rabbits sera and live EV71 virus(10 LD50,50 LD50) were incubated at 37ºC for 1 hour and the virus-sera mixture was injected intraperitoneally (i.p.) into the neonatal BALB/c mice. Mice were observed daily for mortality until 4 weeks after infection. The protection conferred by 100 ul antisera against the 10 LD50 live A and C subtype EV71 virus.

sera\virus	A(BrCr)	C(C4)
P1group protection	100%(11/11)	100%(12/12)
Inactiveted EV71 group	90%(9/10)	100%(11/11)
Before vaccinated	0%(0/11)	0%(0/9)

Table 2B. The protection conferred by 100 ul antisera against the 50 LD50 live A and C subtype EV71 virus.

sera\virus	A(BrCr)	C(C4)
P1group protection	50%(5/10)	60%(6/10)
Inactiveted EV71 group	30%(3/10)	50%(6/12)
Before vaccinated	0%(0/11)	0%(0/10)

Cellular immune responses elicited by immunization with P1 protein

The effect of P1 protein immunization on T cell responses was investigated using splenocyte proliferation and Th1 and Th2 cytokine assays. Analysis of splenocyte proliferation revealed a statistically significant difference in the stimulation index (SI) values in Con A-stimulated splenocytes from the P1 and heat-inactivated EV71 groups compared with the PBS group (Fig. 4). The SI for the P1 protein and the heat-inactivated EV71 groups also shows a significant increase in cell proliferation compared with the PBS group (P < 0.01).

Figure 4. Proliferation of splenocytes derived from immunized mice. Spleen lymphocutes from each group (P1 protein, heat-inactivate EV71 or PBS injected group) were isolated at 2 wk after the last immunization. The proliferation of lymphocytes was detected after stimulation with P1 protein, heat-inactivated EV71, PBS or Con A respectly.

Next, ELISA was used to measure the serum concentrations of IFN-γ and IL-10 in each group. The concentrations of IFN-γ and IL-10 in mice immunized with P1 protein or inactivated EV71 were increased significantly compared with negative control mice (P < 0.01, Fig. 5). In addition, the IFN-γ concentrations in the P1 group were significantly higher than in the inactivated EV71 group (P < 0.05). The results showed that both P1 protein and heat-inactivated EV71 were able to elicit both Th1 (IFN-γ) and Th2 (IL-10) responses although the extent did vary between the two antigens, possibly due to the type of immunogens. These suggest the enhanced T cell proliferation and activation is observed in mice immunized with P1 protein or inactivated EV71.

Figure 5. The production of cytokines by P1 protein and heat-inactivated EV71. The serum collected from different groups of immunized mice and cytokine (IFN-y, IL-10) concentrations in the serum were measured by ELISA. Each sample dilution was measured in triplicate.

Discussion

Several global outbreaks of EV71 infection have been reported, and active EV71 has recently spread in the Asia-Pacific region,^28-31 particularly in mainland China.^32-34 However, there are no available specific treatments for EV71 infections, and so a safe and effective vaccine is needed urgently. Recently, several new candidate EV71 vaccines were developed and provoked good immune responses against EV71 challenge in animals.^{16,22,23,25,34-36} However, given the potential safety, industrial and economic limitations of these vaccines, a safer, low priced and more effective EV71 vaccines need to be developed.³⁷

In this study, we produced the P1 protein in P. pastoris and evaluated its effects against EV71 infection by comparing the immune responses elicited by the purified P1 protein with those elicited by the heat-inactivated EV71 virus. Immunization with P1 protein induced a strong humoral immune response by stimulating persistent anti-EV71 antibodies with high titers (Fig. 2 and Fig. 3).In addition to humoral responses, EV71-P1 protein immunization also induced efficient cellular immune responses (Fig. 4).The high IFN-γ and IL-10 concentrations in the P1 protein group (Fig. 5) also suggest a mixed Th1/Th2 immune response was activated. Importantly, immunizing female mice with P1 protein conferred protection to their newborn offspring against infection with the lethal EV71 (Table 1A and 1B). The anti-P1 antibodies cross-reacted with the A and C subtype strains of EV71, suggesting that the P1 protein vaccine provides cross-protection against different subtype strains. The difference between the protection of the anti-P1 serum (10 LD50, Table 2A and 2B) and the protection conferred from the vaccinating female (10000 LD50, Table 1A and Table 1B) suggests that the antibodies (SIgA) in the milk contributed to the protection against EV71 challenge in the offspring, which agreed with the previous reports.²² It is understood that the protection in human babies conferred by breastfeeding was related to the levels of IgA(secretory IgA, sIgA) class enterovirus antibodies in breast milk, which seemed to be more important than transplacentally acquired antibodies.³⁸ In this study, the results of lethal EV71 challenge also showed that the mice pups can be protected from EV71 infection by it’s mother, if the female mouse has been injected with the P1 protein or an inactivated EV71 vaccine. As safety is the most important factor in the development of a vaccine for lactating and pregnant women, we conclude that the safer P1 protein vaccine is preferable to the inactivated virus vaccine.

Results of antibody titer and neutralizing experiments show that EV71-antibody production in the P1 protein group started one week later than the inactivated EV71 group. However, the level and duration of production in the P1 group was stronger compared with the inactivated EV71 group. We hypothesize that this may be because the immune response was related to the antigen glycosylation,³⁹ as was reported for HCV E2 and HIVGP120.^40,41 Glycosylation can also delay the degradation of the antigen, which could explain the delayed and persistent production of EV71- antibodies in the P1 group compared with the heat-inactivated EV71 group. Based on sequence analysis using software NetOGlyc 4.0 and NetNGlyc 1.0, there was 10 glycosylation sites in the P1 protein, and all of these were retained in our modified P1 protein. The molecular weight of the modified P1 protein was ~100KDa, and10 glycosylation sites were likely responsible for ~30 KDa. Therefore, sufficient glycosylation of P. pastoris-expressed EV71-P1 protein plays an important role in the immune responses.

To date, heat-inactivated EV71 viruses are the only potential vaccine tested in clinical studies. There are currently reports about progress in the development of inactivated EV71 vaccines from researches in Mainland China, Taiwan and Singapore. Phase I clinical trials were completed in Taiwan and Singapore, and Phase III trials for the vaccines were also completed in Mainland China. The clinical trial results showed that the inactivated vaccines have good immunogenicity^42-48; however, there are other factors should be taken into account for a competitive vaccine such as safety, cost, and industrialization.⁴⁸ In our study, the P1 protein elicited improved humoral and cellular immune responses. This suggests that the modified P1 protein retains its natural structure and normal glycosylation of the EV71 capsid proteins. Importantly, the EV71-P1 engineering protein is generally non-infectious and safer than the inactivated viral vaccine. In addition, the low cost of engineering protein vaccines is suitable for use in less developed areas, some of which are high-risk regions for EV71 infection. Further, a protein engineering approach may be more feasible for the industrialization. P. pastoris is a powerful expression system that allows the production of functional recombinant proteins and combines the advantages of high-level expression, low cost, and the capacity for post-translational modification of eukaryotes.⁵⁰ Recombinant glycosylated proteins produced from P. pastoris are ideal medical candidates,⁵¹ and our study suggests that the P1 protein vaccine from P. pastoris is a promising EV71candidate.

Conclusion

We successfully expressed EV71P1 protein in P. pastoris, which displayed strong immunogenicity. The recombinant P1 protein vaccine elicited improved humoral and cellular immune responses compared with the heat-inactivated EV71 virus vaccine, and showed good cross-protection against different EV71 subtypes. Moreover, the engineered protein had improved properties than the inactivated EV71 virus vaccine in terms of safety, cost, and industrial purpose. Therefore, the EV71 P1 protein expressed by P. pastoris is a promising candidate EV71 vaccine for industrial application. Further studies on the optimization of large-scale P1 protein production are in progress

10.4161/hv.29123

Disclosure of Potential Conflicts of Interest

The Author states he has no conflict of interest

Acknowledgments

The authors thank Dr Gianpaolo Dentico for English polish in this paper. The authors also acknowledge the financial support from the Key Project of Science and Technology Department of Wuhan (201160823258), and the Key Project of the Educational Bureau of Wuhan (2010010), and the State Key Laboratory program of Viral Genetic Engineering (2010KF10),and the Key Project of Educational Department of Hubei (D20114501).