<<

Jpn. J. Infect. Dis., 69, 66–74, 2016

Original Article Phage Display-Derived Cross-Reactive Neutralizing against and A16

Xiao Zhang1†, Chunyun Sun2†, Xiangqian Xiao1,LinPang3, Sisi Shen1, Jie Zhang2, Shan Cen4,BurtonB.Yang5, Yuming Huang3, Wang Sheng1*, and Yi Zeng1 1College of Life Science and Bioengineering, Beijing University of Technology, Beijing; 2Sinocelltech, Cell Engineering Center, Chinese Academy of Medical Science, Beijing; 3Beijing Ditan Hospital, Capital Medical University, Beijing; 4Department of Virology, Institute of Medicinal Biotechnology, Chinese Academy of Medical Science, Beijing, China; and 5Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Canada

SUMMARY: Enterovirus 71 (EV71) and coxsackievirus A16 (CVA16) are members of the Picornaviri- dae family and are considered the main causative agents of hand, foot and mouth disease (HFMD). In recent decades large HFMD outbreaks caused by EV71 and CVA16 have become significant public health concerns in the Asia-Pacific region. Vaccines and antiviral drugs are unavailable to prevent EV71 and CVA16 . In the current study, a chimeric antibody targeting a highly conserved peptide in the EV71 VP4 protein was isolated by using a phage display technique. The antibody showed cross- neutralizing capability against EV71 and CVA16 in vitro. The results suggest that this phage display- derived antibody will have great potential as a broad neutralizing antibody against EV71 and CVA16 after affinity maturation and humanization.

potential for new viral recombinants of EV71 and INTRODUCTION CVA16 to emerge have been documented (13–15). These Enterovirus 71 (EV71) and coxsakievirus A16 findings suggest that both EV71 and CVA16 should be (CVA16) are non-enveloped RNA of the targeted for vaccine and therapeutic development for ef- Picornaviridae family. Their are positive-sense fective control and treatment of HFMD. single-stranded of approximately 7,500 nucleo- Immunization of maternal mice with inactivated tides encoding single, large polyprotein that can be EV71 and CVA16 protects newborn mice from lethal digested by into P1, P2, and P3 regions (1–3). challenge with EV71 and CVA16, respectively, which Subsequent cleavage of the P1 regions by viral indicates that neutralizing are critical in the 3CD yields 4 subunit proteins: VP1, VP2, VP3, protection of neonatal mice against viral infection and VP4. The virions of EV71 and CVA16 are penta- (16,17). Two neutralizing epitopes SP55 (amino acids meric icosahedral particles with diameters of 24–30 nm. 163–177 and SP70 (amino acids 208–222), have been Theseparticlesarecomposedof60copiesofstructural identified in the VP1 of EV71 in mice. These peptide-in- proteins organized into 12 pentamers. VP1, VP2, and duced antibodies cross-neutralize EV71 and protect VP3 are located on the virion surface, and myristoylat- mice from lethal challenges with homologous and heter- ed VP4 is located inside (4–6). ologous EV71 strains (18). An immunoglobulin M EV71 and CVA16 have been identified as the major monoclonal antibody (mAb) targeting a highly con- etiological agents of hand, foot and mouth disease served epitope (amino acids 215–219 in EV71 VP1) con- (HFMD). Viral by EV71 and CVA16 result in fers effective protection against lethal EV71 challenge in as well as severe illnesses, such as aseptic , vivo when administrated at 10 mg/g body weight (19). In poliomyelitis-like paralysis, and fatal brainstem ence- addition, a cross-neutralizing epitope in VP2 (amino phalitis (7–9). Numerous large outbreaks of HFMD acids 141–155) has been identified through overlapped caused by EV71 and CVA16 have been reported in the synthetic peptide mapping. Chimeric particles consist- Asia-Pacific region since 2000 (10–12), and HFMD ing of both defined VP2 peptide and hepatitis B core remains a common infectious illness in children in this (HBcAg) induce cross-protective antibody area. Moreover, EV71 and CVA16 co-infection and the responses against EV71. Immune sera elicited by chi- meric particles provide full in vivo passive protection Received February 4, 2015. Accepted April 13, 2015. against lethal EV71 challenge (20). In a recent study, a J-STAGE Advance Publication June 12, 2015. neutralizing antibody targeting a conformational epi- DOI: 10.7883/yoken.JJID.2015.060 tope was isolated from mice immunized with an inacti- *Corresponding author: Mailing address: College of Life vated EV71-B4 strain, and it passively protected ne- Science and Bioengineering, Beijing University of Technol- onatal mice against lethal EV71 infection. The potential ogy, 100 Pingleyuan, Chaoyang District, Beijing, China. epitope was mapped by using escape mutants (21). Tel: +86-10-67392780, Fax: +86-10-67392780, E-mail: In a previous study, we identified an immunodomi- shengwang@bjut.edu.cn; danielle_zx@163.com nant epitope at the N-terminal of the EV71 VP4 protein †These authors contributed equally to this work. (22). The ``core sequence'' of VP4, which is responsible

66 Neutralizing Antibody against EV71 and CVA16 for immune stimulation, is highly conserved among immunodominant epitope, which is highly conserved CVA16 and EV71 viral strains. Immunization with the among EV71 and CVA16 viral strains, is highlighted in VP4 peptide elicits a broad neutralizing antibody gray (Table 1). Seven days after a boost spleen cells response against various EV71 genotypes (22). In the were harvested to extract RNA with the TRIzol method presentstudy,aphagedisplay system was used to (Invitrogen, Carlsbad, CA, USA). A SuperScript III produce a recombinant mAb targeting this conserved First-Strand Synthesis System (Invitrogen) was used to peptide (designated VP4N20) of EV71 VP4. The neu- synthesize cDNA from total RNA according to the tralizing activity of the recombinant antibody against manufacturer's instructions. Vk,Vl,andVH fragments EV71 and CVA16 was evaluated with in vitro neutrali- were amplified with PCR using Vk,Vl,andVH primary zation assay. primers (Table 2) (23,24). For PCR amplification, each 50-mL aliquot of reaction mixture contained 5 mL10× buffer, 100 ng cDNA, 2.0 U FastStart enzyme, 40 nmol MATERIALS AND METHODS of dNTP mix, and 20 pmol each of the appropriate for- Phage library construction: Six-week-old female ward and reverse primers. PCR cycling conditions con- BALB/c mice were immunized with an intraperitoneal sisted of 959C for 5 min followed by 30 cycles of 959C injection of 5 mg recombinant chimeric particles com- for 1 min, 559C for 45 s, 729C for 2 min, and a final posed of HBcAg and VP4N20 as described previously 10-min extension at 729C. Purified VL (Vk + Vl)and (22). The sequence of VP4N20 is shown in Table 1. The VH PCR products were combined randomly to synthe-

Table 1. Amino acid sequence of VP4N20 peptide conserved in EV71 and CVA16 VP4 proteins

Gly Ser Gln Val Ser Thr Gln Arg Ser Gly Ser His Glu Asn Ser Asn Ser Ala Thr Glu EV71 GSQVSTQRSGSHENSNSATE

Gly Ser Gln Val Ser Thr Gln Arg Ser Gly Ser His Glu Asn Ser Asn Ser Ala Ser Glu CVA16 GSQVSTQRSGSHENSNSASE

Table 2. The sequences of primers for the scFv-encoding gene amplification

Hv Primers: MHV.BACK1 (SkiI) 5?-gcggcccagccggccatggccGATGTGAAGCTTCAGGAGTC-3? MHV.BACK2 (SfiI) 5?-gcggcccagccggccatggccCAGGTGCAGCTGAAGGAGTC-3? MHV.BACK3 (SfiI) 5?-gcggcccagccggccatggccCAGGTTCAACTGCAGCAATC-3? MHV.BACK4 (SfiI) 5?-gcggcccagccggccatggccCAGGTTACTCTGAAAGAGTC-3? MHV.BACK5 (SfiI) 5?-gcggcccagccggccatggccGAGGTCCAGCTGCAACAATCT-3? MHV.BACK6 (SfiI) 5?-gcggcccagccggccatggccGAGGTCCAGCTGCAGCAGTC-3? MHV.BACK7 (SfiI) 5?-gcggcccagccggccatggccCAGGTCCAACTGCAGCAGCCT-3? MHV.BACK8 (SfiI) 5?-gcggcccagccggccatggccGAGGTGAAGCTGGTGGAGTC-3? MHV.BACK9 (SfiI) 5?-gcggcccagccggccatggccGAGGTGAAGCTGGTGGAATC-3? MHV.BACK10 (SfiI) 5?-gcggcccagccggccatggccGATGTGAACTTGGAAGTGTC-3? MHV.FOR1 5?-tgaaccgcctccaccTGCAGAGACAGTGACCAGAGT-3? MHV.FOR2 5?-tgaaccgcctccaccTGAGGAGACTGTGAGAGTGGT-3? MHV.FOR3 5?-tgaaccgcctccaccTGAGGAGACGGTGACTGAGGT-3? MHV.FOR4 5?-tgaaccgcctccaccTGAGGAGACGGTGACCGTGGT-3? Lv-k Primers: MKV.BACK1 5?-tctggcggtggcggatcgGATGTTTTGATGACCCAAACT-3? MKV.BACK2 5?-tctggcggtggcggatcgGATATTGTGATGACGCAGGCT-3? MKV.BACK3 5?-tctggcggtggcggatcgGATATTGTGATAACCCAG-3? MKV.BACK4 5?-tctggcggtggcggatcgGACATTGTGCTGACCCAATCT-3? MKV.BACK5 5?-tctggcggtggcggatcgGACATTGTGATGACCCAGTCT-3? MKV.BACK6 5?-tctggcggtggcggatcgGATATTGTGCTAACTCAGTCT-3? MKV.BACK7 5?-tctggcggtggcggatcgGATATCCAGATGACACAGACT-3 MKV.BACK8 5?-tctggcggtggcggatcgGACATCCAGCTGACTCAGTCT-3? MKV.BACK9 5?-tctggcggtggcggatcgGAAATTGTGCTGACTCAATCT-3?

MKV.FORI (NotI) 5?-ctgcggccgcTTTGATTTCCAGCTTGGTGCCTCC-3? Lv-l Primers: MLV.BACK 5?-tctggcggtggcggatcgCAGGCTGTTGTGACTCAGGAA-3? MLV.FOR (NotI) 5?-ctgcggccgcTTTGATTTCCAGCTTGGTCTTGGGCTG-3? Linkers: 5?-GGTGGAGGCGGTTCAGGCGGAGGTGGCTCTGGCGGTGGCGGATCG-3? 3?-CCACCTCCGCCAAGTCCGCCTCCACCGAGACCGCCACCGCCTAGC-5?

67 size genes encoding single-chain variable fragments tical density at 600 nm reached 0.5. Phages were rescued (scFvs) by using two overlapping linkers through with M13K07 helper phage (GE Healthcare) for 1 h, as splicing-overlap-extension PCR (SOE-PCR) method. described previously (25–28). Rescued phage particles in The linker sequences are shown in Table 2. the supernatant were precipitated by adding a 1/5 PCR fragments were subsequently digested with SfiI volume of 20z polyethylene glycol in 2.5 M NaCl and and NotI restriction enzymes and subcloned into phage incubating the mixture on ice for 30 min. The phage display vector pCANTAB5E (GE Healthcare, Pis- pellets were collected via centrifugation at 4,000 × gat cataway, NJ, USA). The gene cloning process is sum- 49C for 40 min and resuspended in 1 mL of 1 × phos- marizedinFig.1.TG1Escherichia coli was used for phate buffered saline. The obtained phage library was recombinant plasmid transformation. Bacterial colonies subjected to five rounds of selection. Panning selection from culture plates were picked for sequencing to exa- of the phage was carried out in enzyme-linked im- mine the diversity of the VH and Vk genes cloned into munosorbent assay (ELISA) plate wells (Nunc, the vector. The final VH/Vk + Vl transformant library Roskilde, Denmark) coated with EV71 VP4 peptide was scraped off and stored as glycerol stocks after sus- (Table 1). The ELISA plate wells were coated with pension in Luria Broth. antigen overnight at 49C as described previously (22). Antigen-specific phage selection via panning: The Antigen-coated wells were blocked with 1z (w/v) glycerol stocks of the scFv library were inoculated into 2 bovine serum albumin (BSA) in Tris-buffered saline × YTAG media containing 1.6z tryptone, 1z yeast (TBS)at379C for 1 h. After the blocking solution was extract, 0.5z NaCl, 1z glucose, and 100 mg/mL am- removed, the phage was added and incubated at 379C picillin and incubated at 379C with shaking until the op- for 1 h. The phage was washed with TBS, and then

Fig. 1. (Color online) Schematic illustration of the amplification of scFv-encoding genes and the cloning of ampli- fied genes into the expression plasmid.

68 Neutralizing Antibody against EV71 and CVA16

100 mM triethylamine (Sigma-Aldrich, St. Louis, MO, Table 3. Signal peptides were subsequently added to the USA) was added and incubated at room temperature for N-terminal of VH and VL domains via SOE-PCR ampli- 10 min to elute the bound phage. Eluted phage was first fication. neutralized by adding 1 M Tris-HCl (pH 7.4) and then PCR amplification was carried out in 50 mLofreac- amplified in E. coli. tion mixture containing 5 mLof10× buffer, 100 ng of Screening of VP4 peptide-specific clones: After five DNA fragments of the VH and VL domains and synthe- rounds of selection, HB2151 cells were infected with the sized DNA oligonucleotides of signal peptides of the amplified phages and cultured onto 2 × YTAG plates heavy and light chains, respectively, 2.0 U FastStart at 379C overnight (29). The obtained colonies were in- enzyme, 40 nmol of dNTP mix, and 20 pmol each of the oculated into 2 × YTAG medium, and the expression appropriate forward and reverse primers shown in of scFvs was induced by the addition of 1 mmol/L Table 3. PCR amplification was performed as described isopropyl-b-D-thiogalactopyranoside and incubation above. The VH and VL domains with signal peptides for 3–5 h. To extract soluble protein from the were eventually combined with CH and CL domains by periplasm, we resuspended the cell pellet in 0.5 mL of using SOE-PCR amplification. The primers SH.BACK ice-cold 1 × TES (0.2 mol/L Tris-HCl [pH 8.0], and CH.FOR. were used to amplify the heavy chain of 0.05 mmol/L EDTA, and 0.5 mol/L sucrose). After the chimeric antibody. The light chain was amplified by being mixed with 0.75 mL 1/5 × TESonice,thecells using the primers SL.BACK and CL.FOR. as shown in were vortexed for 30 s and incubated on ice for 1 h. The Table 3. PCR fragments of the heavy and light chains supernatants containing the soluble scFvs were collected were isolated using TaKaRa MiniBEST Agarose Gel via centrifugation at 12,000 × gfor15minandthen DNA Extraction Kit (Takara, Dalian, China) and in- tested for binding to VP4 peptide (Table 1) in an ELISA serted into pcDNA3.1(+) and pcDNA3.1/zeo(+)plas- with horseradish peroxidase-conjugated goat anti- mids (Invitrogen) to generate pcDNA3.1-H and mouse kappa antibody (Sigma-Aldrich) as a secondary pcDNA3.1/zeo-L constructs, respectively, after diges- antibody. Positive responses were identified with optical tion with BamHI and XbaI as shown in Fig. 2. density at 450 nm readings three times that of the con- Preparation of the chimeric antibody: The recom- trol (1z BSA in 1 × TBS). Plasmid DNA was isolated binant plasmids pcDNA3.1-H and pcDNA3.1/zeo-L from the positive clones for sequencing after confirming were co-transfected into CHO cells to express the chi- thepresenceofthescFvinserts in each clone via enzyme meric antibody. The CHO cells were grown at 379Cin digestion. RPMI-1640 medium with 10z fetal bovine serum (FBS) Gene cloning of the chimeric antibody: Total RNA in a humidified atmosphere with 5z CO2. CHO cells was extracted from B lymphocytes obtained from hu- (1 × 105) were plated in 24-well plates for transfection man peripheral blood by using the TRIzol method. The with Lipofectamine 2000 reagent (Invitrogen) according total RNA was reverse transcribed into cDNA by using to the manufacturer's protocol. High antibody- a SuperScript III First-Strand Synthesis System. Human producing cells were selected in the presence of immunoglobulin heavy-chain constant domain (IgG CH) 450 mg/mL neomycin. The supernatant was collected and light-chain constant domain (CL) of IgG1 subtype and concentrated 10-fold with an Amicon Ultra-15 cen- were amplified via PCR. PCR amplification was per- trifugal filter unit (Millipore, Bedford, MA, USA) and formed at 959C for 5 min followed by 30 cycles of 959C further purified with a Bio-Scale Mini Affi-Prep protein for 1 min, 559C for 45 s, and 729C for 2 min, followed A column (Bio-Rad, Hercules, CA, USA). The eluted by a 10-min extension at 729C. The sequences of the fractions were concentrated 4-fold with the Amicon cen- primers are shown in Table 3. VH and VL domains were trifugal filter. then amplified with PCR, by using the primers shown in

Table 3. The sequences of primers for chimeric antibody-encoding gene amplification

Heavy chain constant region primers: CH.BACK 5?-GTCACTGTCTCTGCAGCCTCCACCAAGG -3? CH.FOR. 5?-TGCTCTAGAGCATCATTTACCCGGGGACAGG-3? (XbaI) Light chain constant region primers: CL.BACK 5?-GTCGGTGGAGGCACCAAGGTGCAGATG-3? CL.FOR. 5?-TGCTCTAGAGCA CTAACACTCTCCCCTG-3? (XbaI) Heavy chain variable region primers: VH.BACK 5?-GCCACCATGGCCCAGGTTCAACTGCAGCAATC-3? (Kozak) VH.FOR. 5?-GGTGGAGGCTGCAGAGACAGTGACCAGAG-3? Light chain variable region primers: VL.BACK 5?-GCCACCATGGAAATTGTGCTGACTCAATCTC-3? (Kozak) VL.FOR. 5?-AAGATTTCCAGCTTGGTGCCTCC-3? Human heavy chain signal peptide: HHS: 5?-ATGAAACACCTGTGGTTCTTCTTCTTCCTGGTGGCAGCTCCCAGATGGGGTCCTGTCCGCCACCATGGCC-3? SH.BACK 5?-CGGGATCCCGATGAAACACCTG-3? (BamHI) Human light chain signal peptide: HLS: 5?-GGTCAGGACACAGCATGGACATGAGGGTCCCCGCTCAGCTCTGGGGCTCTGCTACTCTGGCTCCGCCACCATGGCC-3? SL.BACK 5?-CGGGATCCCGGGTCAGGACACAG-3? (BamHI)

69 Fig. 2. (Color online) Schematic illustration of the amplification and the cloning of heavy- and light-chain-encoding genes into the expression plasmids.

Sodium dodecyl sulfate-polyacrylamide gel elec- proteins were then transferred onto polyvinylidene trophoresis (SDS-PAGE) and western blot analysis: The fluoride membranes, which were blocked with 2z (w/v) purified fractions were subjected to 12z SDS-PAGE BSA in TBS solution for 1 h at room temperature and under reducing conditions (5z v/v b-mercaptoethanol) washed three times with TBS containing 0.05z (v/v) or 12–8z gradient SDS-PAGE under non-reducing Tween 20. The membrane was then probed with either conditions. After electrophoresis, the gels were stained primary mouse anti-EV71 mAb (MAB979, Millipore) or overnight with Coomassie blue and destained with the phage display-derived mAb for 1 h at 379Candwashed solution composed of 10z acetic acid, 40z methanol, three times with TBS buffer. After the membrane was and 50z water. incubated with the secondary goat anti-mouse or mouse EV71 and CVA16 viral particles were collected from anti-human antibody conjugated with the fluorescent cell culture supernatants and lysed with RIPA lysis dye: IRDye 800 CW (Kirkegaard & Perry Laboratories, buffer (Beyotime, Jiangsu, China) for 10 min on ice. Gaithersburg, MD, USA) for 45 min, blotting images After the solution was centrifuged at 12,000 × gfor were acquired by using the Odyssey infrared imaging 10 min, the concentration of proteins was measured by system (Li-COR Biosciences, Lincoln, NE, USA) and using Bradford's reagent (Bio-Rad). The protein sam- analyzed with the software provided by the manufac- ples were denatured via boiling for 10 min and loaded turer. onto an SDS-PAGE (12z) gel for electrophoresis. The In vitro neutralization assay: EV71 BJ08 (genogroup

70 Neutralizing Antibody against EV71 and CVA16

C4), EV71 BrCr-TR (genogroup A), and CVA16 strain via boiling with SDS and loaded onto SDS-PAGE gel CHZH05-01, were propagated in human rhab- for western blot analysis. Supernatant collected from domyosarcoma (RD) cells. titers in RD cells were the cultured RD cells was used as a negative control. determined by using the microtitration method and ex- Commercial anti-VP2 antibody was used as an antibody pressed as the 50z cell culture infective dose (CCID50) control. The results showed that anti-VP2 antibody according to the Reed-Muench method. Twofold serial recognized two protein bands of 28 kDa and 37 kDa, dilutions of sera were prepared with Dulbecco's Modi- which corresponded to the VP2 and VP0 proteins, re- fied Eagle medium (Gibco-BRL, Gland Island, NY, spectively, of EV71. These data indicated that anti-VP2 USA) containing 2z FBS. The EV71/CVA16 stock was antibody recognized the VP0 protein by reacting with diluted to a working concentration of 100 CCID50/ the VP2 protein, as VP0 is a precursor protein com- 50 mL. The neutralization assay was conducted with posed of VP2 and VP4 proteins. Interestingly, one pro- 96-well plates. In each well, 50 mLofdilutedmAbwas tein band of 37 kDa, which correlated to the VP0 pro- mixed with 50 mL of virus at 100 CCID50 and incubated tein, was detected by using the isolated recombinant for 2 h at 379C. A cell suspension (100 mL) containing mAb (Fig. 4), which suggests that the recombinant mAb 8,000 RD cells was further added to wells containing the recognizes VP0 protein by reacting with the linear virus/mAb mixture and cultured at 379C. After 7 days, VP4N20 epitope located in the N-terminal of EV71 VP4 the cells were observed to evaluate the appearance of protein and does not cross-react with other EV71- cytopathic effects by using a CCK8 kit (Beyotime). The encoded viral proteins. However, the VP4 protein could neutralization titer was defined as the highest serum di- not be detected by the recombinant mAb under these lution that could prevent the development of cytopathic experimental western blot conditions owing to its low effects in 50z of the cell. molecular weight (7 kDa). The VP4N20 peptide derived from the N-terminal se- quence of EV71 VP4 protein is widely conserved among RESULTS various EV71 subgenotypes and CVA16 (22). We there- Phage display-derived mAb recognizes the VP0 pro- fore investigated whether recombinant mAb reacts with teins of both EV71 and CVA16: In our recent study, we the VP4 protein of CVA16. As shown in Fig. 4, a com- showed that immunization with VP4N20 peptide der- mercial anti-VP2 antibody recognized the VP0 and VP2 ived from the N-terminal sequence of the VP4 protein of EV71 elicits a broad neutralizing activity against vari- ous EV71 genotypes (22). In the present study, mice were immunized subcutaneously with chimeric HBcAg particles containing VP4N20 peptide and received booster injections 3 weeks later. Splenocytes were then harvested 7 days after the boost for RNA extraction. A phage display technique was used to produce a recom- binant mAb targeting VP4N20 peptide. The recom- binant mAb was successfully selected with a competitive panning technique and produced in mammalian CHO cells. After purification with protein A, the purity of the isolated mAb was evaluated via densitometric analysis after staining with Coomassie blue. The light and heavy chains were visualized with SDS-PAGE gel electropho- resis under reducing and non-reducing conditions (Fig. 3). We were also interested in testing whether recom- Fig. 3. SDS-PAGE analysis. The phage display-derived antibody binant mAb recognizes EV71 protein. EV71 particles was analyzed by SDS-PAGE in non-reducing (NR; A) and harvested from cell culture supernatant were denatured reducing conditions (R; B). M: marker protein.

Fig. 4. Western blot analysis. Antibody MAB979 (A) and the phage display-derived antibody (B) were used to detect EV71 viral proteins. Antibody MAB979 (C) and the phage display-derived antibody (D) were used to detect CVA16 viral proteins.

71 Fig. 5. Neutralization assay against BrCr-TR strain of EV71. (A) RD cells without EV71 infection. (B) Cells infect- ed by EV71 pre-mixed with phage display-derived mAb at the concentration of 0.344 mg/mL. (C) Cells infected by EV71 pre-mixed with phage display-derived mAb at the concentration of 0.043 mg/mL. (D) Cells infected by EV71. Scale bar: 50 mm.

Fig. 6. Neutralization assay against CVA16. (A) RD cells without CVA16 infection. (B) Cells infected by CVA16 pre-mixed with phage display-derived mAb at the concentration of 0.344 mg/mL. (C) Cells infected by CVA16 pre-mixed with phage display-derived mAb at the concentration of 0.043 mg/mL. (D) Cells infected by CVA16. Scale bar: 50 mm. proteins of CVA16. By contrast, the recombinant mAb of HBcAg and EV71 VP4-derived peptide neutralize reacted only with the VP0 protein of CVA16 that con- homologous and heterologous EV71 strains and confer tained the VP4 protein, which suggested that the phage passive protection against a lethal EV71 challenge in ne- display-derived recombinant mAb recognized the VP4 onatal mice (22). Thus, in the present study, we tested protein of CVA16 and did not cross-react with other whether the phage display-derived recombinant mAb CVA16-encoded proteins. Therefore, the phage display- neutralizes the various subtypes of the EV71 virus. In derived mAb can recognize VP0 proteins of both EV71 vitro neutralizing experiments showed that the recom- and CVA16 by targeting the conserved VP4N20 epi- binant mAb prevented the development of cytopathic tope. effects in RD cells. The EV71 A-type BrCr-TR and C4- Phage display-derived mAb neutralizes both EV71 type Bj08 strains were used for the analysis. The median and CVA16 in vitro: Our previous observations showed half maximal inhibitory concentration (IC50) values of that antibodies elicited by chimeric particles consisting the recombinant mAb against the BrCr-TR and Bj08

72 Neutralizing Antibody against EV71 and CVA16 strains were 0.062 mg/mL and 0.098 mg/mL, respec- of both EV71 and CVA16. When used with western blot tively. Photographs of the neutralization experiment analysis, this antibody (MAB979) identified two protein conducted with BrCr-TR are shown in Fig. 5. Because bands in the isolated CVA16 particles that corresponded the phage display-derived mAb reacts with the VP0 pro- to the VP0 and VP2 proteins. The data indicated that tein of CVA16, we tested whether the recombinant mAb CVA16 particles harvested from the cell culture were a neutralizes the CVA16 virus. In vitro neutralizing data mixture of P- and R-particles, an outcome correlated showed that the recombinant mAb inhibited CVA16-in- with previous observations (32). However, the phage duced cytopathic effects in RD cells. The median IC50 of display-derived mAb reacted only with the VP0 precur- the mAb for CVA16 was 0.068 mg/mL (Fig. 6). These sor protein of CVA16, which indicates that this anti- data suggest that the phage display-derived anti-VP4 body specifically recognizes the VP4 protein of CVA16 mAb has broad cross-reactive neutralizing activity and does not cross-react with other capsid proteins of against both EV71 and CVA16, which are the major CVA16. etiological factors of HFMD. The results of in vitro neutralization assays further showed that the phage display-derived anti-VP4 mAb neutralizes both EV71 and CVA16 and prevents DISCUSSION cytopathic effects induced by either virus. Located in- Phage display is a widely used technology for the high side virions, VP4 is a myristoylated protein composed throughput screening of human therapeutic antibody of approximately 70 amino acids. However, the exter- candidates. It involves the establishment of an antibody nalization of VP4 can be triggered by the receptor library by expressing exogenous proteins on the surface binding-induced conformational alteration of the cap- of filamentous phage particles through fusion with sid, which forms the channels responsible for a safe phage surface proteins. This library can subsequently be release of the to the cell used to select the antibodies that bind the desired target cytoplasm. The picornavirus capsid structure is thus (30). Thus, in combination with antibody library screen- more dynamic than the crystal structure suggests (33, ing, phage display usually allows for the rapid isolation 34). The exposure of VP4 to the outside of the capsid of antigen-directed antibodies in vitro. In the current may potentially lead to anti-VP4 antibody-mediated study, the VP4N20 peptide was used to immunize mice, neutralization against picornavirus. The results of this after which mouse spleen cells were harvested to con- study are consistent with those of previous reports struct an antibody library and isolate a recombinant showing that antibodies elicited against the VP4 protein mAb targeting VP4N20 through phage display. The of neutralize infectivity in vitro (35). results of western blot analysis further showed that Although the phage display-derived anti-VP4 mAb phage display-derived recombinant antibody recognizes had dual neutralizing activities against both EV71 and the VP0 precursor protein consisting of the VP2 and CVA16, the IC50 data revealed that this activity was VP4 proteins by binding to the VP4N20 epitope, which low. Low affinity for the antigen of interest is the main indicates that an antibody targeting the VP4N20 epitope problem encountered in antibody selection with phage- was isolated using the phage display technique. display technology. However, the affinity of a selected Two forms of EV71 particles, E-particles and F-parti- antibody for a specific antigen can be improved with cles have previously identified found in the supernatant random mutation strategies (36,37). Recently, we iso- of cell cultures. EV71 E-particles are immature viral lated EV71-specific mAb with high neutralizing activity particles with low infectivity in which the P1 and VP0 from an EV71-infected infant by using a single B cell polyproteins are incompletely processed, whereas F- screening technique (unpublished data). Compared with particles are mature forms of EV71 (31). In our experi- antibodies obtained with phage display technology ments, EV71 viral particles were harvested from the su- which is a complete in vitro process for antibody screen- pernatant of cell cultures and loaded onto SDS-PAGE ing, B-cells activated by virus infection undergo affinity gel, and the commercial anti-VP2 antibody MAB979 maturation and generate high-affinity antibody was used to monitor the expression of VP2 and VP0 via responses to external antigenic stimuli, which are fun- western blot. Two positive protein bands corresponding damental processes in adaptive immunity (38). to the VP0 and VP2 proteins were detected with this sys- Neutralizing antibodies are both useful in the identifi- tem. This result is consistent with that of a previous cation of neutralizing epitopes for vaccine design and report (31) and indicates that the collected-EV71 parti- important for understanding the mechanism of broadly cles were a mixture of E- and F-particles. As expected, potent cross-neutralization, a modality that has poten- the VP0 precursor protein was visualized in collected tial preventive and therapeutic value. In recent years, EV71 particles by using the phage display-derived mAb substantial progress has been made in the discovery of under the applied electrophoresis conditions, and the an anti-EV71 antibody, and highly potent and broad results suggest that the isolated antibody is specific for neutralizing antibodies have been isolated from a varie- the VP4 protein and has no cross-reactive response with ty of animal models (19–21,39). These antibodies confer other EV71 capsid proteins. protection against EV71 lethal challenge in passively Our data further show that phage display-derived immunized animals. However, progress toward anti- mAb reacts with the VP4 protein of CVA16, which sug- CVA16 neutralizing antibodies remains limited. The gests that the isolated antibody can cross-reactively VP4 protein may be a potential target for neutralizing recognize the VP4 proteins of both EV71 and CVA16 by antibody development for CVA16 infection control. binding to the VP4N20 epitope that is well conserved In this report, a recombinant mAb against the VP4 between the two viral proteins. MAB979 is a cross-reac- protein of EV71 was isolated via phage display. The tive antibody that recognizes the VP2 and VP0 proteins phage display-derived mAb cross-reacted with the VP4

7373 proteins of both EV71 and CVA16 and cross-neutral- lethal enterovirus 71 infection in newborn mice by neutralizing ized both EV71 and CVA16 in vitro. The results suggest antibodies elicited by a synthetic peptide. Microbes Infect. 2007; that this mAb has great potential as a candidate for 9:1299-306. 19. Lim XF, Jia Q, Khong WX, et al. Characterization of an isotype broad neutralizing antibody against EV71 and CVA16 dependent monoclonal antibody against linear neutralizing epi- infection after affinity modification and humanization. tope effective for prophylaxis of enterovirus 71 infection. PLoS One. 2012;7:e29751. 20.XuL,HeD,LiZ,etal.Protectionagainstlethalenterovirus71 Acknowledgments This work was supported by grants from Inter- challenge in mice by a recombinant vaccine candidate containing national Science & Technology Cooperation (No. 2011DFG33200), a broadly cross-neutralizing eitope within the VP2 EF loop. Ther- National High Technology Research and Development Program of anostics. 2014;4:498-513. China (863 Program, No. 2012AA02A400), Project supported by the 21. Kiener TK, Jia Q, Meng T, et al. A novel universal neutralizing National Natural Science Foundation of China (No. 31470076). monoclonal antibody against enterovirus 71 that targets the high- ly conserved ``knob'' region of VP3 protein. PLoS Negl Trop Conflict of interest None to declare. Dis. 2014;8:e2895. 22. Zhao M, Bai Y, Liu W, et al. Immunization of N terminus of en- terovirus 71 VP4 elicits cross-protective antibody responses. BMC REFERENCES Microbiol. 2013;13:287. 1. Schmidt NJ, Lennette EH, Ho HH. An apparently new enterovi- 23. Zhou H, Fisher RJ, Papas TS. Optimization of primer sequences rus isolated from patients with disease of the central nervous sys- for mouse scFv repertoire display library construction. Nucleic tem. J Infect Dis. 1974;129:304-9. Acids Res. 1994;22:888-9. 2. Brown BA, Pallansch MA. Complete nucleotide sequence of en- 24. Marks JD, Hoogenboom HR, Bonnert TP, et al. By-passing im- terovirus 71 is distinct from . Virus Res. 1995;39: munization. Human antibodies from V-gene libraries displayed 195-205. on phage. J Mol Biol. 1991;222:581-97. 3. Tuthill T, Groppelli E, Hogle J, et al. . In: Joh- 25. Clark MA, Hawkins NJ, Papaioannou A, et al. Isolation of hu- nson JE editor. Cell Entry by Non-Enveloped Viruses. Berlin: man anti-c-erbB-2 Fabs from a lymph node-derived phage display Springer Berlin Heidelberg; 2010. p. 43-89. library. Clin Exp Immunol. 1997;109:166-74. 4. Wang X, Peng W, Ren J, et al. A sensor-adaptor mechanism for 26. Coomber DWJ, Hawkins NJ, Clark MA, et al. Generation of an- enterovirus uncoating from structures of EV71. Nat Struct Mol ti-p53 Fab fragments from individuals with colorectal cancer us- Biol. 2012;19:424-9. ing phage display. J Immunol. 1999;163:2276-83. 5. Chow M, Newman JF, Filman D, et al. Myristylation of picor- 27. Makvandi-Nejad S, Sheedy C, Veldhuis L, et al. Selection of sin- navirus capsid protein VP4 and its structural significance. Na- gle chain variable fragment (scFv) antibodies from a hyperim- ture. 1987;327:482-6. munized phage display library for the detection of the antibiotic 6. Ren J, Wang X, Hu Z, et al. Picornavirus uncoating intermediate monensin. J Immunol Methods. 2010;360:103-18. captured in atomic detail. Nat Commun. 2013;4:1929. 28. Zhang JL, Gou JJ, Zhang ZY, et al. Screening and evaluation of 7. McMinn PC. An overview of the evolution of enterovirus 71 and human single-chain fragment variable antibody against hepatitis its clinical and public health significance. FEMS Microbiol Rev. B virus surface antigen. Hepatobiliary Pancreat Dis Int. 2006;5: 2002;26:91-107. 237-41. 8. Yen FB, Chang LY, Kao CL, et al. infection in 29. He FT, Nie YZ, Chen BJ, et al. Expression and identification of northern Taiwan–epidemiology and clinical characteristics. J recombinant soluble single-chain variable fragment of monoclo- Microbiol Immunol Infect. 2009;42:38-46. nal antibody MC3. World J Gastroenterol. 2002;8:258-62. 9. Suzuki Y, Taya K, Nakashima K, et al. Risk factors for severe 30. Pande J, Szewczyk MM, Grover AK. Phage display: concept, hand foot and mouth disease. Pediatr Int. 2010;52:203-7. innovations, applications and future. Biotechnol Adv. 2010;28: 10. Wu Y, Yeo A, Phoon MC, et al. The largest outbreak of hand; 849-58. foot and mouth disease in Singapore in 2008: the role of enterovi- 31. Liu CC, Guo MS, Lin FH, et al. Purification and characterization rus 71 and coxsackievirus A strains. Int J Infect Dis. 2010;14: of enterovirus 71 viral particles produced from Vero cells grown e1076-81. in a serum-free microcarrier bioreactor system. PLoS One. 2011; 11. Chen KT, Chang HL, Wang ST, et al. Epidemiologic features of 6:e20005. hand-foot-mouth disease and caused by enterovirus 32. Chong P, Guo MS, Lin FH, et al. Immunological and biochemi- 71 in Taiwan, 1998–2005. Pediatrics. 2007;120:e244-52. cal characterization of coxsackie virus A16 viral particles. PLoS 12. Solomon T, Lewthwaite P, Perera D, et al. Virology, epidemiolo- One. 2012;7:e49973. gy, pathogenesis, and control of enterovirus 71. Lancet Infect 33. Fricks CE, Hogle JM. Cell-induced conformational change in Dis. 2010;10:778-90. poliovirus: externalization of the amino terminus of VP1 is 13. Yan XF, Gao S, Xia JF, et al. Epidemic characteristics of hand, responsible for liposome binding. J Virol. 1990;64:1934-45. foot, and mouth disease in Shanghai from 2009 to 2010: Enterovi- 34. Greve JM, Forte CP, Marlor CW, et al. Mechanisms of receptor- rus 71 subgenotype C4 as the primary causative agent and a high mediated rhinovirus neutralization defined by two soluble forms incidence of mixed infections with coxsackievirus A16. Scand J of ICAM-1. J Virol. 1991;65:6015-23. Infect Dis. 2012;44:297-305. 35. Katpally U, Fu TM, Freed DC, et al. Antibodies to the buried N 14. Zhao K, Han X, Wang G, et al. Circulating coxsackievirus A16 terminus of rhinovirus VP4 exhibit cross-serotypic neutralization. identified as recombinant type A human enterovirus, China. J Virol. 2009;83:7040-8. Emerg Infect Dis. 2011;17:1537-40. 36. Casson LP, Manser T. Evaluation of loss and change of speci- 15. Li Y, Zhu R, Qian Y, et al. Comparing Enterovirus 71 with Cox- ficity resulting from random mutagenesis of an antibody VH sackievirus A16 by analyzing nucleotide sequences and antigenici- region. J Immunol. 1995;155:5647-54. tyofrecombinantproteinsofVP1sandVP4s.BMCMicrobiol. 37. Kobayashi N, Oyama H, Kato Y, et al. Two-step in vitro antibody 2011;11:246. affinity maturation enables estradiol-1b assays with more than 16. Wu CN, Lin YC, Fann C, et al. Protection against lethal enterovi- 10-fold higher sensitivity. Anal Chem. 2010;82:1027-38. rus 71 infection in newborn mice by passive immunization with 38. Babcook JS, Leslie KB, Olsen OA, et al. A novel strategy for subunit VP1 vaccines and inactivated virus. Vaccine. 2001;20: generating monoclonal antibodies from single, isolated lympho- 895-904. cytesproducingantibodiesofdefined specificities. Proc Natl 17. Li J, Chang J, Liu X, et al. Protection from lethal challenge in a Acad Sci USA. 1996;93:7843-8. nenonatal mouse model by circulating recombinant form cox- 39. Li Z, Xu L, He D, et al. In vivo time-related evaluation of a ther- sackievirus A16 vaccine candidates. J Gen Virol. 2014;95: apeutic neutralization monoclonal antibody against lethal en- 1083-93. terovirus71infectioninamousemodel.PLoSOne.2014;9: 18. Foo DG, Alonso S, Chow VT, et al. Passive protection against e109391.

7474