Purification and assembling a fused capsid protein as an enterovirus 71 vaccine candidate from inclusion bodies to pentamer-based nanoparticles

Biochemical Engineering Journal 117 (2017) 139–146

Contents lists available at ScienceDirect

Biochemical Engineering Journal journal homepage: www.elsevier.com/locate/bej

Regular article

Purification and assembling a fused capsid protein as an enterovirus 71 vaccine candidate from inclusion bodies to pentamer-based nanoparticles Ling Xue a,c,1 , Jiangning Liu b,1 , Qi Wang a,c , Chun Zhang a , Longfu Xu a,c , Jian Luo a , Jian Wang d , Chuan Qin b,∗ , Yongdong Liu a,∗ , Zhiguo Su a a

National Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing, 100190, PR China Institute of Laboratory Animal Science, Chinese Academy of Medical Sciences, Beijing, 100021, PR China c University of Chinese Academy of Sciences, Beijing, 100049, PR China d National Vaccine & Serum Institute Co. Ltd, Beijing, 100024, PR China b

a r t i c l e

i n f o

Article history: Received 20 May 2016 Received in revised form 1 October 2016 Accepted 6 October 2016 Available online 7 October 2016 Keywords: Enterovirus 71 (EV71) Virus-like particles (VLPs) Assembling Fused antigen Inclusion bodies

a b s t r a c t An efficient preparation process for a novel Enterovirus 71 (EV71) vaccine was developed in this paper, which is a fused antigen by connecting the truncated capsid proteins of VP1, VP2 and VP3 into one molecule through flexible peptide linkers and expressed in E. coli as inclusion bodies. The fused protein was purified at denatured state through two-step ion exchange chromatography, with final purity above 95%, host cell proteins below 0.003% and residual DNA less than 50 ng/mL. During the following refolding and assembling process through dilution, the fused antigen precipitated completely, while the precipitation was efficiently inhibited with 2 M urea or 0.5 M arginine as an additive. Size exclusion chromatography analysis indicated the protein formed soluble aggregates with linear or rod-like appearance in transmission electron microscopy. These aggregates transformed into pentamers with a size of 15 nm at pH 8.0 after the additive removing. Moreover, most of the pentamers assembled as sphere-like particulates about 25–40 nm after being induced by calcium chloride. High antigen-specific IgG titer was elicited by immunization with the nanoparticles in mouse model. Splenocytes proliferative responses and cytokines analysis indicated this particulate antigen could induce humoral and cellular immune responses. These results lay foundations for developing the fused antigen as an alternative vaccine against hand-foot-and-mouth disease (HFMD) and for the large-scale production for E.coli-based vaccines. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Human enterovirus 71 (EV71) is one of the main causative pathogens for hand-foot-and-mouth disease (HFMD) which often infects infants and children under 5 years old [1,2]. Since it was first found in 1969, it has caused several HFMD outbreaks in Asian countries and resulted in severe morbidity and mortality [1,3,4]. Although some compounds could be used to cure HFMD, prophylactic vaccine is believed as the most effective way to eradicate this disease [5,6].

∗ Corresponding authors. E-mail addresses: [email protected] (C. Qin), [email protected] (Y. Liu). 1 Both authors contributed equally to this work. http://dx.doi.org/10.1016/j.bej.2016.10.009 1369-703X/© 2016 Elsevier B.V. All rights reserved.

Until now, a number of EV71 vaccine candidates have been developed at different stages. Among these candidates, traditional inactivated EV71 virus was researched most extensively and just received the license to be used in clinic in China. Inactivated virus was reported to exhibit almost 100% protective efficacy, however, low productivity and safety concerns still prompt scientists to develop other vaccine substitutes [7]. Through epitope mapping analysis, some vital antigen epitopes of EV71 were identified, some of them were synthesized to be explored as vaccine candidates, such as SP55 and SP70, but these subunit vaccines often have low immune responses [8,9]. With repeated antigens on the surface but no infective genetic materials inside, recombinant virus-like particles (VLPs) represent a safer and even more effective vaccine platform [10–13]. Several EV71-derived VLPs were prepared by infecting Sf9 insect cells with a recombinant baculovirus vector containing P1 and 3CD genes of EV71 [14,15]. However, insect

140

L. Xue et al. / Biochemical Engineering Journal 117 (2017) 139–146

expression system is usually cost expensive and low productive. Recently, two chimeric VLP vaccines for EV71 were prepared in the bacterial system of E. coli by inserting the core and highly conserved antigen peptides into hepatitis B virus core protein (HBc) as a carrier [16,17]. These chimeric VLP vaccine candidates were reported to be able to induce neutralization antibodies protection for female BALB/c mice, but limited types of antigen epitope might not guarantee an overall protection against virus infection. With the benefit of growing fast and easy control, prokaryotic system is increasingly explored to produce VLPs. Virus capsid proteins could be expressed in prokaryotic cells as organized VLPs such as HBc, or as capsomeres such as L1 protein of human papillomavirus [18]. But more often, for bacterial system, foreign proteins tend to form insoluble aggregates as inclusion bodies that need to be firstly denatured and solubilized before reassembling in vitro. VLPs assembling from denatured protein might be more difficult than from capsomeres of pentamer or hexamer units, but on the other hand, there are some advantages for expressing capsid proteins as inclusion bodies. Firstly, they are easy to be separated from cell debris simply by centrifugation [19]; secondly, they could be purified at denatured state with efficient removal of the nucleic acids and some host cell proteins (HCPs) that may be enveloped inside the particles when assembled in vivo [20]; thirdly, that may be the most important point, they could be produced fast, in large scale and in low cost through the E. coli system [11]. Hence, it is of great importance to investigate the mechanism and develop general techniques of assembling capsid proteins from inclusion bodies for bacterial-based vaccine preparation. EV71 is a kind of non-enveloped, single-stranded RNA virus. Matured virus is an icosahedral sphere which consists of 60 copies of VP1, VP2, VP3 and VP4. VP4 is buried inside the icosahedron with the function of binding nucleic acids, having little impacts on the structure of icosahedral particles [21–23]. VP1, VP2 and VP3 contain most of the epitopes and constitute the virus outer shell through a pentameric assembly unit. Recently, antigen epitopes mapping for EV71 was investigated in detail, and we found that combining parts of the three capsids together could provide effective protections on neonatal mice against virus infection [24]. However, being expressed as inclusion bodies in E.coli and meanwhile missing an efficient production procedure impede further developing this fused capsid protein as a new potential EV71 vaccine candidate. Here, based on the previous study, the fused capsid protein containing the truncated VP1, VP2 and VP3 was constructed and expressed. After extraction, the inclusion bodies were solubilized and purified at denatured state. Then a refolding and assembling procedure was developed through an intermediate step with 2 M urea or 0.5 M arginine to suppress precipitate. Transmission electron microscopy (TEM) demonstrated that this fused protein could assemble into pentamer-derived particulate with a size of 25–40 nm. High antigen-specific IgG titer could be induced by immunization with the particulate antigen in mouse model. For the first time, to our knowledge, the fused truncated EV71 capsid protein was demonstrated to assemble itself into sphere-like nanoparticles which hold great promise to be developed as a new kind of vaccine candidate against HFMD. Furthermore, the robust purification and assembling approaches described here lay a foundation for the large-scale production of E. coli-based vaccines.

2. Materials and methods 2.1. DNA manipulation EV71 FY0805, isolated from Anhui province of China at 2008 (Genbank accession number: HQ882182.1), was used as the template for protein designation. Briefly, the total RNA was extracted

with TRIzol from virus stock and reverse-transcripted into cDNA with random primers. Then, the encoding region of 70–249 (a peptide consists of 180 amino acids located at N-terminal of VP2), 324–443 (a peptide consists of 120 amino acids located at Nterminal of VP3), and 746–876 (a peptide consists of 131 amino acids located at C-terminal of VP1) of the viral polypeptides were cloned respectively with fusion primers shown in Table 1. Subsequently, the DNA sequence encoding the truncated fused protein was amplified from the PCR products of former step by fusion PCR. The target fragment from the PCR product was extracted and digested with NdeI and EcoRI and ligated into the pET30a(+) (Novagen, USA). Then the resulting recombinant plasmid pET30avacA was transformed into E. coli BL21 (DE3).

2.2. Expression and induction of EV71 fused capsid protein The recombinant E.coli BL21 (DE3) was first grown at 37 ◦ C in five shake-flasks each containing 100 mL LB medium supplemented with 100 ␮g/mL kanamycin and then inoculated in a 20 L bioreactor (NBS, USA) when cell density reached OD600 of 5.0–8.0. Cells were continuously grown in 15 L fermentation medium (yeast extract 5 g/L, tryptone 10 g/L, NaCl 10 g/L) supplemented with 100 ␮g/mL kanamycin. EV71 fused capsid protein expression was induced when cell growth reached mid-exponential phase with 1 mM isopropyl-d-thiogalactopyranoside (IPTG). After 4 h induction, cells were harvested by centrifugation at 4,000g for 20 min at 4 ◦ C.

2.3. Inclusion bodies extraction and solubilization The cell pellets were resuspended by Lysis Buffer (20 mM TrisHCl, 1 mM EDTA, pH 8.5) at 1: 10 (w/v). The suspension was disrupted by high pressure homogenization (APV2000, Germany) at 3 runs and the pellets of inclusion bodies were collected by centrifugation at 10,000g for 30 min. The inclusion bodies of the fused capsid protein were washed three times by Lysis Buffer containing 1% Triton X-100, 1 M NaCl and 2 M urea, respectively. For each purification experiment, the inclusion bodies were solubilized and denatured in the denaturant buffer (6 M guanidine chloride, 50 mM dithiotretiol (DTT), 20 mM Tris-HCl, 1 mM EDTA, pH 8.5) at room temperature by continuous shaking using rotary drum for 8 h. After centrifugation at 10,000g for 20 min, the suspension was collected and stored at −70 ◦ C. After each centrifugation, the supernatant and pellets were collected and detected by SDSPAGE.

2.4. Purification of solubilized denatured EV71 fused capsid protein Two-column chromatography of ions exchange was used to purify the denatured fused capsid protein. Before ion exchange chromatography, the denatured protein was firstly exchanged into buffer A (8 M urea, 5 mM DTT, 10 mM PB, 1 mM EDTA, pH 7.0) by a desalting column (XK 200 × 16 mm ID, GE Healthcare) containing 50 mL Sephadex G25 (GE Healthcare, USA) and connecting to a ÄKTA Purifier 100 (GE Healthcare, USA). Then the fused protein in buffer A was loaded onto a column containing 20 mL Poros HQ (ABI, USA) equilibrated by buffer A. The pass-through peak was adjusted to pH 6.5 and applied to another column containing 20 mL Poros HS (ABI, USA) equilibrated by buffer B (8 M urea, 5 mM DTT, 10 mM PB, 1 mM EDTA, pH 6.5). The adsorbed protein was eluted by buffer B containing 0.3 M NaCl through linear gradient elution of 5 CV. Eluted peaks were collected and subjected to SDS-PAGE analysis.

L. Xue et al. / Biochemical Engineering Journal 117 (2017) 139–146

141

Table 1 The primer sequences for the construction of the recombinant plasmid pET30a-vacA. Primers

Sequence (5 →3 )

VP2F VP2R VP3F VP3R VP1F VP1R

GGAATTCCATATGTCCCCATCCGCTGAGGCATG (NdeI) CGACCCTCCGCCTCCGCTACCGCCTCCACCAGAGCCTCCTCCACCGTGTGGGCACACTGTTAACTG GGTGGAGGAGGCTCTGGTGGAGGCGGTAGCGGAGGCGGAGGGTCGGGGTTCCCCACCGAGCTAAAACC CGACCCTCCGCCTCCGCTACCGCCTCCACCAGAGCCTCCTCCACCCATGAAGGTGACTTCCAATG GGTGGAGGAGGCTCTGGTGGAGGCGGTAGCGGAGGCGGAGGGTCGGTCAAGCTGTCAGACCCTCC CGGAATTCTTAAAGAGTGGTGATCGCTGTGC (EcoRI)

2.5. Assembly of virus-like nanoparticles

2.11. Dynamic light scattering (DLS) analysis

The purified fused protein collected from Poros HS was diluted to buffer C (20 mM PB, 1 mM EDTA, pH 8.0) with an additive of 2 M urea or 0.5 M arginine and incubated at 4 ◦ C for 12 h. Then the additive in the diluted sample was removed by a desalt column Sephadex G25 with buffer D (20 mM Tris-HCl, pH 8.0). Then 1 mM CaCl2 was added into the desalted sample to initiate the fused protein to assemble into virus-like nanoparticles.

A Zetasizer Nano ZS90 instrument (Malvern Instruments Ltd., UK) was used to determine the protein particle size. Samples were centrifuged at 12,000g for 15 min, and equilibrated to 25 ◦ C. 5 runs of measurements were performed and the data of the same sample were averaged.

2.6. SDS-PAGE According to the method of Laemmli [25], SDS-PAGE was used to analyze protein expression and purity using a 12% gel, and protein bands were developed by Coomassie brilliant blue R250 staining.

2.7. Size exclusion chromatography (SEC) analysis 500 ␮L diluted protein in 2 M urea or 0.5 M arginine was loaded on the column of Superdex200 (300 × 10 mm ID, GE Healthcare), which was already equilibrated with 20 mM PB, 1 mM EDTA, 0.15 M Na2 SO4 , 2 M urea, pH 8.0, and eluted at 0.5 mL/min. Absorbance was recorded at 280 nm.

2.12. TEM analysis 5 ␮L sample of 100 ␮g/mL was dropped onto glow-discharged and carbon-coated, 230-mesh copper grid and allowed to absorb at room temperature for 1 min. The excess solution was removed by filter paper. Then, the copper grid was washed with ddH2 O, and negatively stained with 2% uranyl acetate for 45 s. Excess dye was removed and dried in air before observation using a HT7700 transmission electron microscopy (Hitachi, Japan). 2.13. Circular dichroism (CD) analysis The CD spectra analysis was carried out by Jasco J-810 Spectropolarimeter (Jasco, Japan), and the path length of the quartz cell was 1 mm. The spectra was corrected by the background of the sample buffer. 2.14. Immune experiment

2.8. Measurement of the remained host DNA The Quant-iTTM PicoGreen dsDNA reagent and Kits (Invitrogen, USA) were used to detect and quantitate the remained DNA level, following the manufacturer’s instructions. The fluorescence emission intensity was measured using a microplate reader (Varioskan flash, Thermo scientific, USA) at excitation wavelength of 480 nm and emission wavelength of 520 nm.

2.9. Measurement of remained HCPs HCPs were measured through the E. coli HCPs ELISA Kit (Cygnus, USA) according to the manufacturer’s instructions.

2.10. Measurement of endotoxins The level of endotoxins was determined by the Limulus amebocyte lysate (LAL) test. The lyophilized lysate (Zhanjiang A & C Biological Ltd., China) with the sensitivity of 0.125 EU/mL was reconstituted using 0.1 mL of endotoxin free water. A 0.1 mL sample was added into the depyrogenated borosilicate glass tubes (10 × 75 mm, Zhanjiang A & C Biological Ltd., China) containing 0.1 mL of reconstituted LAL reagent and the mixture was incubated in water bath of 37 ◦ C for 60 min. A positive reaction was indicated by the formation of a solid gel that did not collapse upon inversion of the tube. A negative result was characterized by the absence of such a gel. The endpoint of the assay is defined as the lowest concentration of endotoxin to yield a positive result.

For immunization, the purified and assembled antigen was formulated with a commercial Alu-Vac 15 adjuvant (Serva, Germany) which contains 15 mg/mL of aluminum hydroxide. Briefly, the antigen was diluted to a concentration of 0.2 mg/mL in a volume of 50 ␮L and then mixed with 10-fold diluted Alu-Vac 15 adjuvant at a volumetric ratio of 1:1 according to the manufacturer’s instructions. The heat inactivated EV71 virus were dissolved as 5.0 × 107 TCID50 /mL in the same solution and formulated as the antigen. Female BALB/c mice were purchased from the Beijing Laboratory Animal Center and bred in a specific-pathogen-free facility. Four groups each with 6 female BALB/c mice of 6–8 weeks were used in the immune experiments. The injected dose was 10 ␮g/mouse and given through intramuscular injection. Three groups were immunized with the fused antigens prepared at pH 8.0 and 9.5, and with the inactivated EV71, respectively. The fourth group was injected with PBS as the control. Booster doses were given on days 14 after the first immunization. Blood samples were collected at days 28 after the first immunization and assayed for antigen-specific antibody titer. 2.15. Antigen-specific antibody titer analysis Indirect ELISA was used to detect EV71 fused capsid-specific antibody titer of the sera. Each well of the 96-well plate was coated with 0.2 ␮g assembled fused antigen overnight at 4 ◦ C and washed with PBST for 3 times. After being blocked with PBST containing 0.5% BSA at 37 ◦ C for 1 h, the plates were washed with PBST for 3 times. Then, 100 ␮L/well diluted sera was added and incubated at 37 ◦ C for 45 min. After 3 times washing, 100 ␮L/well

142

L. Xue et al. / Biochemical Engineering Journal 117 (2017) 139–146

HRP-labeled goat anti mouse IgG (Abcam, UK) at 1:50,000 dilution was added and incubated for 30 min at 37 ◦ C. TMB was used for color development and 2 M H2 SO4 was used to terminate the reaction. Absorbance was determined at 450 nm, with 620 nm as the reference wavelength. 2.16. Splenocytes proliferation assay and cytokines production Splenocytes were harvested from immunized mice 35 days post primary immunization. 5 × 105 cells were inoculated in 96-well plates, stimulated with 2 ␮g EV71 fused antigen, or kept unstimulated as control, and co-cultured at 37 ◦ C in 5% CO2 for 48 h. For blank, RPMI-1640 was substituted for cells. A CCK8 assay was applied to evaluate splenocytes proliferation. Cells were treated with 10 ␮L CCK8 for 2 h. Then, an Infinite M200 microplate spectrophotometer (Tecan, Switzerland) was used to measure the absorbance on 450 nm. Proliferation Index was determined according to the formula: Proliferation Index = (Asample − Ablank )/(Acontrol − Ablank ). For cytokines production analysis, the supernatants of splenocytes were collected after 48 h stimulation. IFN-␥, IL-2, IL-4 and IL-5 production were detected using ELISA assay kits (Ebioscience, USA). 3. Results 3.1. Designing and expressing the EV71 fused capsid protein in E. coli Some vital epitopes for EV71 have been identified and several ones were adopted for developing subunit vaccines, such as SP55 and SP70 [9]. For the fused protein antigen designed in this study, three parts from VP1, VP2 and VP3 were adopted according to our previous study [24]. For detail, it contains the 180 amino acids (P70-249 ) and 120 amino acids (P324-443 ) at the N-terminal of VP2 and VP3, and the 131 amino acids (P746-876 ) at the C-terminal of VP1. Two flexible peptides of (Gly4 Ser)3 were adopted to connect the three parts together to form a fused molecule with a molecular weight of 48 kDa (Fig. 1A). Then a plasmid of pET30a(+)-vacA was constructed and translated into BL 21, with the basic sequences of the fused protein inserted between NdeI and EcoRI. The fused protein was induced by IPTG and about 250 g wet cells were harvested in 15 L culture. After cell disruption and centrifugation, the fused protein was found in the pellets (Fig. 1B), indicating that the fused protein was expressed in the form of inclusion bodies. 3.2. Purification of the fused capsid protein at denatured state After washing three times, about 42 g inclusion bodies were obtained with the purity of 60%, and a lot of HCPs and nucleic acids still remained in the pellets (date not shown). Due to the unique features, virus capsid proteins are prone to be combined with certain HCPs and nucleic acids. High concentration of salts, such as denaturants of urea and guanidine hydrochloride, can break the interactions among proteins and nucleic acids. So, after the inclusion bodies were solubilized, it was firstly purified at denatured state in the presence of 8 M urea. Two-step chromatography of ions exchanges were used to remove the contaminants and the eluted peaks were analyzed by SDS-PAGE. For the first column of Poros HQ, the pass-through peak was collected (Fig. 2A). At the pH value for the anion exchange chromatography, nucleic acids were negative and absorbed on the media, so the fused protein could be greatly separated from host nucleic acids at this step. For the second column of Poros HS, the fused protein was absorbed and mainly

Fig. 1. (A) Schematic representation for the EV71 fused capsid protein construction. (B) SDS-PAGE analysis of the expression of EV71 fused capsid protein. Lane 1–2: before and after induction by IPTG; Lane 3–4: pellets and supernatant after cell disruption; M: molecular weights stand. The band for EV71 protein is marked by a red triangle. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

eluted in the peak 2 by 0.15–0.3 M NaCl (Fig. 2B). SDS-PAGE analysis showed the purity of the target protein in peak 2 was above 95%, with a final yield of 16.3% (Fig. 2C, Table 2). Moreover, when this purification protocol was operated following GMP request, the residual DNA was less than 50 ng/mL, HCPs were below 0.003% and the endotoxins were measured below 5 EU/mL. 3.3. Assembling EV71 fused capsid protein into pentamer-based nanoparticles through an intermediate step To recover the correct structure of the fused protein, dilution was tried to switch on the refolding and assembling simply by decreasing the denaturant concentration. Unfortunately, when the denatured protein was diluted directly into the buffer of 20 mM PB (pH 8.0), precipitate formed immediately and almost no protein remained in the supernatant. However, when the denatured protein was diluted in the buffer containing 2 M urea or 0.5 M arginine, no precipitate could be observed even at the protein concentration up to 1.0 mg/mL (Fig. 3A). When the diluted sample was subjected to SEC analysis of Superdex 200, the fused capsid protein was eluted at the exclusion volume (Fig. 3B) for the sample with 2 M urea, while a much wider peak was observed for the case of 0.5 M arginine addition. No monomer could be detected in both cases. DLS analysis showed the fused capsid protein formed some particles with an average size about 22 nm in the presence of 2 M urea and 20 nm in the presence of 0.5 M arginine (Fig. 3C). In order to find out whether these soluble aggregates were regular particles, the diluted samples were subjected to TEM analysis. As shown in Fig. 3D and E, the protein aggregates have linear or rod-like appearances in the presence of 2 M urea or 0.5 M arginine. Then the additive in the diluted sample was removed by desalting into buffers with different pH values. As showed in Fig. 4, obviously different appearances of the aggregates were observed for different pH values. For pH 4.5 and 9.5, aggregation was intensified but smaller protein spots with a size about 15 nm and a shape of pentamer (Fig. 4B, inserted) were detected at pH 8.0. CD analysis (Fig. 4E) showed the secondary structure for the fused protein was mainly ␣-helix at pH 8.0 and random coils at pH 4.5 and 9.5. Urea

L. Xue et al. / Biochemical Engineering Journal 117 (2017) 139–146

143

Fig. 2. Chromatographic profiles and the purity analysis of peaks collected from each column. (A) Chromatographic profile of Poros HQ, (B) chromatographic profile of Poros HS and (C) SDS-PAGE analysis of the peaks from each column. Lane 1: solubilized inclusion bodies of fusion protein; lane 2: the pass-through peak of Poros HQ; lane 3–6: pass-through peak, Peak 1, Peak 2 and Peak 3 from Poros HS. M: molecular weight standard. The band for EV71 protein is marked by a red triangle. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Table 2 The purification process for fused protein with two-column chromatography. Steps

Total protein (mg)

Fusion protein (mg)

Purity (%)

Recovery (%)

Solubilizing inclusion bodies Poros HQ Poros HS

463.40 232.35 47.61

278.04 162.65 45.23

60 70 95

– 58.5 16.3

2.5 g inclusion bodies were solubilized in the denaturant buffer (6 M guanidine chloride, 50 mM DTT, 20 mM Tris-HCl, 1 mM EDTA, pH 8.5) at the ratio of 1: 10 (w/v).

Fig. 3. (A) The mass yield for the dilution of the denatured fused capsid protein. (B) SEC analysis of the diluted sample through Superdex200 with a buffer containing 2 M urea. The flow rate was 0.5 mL/min and absorbance was recorded at 280 nm. (C) DLS profile, (D) and (E) TEM images of the diluted sample in 2 M urea and 0.5 M arginine. The basic buffer for the dilution is 20 mM PB, 1 mM EDTA, pH 8.0.

or arginine has been widely used to inhibit aggregation and assist protein refolding. However, our results indicated that the presence of urea or arginine might interfere in the capsid protein’s assembling process although it did increase the solubility of the capsid protein. The results also indicated that pH value had great influences on the fused protein conformation and the morphology of the intermediate assembly. To find whether the pentamers could further assemble inito VLPs, 1 mM CaCl2 was added into the desalted sample at pH 8.0. Fig. 4D showed most of the pentamers formed larger round particles with a size about 25–40 nm after inducing by divalent ion.

Although the morphology of the assembly did not show as welldefined VLPs when compared with the real EV71 virus (inserted in Fig. 4D), these results indicated that this fused protein could assemble itself into pentamers and then spherical particles from inclusion bodies.

3.4. Immune responses elicited by the nanoparticles in mouse model Our previous study has demonstrated that the peptides comprised in the fused protein could induce high neutralizing antibody

144

L. Xue et al. / Biochemical Engineering Journal 117 (2017) 139–146

results, the antigen-specific antibody responses, splenocytes proliferation and cytokines production were further analyzed to assess the immune responses for the assembled particles at pH 8.0 and 9.5, as well as compared with the inactivated EV71. The corresponding antigen-immunized sera were assayed for the antigen-specific antibodies with the fused protein as antigen in ELISA. The antigenspecific antibodies from mice immunized with the fused antigen assembled at pH 8.0 was obviously higher than that at pH 9.5; meanwhile the reactivity of serum samples from mice immunized with the inactivated EV71 protein to the fused protein was very low. From the CD spectra and morphology analysis, different conformational and spatial structures were found for the fused antigens prepared at different conditions. Low specific-antigen antibody for the nanoparticles prepared at pH 9.5 assured that most epitopes for the fused capsid protein were conformational ones (Fig. 5A). During the inactivation process, the spatial structure of EV71 virus would be disrupted, and then some antibodies elicited in this case would not bind to the fused antigen assembly, also resulting low specific-antigen antibody response. Splenocytes proliferation experiment showed that the proliferation index for the particulate antigen for pH 8.0 was higher than for pH 9.5, but both were a little lower than for the inactivated virus (Fig. 5B). To make sure whether cellular immunity was induced by the assembled fusion protein, splenocytes proliferative responses in immunized mice were determined by measuring the levels of IFN-␥, IL-2, IL-4 and IL-5. The levels of four cytokines were higher for the assembly at pH 8.0 than pH 9.5 s. Compared with the inactivated EV71 virus group, higher levels of IFN-␥ and IL-2 were predominantly produced in mice immunized with particulate antigen at pH 8.0. Meanwhile, the levels of IL-4 and IL-5 were lower for particulate antigens than for the inactivated EV71 virus (Fig. 5C). Taken together, these results demonstrated immunization with particulates of EV71 fused capsid protein was capable of inducing EV71-specific cellular and humoral immune responses in mouse model. The antigen assembled at pH 8.0 could induce much better immune response than at pH 9.5. Fig. 4. TEM images of the fused capsid protein in different buffers after the additive was removed. (A) pH 4.5; (B) pH 8.0, with a schematic diagram of a pentamer and a magnified image inserted. Particles with typical morphology of pentamer are indicated by red arrows; (C) pH 9.5; (D) pH 8.0 with 1 mM CaCl2 . TEM image of inactivated EV71 was inserted in Fig. 4D as a comparison. The scale bar is 100 nm. (E) CD analysis of the fused capsid protein at different pH after the additive was removed. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

titer and provide effective protections on neonatal mice against EV71 infection [24]. In this study, the fused protein was found to assemble into various morphologies at different solutions. As particle’s morphology might have great influence in the immune

4. Discussion For the EV71 native capsid, the assembly unit of protomer contains more than one kind of viral protein, so the sophisticated host system of baculovirus/insect is usually adopted to express EV71derived VLPs which generally had the full length of the capsid protein [15,26]. Different from the traditional strategy, the vaccine candidate designed in this study contained the main parts of EV71 structural proteins of VP1, VP2 and VP3, including almost all the reported epitopes related to EV71 and successfully expressed by E.coli. As it was expressed as inclusion bodies, developing a purifi-

Fig. 5. (A) Antigen-specific IgG titer of sera from mice immunized with assembled fused antigens at pH 8.0 and 9.5, as well as inactivated EV71, (B) splenocytes proliferation experiment and (C) cytokines analysis with mice vaccinated with assembled fused antigens or inactivated EV71 virus. Four groups of BALB/c mice were intramuscularly injected with PBS, assembled fused antigens and inactivated EV71 virus, respectively, and booster doses were given on days 14 post primary immunization. Sera collected from retro-orbital plexus of the immunized group were detected for antigen-specific IgG titer by indirect ELISA. All the four groups were sacrificed on days 35 and splenocytes were harvested and stimulated by 2 ␮g assembled nanoparticles in vitro for 48 h. Proliferation Index was determined by the formula: Proliferation Index = (Asample − Ablank )/(Acontrol − Ablank ). Cytokines, IFN-␥, IL-2, IL-4 and IL-5, in the culture supernatants were detected by ELISA.

L. Xue et al. / Biochemical Engineering Journal 117 (2017) 139–146

cation, refolding and assembling process is necessary to produce amounts of qualified proteins for overall assessing it as a useful vaccine candidate. Generally, nucleic acids are prone to interact with capsid protein [20]. When VLPs are assembled in vivo, host nucleic acids and HCPs might be enveloped inside the spheres, which would be hardly removed by traditional purification methods. In such cases, disruption the spheres to release the contaminants may be the first step for the purification. Host DNA/RNA impurities could lead to serious clinical consequences. Therefore, the amount of residual host DNA and proteins in the final products should be quantified and cleared during the downstream process for quality control. In this study, purifying EV71 fused capsid protein at denatured state is a reasonable and efficient downstream protocol which could avoid enveloping host contaminants into particles during the following assembling process. Following the GMP requests, low levels of HCPs, residual nucleic acids and endotoxins in final product demonstrated the efficiency of the purification procedure. It was also indicated that high concentration of urea did not affect the removal of host nucleic acids and endotoxins through ion exchange chromatography. However, the physicochemical discrepancies among different proteins might shrink at denatured state compared with the native state, so purification at denatured state might need chromatographic media with higher resolution. With the ability of expression and assembling viral protein as VLPs with high architectural consistency, insect and mammal cells are widely used to produce vaccines of VLPs. Generally, culturing insect cell is very expensive, time-consuming, and low productivity [26,27]. Moreover, ultracentrifugation used for VLPs purification is labour intensive and low efficient. Currently, the rapid and lowcost bacterial system is becoming a prevalent platform for VLPs and some antigens’ production. Moreover, large production of recombinant proteins through E.coli system is easily achieved now. For the EV71 fused antigen in this study, it would take about one week for the production including expression, purification, refolding and assembling in vitro. The yield for the assembled particles is about 60 mg/L, which is much higher than for the insect expression system. Although having the benefits of economic and fast production of large amount of proteins, only limited capsid proteins were expressed with the morphology of virus in E.coli. Most viral proteins were expressed as capsomeres or even more as inclusion bodies. Reassembling in vitro is indispensable meanwhile a bottleneck step for recovering the viral morphology. For the case of inclusion bodies, assembling VLPs starting from denatured protein would face more challenges than from soluble capsomeres because this process was simultaneously entangled with an extra refolding process. Low concentration of urea and arginine are widely used additives to inhibit precipitate during protein refolding. However, the assembly of EV71 fused capsid protein in the additive of urea or arginine only formed soluble irregular aggregates. Intriguingly, these aggregates could transform to structures of pentamers along with the removal of the additive. pH and Ca2+ were found to be the important factors affecting the formation of the pentamers and the following particles [28]. As we know, weak interaction exits among capsid proteins which benefits virus disassembling and reassembling during its proliferation [29,30]. It is putative that due to the same weak interaction among the assembled intermediate of the fused antigen, disordered aggregates could rearrange to the correct form of pentameric assembly unit once the interfering factors disappeared. SEC and DLS are methods conveniently used to detect the aggregates or particle size, but generally only TEM could detect the correct morphology of the assembly. Here, we found that CD analysis was another approach that could be used to reflect whether the assembly formed the correct structure.

145

In spite that VLPs were anticipated for this EV71 fused antigen, it just re-assembled into pentamer-based larger nanoparticles, without well-defined morphology of VLPs. The immune results in mouse model for particles with different appearances varied greatly, indicating that most epitopes for the fused antigen were conformational ones. However, the pentamer-based nanoparticles could induce high antigen-specific antibody titer and stimulate splenocytes proliferation, which were comparable to the inactivated EV71 virus. All the results lay a foundation to develop this fused capsid protein as a novel vaccine candidate for HFMD. Moreover, the general process established in this paper was also valuable for the large-scale production for E.coli-based vaccines. Acknowledgements We thank Dr. Jie Wu and Yufei Xia for the mice immune experiments. This research was financially supported by National Science and Technology Major Project of “National Key Program on Drug Innovation” (Grant No. 2012ZX09101319004 and 2016ZX09101120-006), the National Natural Science Foundation of China (Grant No. 21576267), NN-CAS research fund (Grant No. NN-CAS-2014-2) and Open Funding Project of the National Key Laboratory of Biochemical Engineering (Grant No. 2014KF-05). References [1] S.M. Wang, C.C. Liu, Update of enterovirus 71 infection: epidemiology, pathogenesis and vaccine, Expert Rev. Anti Infect. Ther. 12 (2014) 447–456. [2] W. Xing, Q. Liao, C. Viboud, J. Zhang, J. Sun, J.T. Wu, Z. Chang, F. Liu, V.J. Fang, Y. Zheng, B.J. Cowling, J.K. Varma, J.J. Farrar, G.M. Leung, H. Yu, Hand, foot, and mouth disease in China, 2008–12: an epidemiological study, Lancet Infect. Dis. 14 (2014) 308–318. [3] P.C. McMinn, An overview of the evolution of enterovirus 71 and its clinical and public health significance, FEMS Microbiol. Rev. 26 (2002) 91–107. [4] T. Solomon, P. Lewthwaite, D. Perera, M.J. Cardosa, P. McMinn, M.H. Ooi, Virology, epidemiology, pathogenesis, and control of enterovirus 71, Lancet Infect. Dis. 10 (2010) 778–790. [5] E.J. Bek, P.C. McMinn, Recent advances in research on human enterovirus 71, Future Virol. 5 (2010) 453–468. [6] C.W. Tan, J.K.F. Lai, I.C. Sam, Y.F. Chan, Recent developments in antiviral agents against enterovirus 71 infection, J. Biomed. Sci. 21 (2014). [7] S.L. Zhou, X.L. Ying, X. Han, X.X. Sun, Q. Jin, F. Yang, Characterization of the enterovirus 71 VP1 protein as a vaccine candidate, J. Med. Virol. 87 (2015) 256–262. [8] C.H. Chiu, C. Chu, C.C. He, T.Y. Lin, Protection of neonatal mice from lethal enterovirus 71 infection by maternal immunization with attenuated Salmonella enterica serovar Typhimurium expressing VP1 of enterovirus 71, Microbes Infect. 8 (2006) 1671–1678. [9] D.G.W. Foo, S. Alonso, M.C. Phoon, N.P. Ramachandran, V.T.K. Chow, C.L. Poh, Identification of neutralizing linear epitopes from the VP1 capsid protein of Enterovirus 71 using synthetic peptides, Virus Res. 125 (2007) 61–68. [10] J. Chroboczek, I. Szurgot, E. Szolajska, Virus-like particles as vaccine, Acta Biochim. Pol. 61 (2014) 531–539. [11] L.H.L. Lua, N.K. Connors, F. Sainsbury, Y.P. Chuan, N. Wibowo, A.P.J. Middelberg, Bioengineering virus-like particles as vaccines, Biotechnol. Bioeng. 111 (2014) 425–440. [12] M.R. Schmiedeskamp, D.R. Kockler, Human papillomavirus vaccines, Ann. Pharmacother. 40 (2006) 1344–1352. [13] T. Wu, S.W. Li, J. Zhang, M.H. Ng, N.S. Xia, Q. Zhao, Hepatitis E vaccine development: a 14 year odyssey, Hum. Vaccin Immunother. 8 (2012) 823–827. [14] Z.Q. Ku, X.H. Ye, X.L. Huang, Y.C. Cai, Q.W. Liu, Y. Li, Z.G. Su, Z. Huang, Neutralizing antibodies induced by recombinant virus-like particles of enterovirus 71 genotype C4 inhibit infection at pre- and post-attachment steps, PLoS One 8 (2013) e57601. [15] Y.C. Chung, J.H. Huang, C.W. Lai, H.C. Sheng, S.R. Shih, M.S. Ho, Y.C. Hu, Expression, purification and characterization of enterovirus-71 virus-like particles, World J. Gastroenterol. 12 (2006) 921–927. [16] L.F. Xu, D. He, L.S. Yang, Z.Q. Li, X.Z. Ye, H. Yu, H. zhao, S.X. Li, L.Z. Yuan, H.L. Qian, Y.Q. Que, J.W. Shih, H. Zhu, Y.M. Li, T. Cheng, N.S. Xia, A broadly cross-protective vaccine presenting the neighboring epitopes within the VP1 GH loop and VP2 EF loop of enterovirus 71, Sci. Rep. 5 (2015) 12973. [17] X.H. Ye, Z.Q. Ku, Q.W. Liu, X.L. Wang, J.P. Shi, Y.F. Zhang, L.L. Kong, Y. Cong, Z. Huang, Chimeric virus-like particle vaccines displaying conserved enterovirus 71 epitopes elicit protective neutralizing antibodies in mice through divergent mechanisms, J. Virol. 88 (2014) 72–81. [18] M.L. Li, T.P. Cripe, P.A. Estes, M.K. Lyon, R.C. Rose, R.L. Garcea, Expression of the human papillomavirus type 11 L1 capsid protein in Escherichia coli:

146

[19] [20]

[21]

[22]

[23] [24]

L. Xue et al. / Biochemical Engineering Journal 117 (2017) 139–146 characterization of protein domains involved in DNA binding and capsid assembly, J. Virol. 71 (1997) 2988–2995. S.M. Singh, A.K. Panda, Solubilization and refolding of bacterial inclusion body proteins, J. Biosci. Bioeng. 99 (2005) 303–310. D.I. Lipin, Y.P. Chuan, L.H.L. Lua, A.P.J. Middelberg, Encapsulation of DNA and non-viral protein changes the structure of murine polyomavirus virus-like particles, Arch. Virol 153 (2008) 2027–2039. J.O. Cifuente, H. Lee, J.D. Yoder, K.L. Shingler, M.S. Carnegie, J.L. Yoder, R.E. Ashley, A.M. Makhov, J.F. Conway, S. Hafenstein, Structures of the procapsid and mature virion of enterovirus 71 strain 1095, J. Virol. 87 (2013) 7637–7645. C.C. Liu, M.S. Guo, F.H. Lin, K.N. Hsiao, K.H. Chang, A.H. Chou, Y.C. Wang, Y.C. Chen, C.S. Yang, P.C. Chong, Purification and characterization of enterovirus 71 viral particles produced from vero cells grown in a serum-free microcarrier bioreactor system, PLoS One 6 (2011) e20005. P. Plevka, R. Perera, J. Cardosa, R.J. Kuhn, M.G. Rossmann, Crystal structure of human enterovirus 71, Science 336 (2012) 1274. J.N. Liu, W. Wang, J.Y. Duo, Y. Hao, C.M. Ma, W.B. Li, S.Z. Lin, X.Z. Gao, X.L. Liu, Y.F. Xu, W.-B. Xu, C. Qin, L.F. Zhang, Combined peptides of human enterovirus 71 protect against virus infection in mice, Vaccine 28 (2010) 7444–7451.

[25] U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227 (1970) 680–685. [26] Y.C. Chung, M.S. Ho, J.C. Wu, W.J. Chen, J.H. Huang, S.T. Chou, Y.C. Hu, Immunization with virus-like particles of enterovirus 71 elicits potent immune responses and protects mice against lethal challenge, Vaccine 26 (2008) 1855–1862. [27] S.Y. Lin, H.Y. Chiu, B.L. Chiang, Y.C. Hu, Development of EV71 virus-like particle purification processes, Vaccine 33 (2015) 5966–5973. [28] G.H. Chang, Y.J. Luo, X.Y. Wu, B.Y. Si, L. Lin, Q.Y. Zhu, Monoclonal antibody induced with inactived EV71-Hn2 virus protects mice against lethal EV71-Hn2 virus infection, Virol. J. 7 (2010) 106. [29] E. Gharakhanian, J. Takahashi, J. Clever, H. Kasamatsu, In vitro assay for protein-protein interaction: carboxyl-terminal 40 residues of simian virus 40 structural protein VP3 contain a determinant for interaction with VP1, Proc. Natl. Acad. Sci. U. S. A. 85 (1988) 6607–6611. [30] D.H. Barouch, S.C. Harrison, Interactions among the major and minor coat proteins of polyomavirus, J. Virol. 68 (1994) 3982–3989.