|Home | About | Journals | Submit | Contact Us | Français|
Enteroviruses (EVs) are common seasonal viruses that are associated with a variety of diseases. High-quality monoclonal antibodies (MAbs) are needed to improve the accuracy of EV diagnosis in clinical laboratories. In the present study, the full-length VP1 genes of poliovirus 1 (Polio 1) and coxsackievirus B3 (Cox B3) were cloned, and the encoded proteins were expressed and used as antigens in an attempt to raise broad-spectrum MAbs to EVs. Two pan-EV MAbs were isolated: one raised against Polio 1 VP1 and the other against Cox B3 VP1. The binding sites of both pan-EV MAbs were mapped to an amino acid sequence within a conserved region in the N terminus of Polio 1 VP1 by peptide and competition enzyme-linked immunosorbent assay. Two additional MAbs, an EV70-specific MAb and an EV71/Cox A16-bispecific MAb, developed against EV70 and 71 VP1 proteins, were pooled with the two pan-EV MAbs (pan-EV MAb mix) and tested for their sensitivity and specificity in the staining of various virus-infected cells. The pan-EV MAb mix detected all 40 prototype EVs tested and showed no cross-reactivity to 18 different non-EV human viruses. Compared with two commercially available EV tests, the pan-EV MAb mix exhibited higher specificity than one test and broader spectrum reactivity than the other. Thus, our study demonstrates that full-length Polio 1 VP1 and Cox B3 VP1 can serve as effective antigens for developing a pan-EV MAb and that the pan-EV MAb mix can be used for the laboratory diagnosis of a wide range of EV infections.
Human enteroviruses (EVs) are classified into four species: Human enterovirus A (coxsackievirus A2 [Cox A2], A3, A5, A7, A8, A10, A12, A14, and A16 and EVs 71, 76, 89, 90, and 91), Human enterovirus B (Cox A9 and B1 to B6; echoviruses 1 to 7 [Echo 1 to 7], 9, 11 to 27, and 29 to 33; and EVs 69, 73 to 75, 77 to 88, 97, 100, and 101), Human enterovirus C (Cox A1, A11, A13, A17, A19 to A22, and A24; polioviruses 1 to 3 [Polio 1 to 3]; and EV96), and Human enterovirus D (EV68 and EV70) (14). Infection with these EVs causes a wide range of diseases in humans, from mild respiratory illness to severe aseptic meningitis. Diagnosis of EV infection depends mainly upon laboratory testing, since the clinical symptoms vary and overlap with other diseases. Laboratory diagnosis of EV infection is currently determined with either reverse transcription-PCR to detect EV RNA or by isolating the virus in cell culture followed by monoclonal antibody (MAb) staining (15, 19). However, the two commercially available pan-EV MAbs used in the staining assay either fail to react with multiple EV serotypes (9, 19) or cross-react with other non-EVs (8, 22). The lack of availability of highly reactive and specific pan-EV MAbs for diagnosis of EV infection could be due to the difficulties in developing specific MAbs against the extensive antigenic diversity among EVs.
EV capsid protein VP1 is one of four structural proteins of EV, and its antigenic homology among many different EV serotypes has been well documented (2, 7, 16-18). The N termini of most EV VP1 proteins contain highly conserved immunogenic regions that are recognized by sera from most EV-infected patients (2). This finding suggests that the EV VP1 protein can potentially be used as a target to raise pan-EV MAbs for the detection of a broad spectrum of EVs.
With the aim of developing highly specific and reactive pan-EV MAbs, we prepared full-length recombinant VP1 proteins from four different EV serotypes (Polio 1, Cox B3, EV70, and EV71) to use as immunogens for hybridoma production. The best-performing MAbs derived from each of the immunogens described above were combined to make a pool of four MAbs (pan-EV MAb mix). This pan-EV MAb mix compared favorably with the other two commercially available EV MAb reagents and showed great promise in its utility for the laboratory diagnosis of a wide range of EV infections.
Full-length sequences of Polio 1 VP1 and EV71 VP1 genes (GenBank accession no. AY531200 and BAD89965) were optimized with codons that are preferentially used in Escherichia coli. Multiple overlapping complementary oligonucleotides which encompass the entire coding sequences of both the VP1 genes were designed based on the codon-optimized sequences. The Polio 1 VP1 gene was synthesized by assembly PCR in two groups of reactions (A and B) with the oligonucleotides containing appropriate restriction endonuclease sites. The assembled products were then amplified using the outermost primers in a second PCR for gene amplification. The two amplified PCR products (fragments A and B) were each cloned into TA vector (Invitrogen, Carlsbad, CA), and the clones were confirmed by DNA sequencing. Fragments A and B were isolated by double restriction enzyme digestion (NcoI/BstEII for fragment A and BstEII/BamHI for fragment B) and then subcloned into pQE 60 expression vector (Qiagen, Valencia, CA) via three-piece ligation. The EV71 VP1 gene was synthesized by an overlap extension PCR method similar to the Polio 1 VP1 gene synthesis described above and was also cloned into pQE60 expression vector.
Viruses were propagated in a Buffalo green monkey kidney (BGMK) cell line. Total RNA was extracted from the cells infected with Cox B3 and EV70 viruses using the RNeasy Plus kit (Qiagen, Valencia, CA). PCR amplification of the full-length genes of Cox B3 VP1 and EV70 VP1 was carried out using primers with sequences as follows: 5′-ATCTCCATGGGAGGTCCAGTTGAAGATG-3′ (Cox B3 VP1 Fwd), 5′-ATCGGGATCCAAATGCACCAGTATTC-3′ (Cox B3 VP1 Rev), 5′-ACTGCCATGGGCGCTGCAACAACACAA-3′ (EV70VP1 Fwd), and 5′-AGTCGGATCCGGCTGTTGTCAAATTG-3′ (EV70 VP1 Rev).
The two PCR products were each cloned into pQE 60 expression vectors via NcoI and BamHI restriction sites (bold font). Both constructs were confirmed by DNA sequencing.
E. coli M15 (pREP4) cells were transformed with the four pQE60-EV VP1 constructs to express His6-tagged VP1 proteins of Polio 1, Cox B3, EV70, and EV71. Protein expression was induced with isopropyl-β-d-thiogalactopyranoside (IPTG), and each was purified under denaturing conditions using a nickel affinity chromatography column (Qiagen, Valencia, CA). The proteins were eluted with 6 M guanidine hydrochloride buffer of various pH values, and fractions containing the protein were identified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) after trichloroacetic acid precipitation. The proteins were refolded by serial dialysis into 1× phosphate-buffered saline (PBS)-0.25 M arginine buffer, and their purity was analyzed by SDS-PAGE.
Six-week-old female BALB/c mice were immunized with 100 μg of each recombinant VP1 protein emulsified in complete Freund's adjuvant (Sigma, St. Louis, MO) individually via intraperitoneal injection. A booster immunization using the same antigen in incomplete Freund's adjuvant (Sigma, St. Louis, MO) was carried out on days 25 and 45. Two mice from each group with the highest serum antibody titers were boosted intraperitoneally with the same antigen in PBS 4 days prior to cell fusion.
Fusion of splenocytes with Sp2/0Ag-14 myeloma cells (University of Pavia, Lombardy, Italy) was performed using methods described in detail elsewhere (6), with some modifications. Briefly, hybridomas were selected with hypoxanthine-aminopterin-thymidine medium containing 5% hybridoma cloning factor (BioVeris, Gaithersburg, MD), and supernatants were screened by both indirect fluorescence assay (IFA) on virus-infected cell monolayers and indirect enzyme-linked immunosorbent assay (ELISA) against VP1 recombinant proteins. Supernatants that were positive in both assays were further screened on antigen slides containing six Cox B viruses, six echoviruses, and three polioviruses (Bion, Des Plaines, IL; Chemicon, Temecula, CA). Hybridomas that secreted a broad range of cross-reactive antibodies were selected for cloning and MAb production. Isotypes of the selected MAbs were determined using an isotyping ELISA kit (Southern Biotech, Birmingham, AL). The best-performing MAbs, with an immunoglobulin G (IgG) isotype, were purified from the hybridoma supernatants by fast protein liquid chromatography using a protein G column.
Lysates of Polio 1-infected Vero cells and MRC-5 cells that were infected with Cox B3, Cox A16, EV70, and EV71 each were prepared individually with lane marker-reducing sample buffer (Pierce, West Chester, PA). The cell lysates were run in 4 to 20% SDS-PAGE, and the separated proteins were electrophoretically transferred to polyvinylidene difluoride membranes. The blotted membranes were blocked with PBS containing 1% Tween 20 and 3% skim milk and then incubated with 1 to 2 μg/ml of each EV MAb. The MAbs that bound to the membrane were detected by anti-mouse IgG-horseradish peroxidase (HRP) conjugate, followed by color development with Opti-4CN substrate (Bio-Rad, Hercules, CA).
The two most reactive and conservative peptides of EV VP1, PALTAVETGATNPL and SRSESSIENF, were synthesized (Proteos, Kalamazoo, MI). Plates were individually coated with 5 μg/ml of each peptide in coating buffer (0.1 M NaHCO3, pH 9.5). The four MAbs of Polio 1, Cox B3, EV70, and EV71 were serially diluted in PBS-Tween 20 and added to each immobilized VP1 peptide in triplicate. After a 1-h incubation at room temperature, the binding of the MAbs to the immobilized peptides was detected by anti-mouse IgG-HRP conjugate, followed by color development with BM blue POD substrate (Roche, Mannheim, Germany). The absorbance at 450 nm was read with a multiwell spectrophotometer (Molecular Devices, Sunnyvale, CA).
Super E-mix cells (Diagnostic Hybrids, Inc., Athens, OH) were infected with 41 prototype EVs for 20 to 48 h. The infected cell monolayers were either harvested to prepare slides or used directly for testing. Both the cell monolayers and slides were fixed with acetone and then incubated with the pan-EV MAb mix at 37°C for 30 min. After three washes with PBS, the cells were incubated with anti-mouse IgG-fluorescein isothiocyanate for another 30 min. Two commercial EV MAb reagents from Millipore and Dako were used side-by-side for comparison following the manufacturer's instructions. The stained cell monolayers or slides were mounted in mounting fluid and examined under a fluorescence microscope. Eighteen different non-EV human viruses were also inoculated into appropriate cell lines, including R-mix (Diagnostic Hybrids, Inc. Athens, OH), MRC-5, and CV-1. After a 20- to 48-h incubation, the infected cells were processed as described above. Cross-reactivity to mumps, rubella, and Epstein-Barr viruses were determined by IFA using commercially prepared viral antigen slides (Bion, Des Plaines, IL). All non-EV human viruses used in this test had been verified with MAbs specific to each virus.
To prepare immunogens for pan-EV hybridoma MAb generation, we cloned VP1 genes of Polio 1, Cox B3, EV70, and EV71 into pQE60 expression vectors. The four constructs were confirmed by DNA sequencing, and their deduced protein sequences were aligned to show the two most reactive, conserved regions, PALTAVETGATNPL and SRSESSIENF, located at the N termini of the VP1 proteins (Table (Table1).1). As shown in Table Table1,1, the relative amino acid identity at the N terminus is higher between Polio 1 VP1 and Cox B3 VP1 compared to the other two VP1 sequences. These four recombinant proteins were expressed in E. coli M15 (pREP4), purified by affinity chromatography, and analyzed by SDS-PAGE. The four expressed VP1 proteins migrated between the 30 and 40 kDa markers, which were consistent with the predicted size of each VP1 protein (data not shown). All four VP1 recombinant proteins were used to immunize mice for hybridoma production.
To determine whether the full-length VP1 proteins could be used to generate pan-EV MAbs, four groups of mice were immunized with each protein. A total of nine MAbs were identified as pan-EV MAbs which were able to recognize either all or most of 15 EVs on the antigen slides. Of the nine pan-EV MAbs, five were derived from Polio 1 VP1 and four were derived from Cox B3 VP1. Twelve EV70-specific MAbs were developed with EV70 VP1. Twenty-three MAbs were developed with EV71 VP1, nine of which recognized both EV71 and Cox A16 viruses, while the remaining 14 MAbs were specific to only EV71. Table Table22 summarizes the IFA screening results of the best-performing MAb from each group on the EV antigen slides and EV-infected cell monolayers. The Polio 1 VP1 MAb stained all 15 EVs, while the Cox B3 VP1 MAb recognized 14 of 15 EVs. The MAb derived from EV70 VP1 was specific for EV70, and the MAb derived from EV71 VP1 was bispecific to both EV71 and Cox A16 viruses. All four MAbs were isotyped as IgG1 and selected for further characterization.
Due to the high cross-reactivity of the Polio 1 VP1 MAb and the Cox B3 VP1 MAb observed among EVs (Table (Table2),2), we assumed that the MAbs might recognize a certain group-common epitope of the VP1 protein. To further characterize the MAbs, Western blot analysis was performed using cell lysates prepared from Vero cells infected with Polio 1 virus and MRC-5 cells individually infected with Cox B3, EV70, EV71, and Cox A16. Uninfected Vero and MRC-5 cell lysates were employed as negative controls. As shown in Fig. Fig.1,1, both the Polio 1 and Cox B3 VP1 MAbs were able to recognize the VP1 proteins of Polio 1 and Cox B3 but failed to recognize EV70 and EV71. Compared to the Cox B3 VP1 MAb, the Polio 1 VP1 MAb exhibited stronger binding to the VP1 proteins, as evidenced by the higher intensity of the bands. These results further suggest that the Polio 1 and Cox B3 VP1 MAbs recognize a common epitope shared by Polio 1 and Cox B3 viruses but that the affinity of the two MAbs may differ. The EV71-derived MAb recognized both EV71 and CoxA16, while the EV70 MAb showed specific binding to EV70 only.
To identify the linear epitope of the VP1 protein recognized by the Polio 1 and Cox B3 VP1 MAbs, two conserved peptides, PALTAVETGATNPL (P1) and SRSESSIENF (P2), were used as targets in epitope-mapping ELISA. As shown in Fig. Fig.2,2, both the Polio 1 and Cox B3 VP1 MAbs recognized peptide P1 and showed no binding to peptide P2. The EV70 VP1 MAb recognized only P1 at a low level when a high concentration of MAb was used. The relative binding affinity of the EV VP1 MAbs to P1 varied depending on the similarity of each P1 sequence to the EV group-common P1 sequence (Table (Table1).1). The EV71 VP1 MAb recognized neither P1 nor P2. These results indicate that the amino acid sequences in the target regions of EV VP1 proteins for both the Polio 1 and Cox B3 VP1 MAbs are closely related.
After characterization, the four MAbs were combined to make a single pan-EV MAb reagent (pan-EV MAb mix). To demonstrate that the pan-EV MAb mix could be used as a diagnostic tool for identifying a broad range of EV-infected cells, Super E-mix cells were infected with a total of 41 prototype EVs. Eighteen different non-EV human viruses were also used to infect appropriate cell lines for analyzing the specificity of the MAbs. Two other commercial EV MAb reagents from Millipore (Chemicon) and Dako were tested side-by-side to stain the virus-infected cells by IFA. As shown in Tables Tables33 and and4,4, the pan-EV MAb mix positively stained 40 of 41 prototype EVs, excluding only EV 68 with no cross-reactivity to any of the non-EVs. The Dako MAb also did not cross-react with any of the non-EVs but failed to stain Cox A16, Echo 2, Echo 8, Echo 19, EV68, EV69, EV70, and EV71 and reacted weakly with Cox A9, Cox A24, Cox B6, and Echo 25. The Millipore EV MAb blend stained all the EVs but showed cross-reactivity to seven nonhuman EVs, including parechovirus 2; rhinoviruses 2, 39, and 88; reovirus 3; herpes simplex virus type 2; and adenovirus 48. The pan-EV MAb mix was also tested against 10 additional non-EVs, including influenza A and B, respiratory syncytial virus, parainfluenza virus 1 (PIV1), PIV2, PIV3, varicella-zoster virus, and mumps, rubella, and Epstein-Barr viruses and showed no cross-reactivity to any of these viruses (data not shown). The intensity of fluorescent staining of all the positives was similar (a score of 3 to ≥4, a high level of staining intensity) in all the specimens except the weak positives that presented a low fluorescent signal but were distinguishable from negative specimens. Figure Figure33 shows images from representatives of the pan-EV MAb-stained cells infected with EVs that have a high prevalence or clinical significance in humans.
To develop a high-quality EV identification kit, we generated numerous MAbs against full-length VP1 proteins of different EV species. Two IgG1 MAbs, derived from Polio 1 and Cox B3, recognize the most conserved region PALTAVETGATNPL in the N terminus of Polio 1 VP1 protein, while neither showed any binding to another conserved group-common peptide SRSESSIENF. These two conserved group-common epitopes have been reported to be the most reactive peptides recognized by serum samples from patients infected with different EV serotypes (2). The relative binding affinity of the Polio 1 MAb is higher than that of the Cox B3 MAb (Fig. (Fig.2),2), in agreement with the Western blot results obtained with the same MAb (Fig. (Fig.1).1). These results indicate that the epitopes recognized by the two MAbs may not be identical but share high sequence homology (Table (Table1).1). This assumption is supported by both the IFA result, which shows that the Polio VP1 MAb detects one more EV than the Cox B3 VP1 MAb (Table (Table2),2), and the competition ELISA result, which shows that unlabeled Polio 1 VP1 MAb inhibits the binding of the biotinylated Cox B3 VP1 MAb to the Polio 1 VP1 (data not shown). These two MAbs are capable of recognizing most EVs, which is consistent with the previous finding that sera from most EV-infected patients react to the PALTAVETGATNPL peptide (2).
The exact epitopes recognized by the other two MAbs, EV70 and EV71, were not mapped, but the low-level binding of the EV70 MAb to the Polio VP1 peptide P1 in the epitope-mapping assay (Fig. (Fig.2)2) suggests that EV70 MAb might bind to the same conserved region. The low homology of the N-terminal sequence of the EV70 peptide might explain the low level of binding to the Polio 1 VP1 P1 peptide (Table (Table1).1). The MAb derived from EV71 VP1 showed no binding to the Polio 1 VP1 peptide (Fig. (Fig.2),2), indicating that the EV71 MAb might recognize a different peptide. The lack of binding may also be due to the high diversity of this epitope in EV71 VP1 compared with that in Polio 1 VP1 (Table (Table1).1). The ability of the EV71 MAb to recognize both EV71 and Cox A16 VP1 proteins is not surprising given the high degree of sequence homology between EV71 and Cox A16 VP1 proteins (12).
Initially, the mixture of the two pan-EV MAbs, Polio 1 VP1 and Cox B3 VP1, was evaluated in IFA for the identification of prototype EVs and archived clinical isolates. Both pan-EV MAbs reacted strongly with the majority of EVs tested in the assay. Not surprisingly, the two MAbs failed to recognize EV70, EV71, and Cox A16. This result is likely due to the fact that the degree of amino acid identity of these three VP1 proteins compared with the Polio 1 VP1 is low (Table (Table1).1). Since EV70, EV71, and Cox A16 are three important pathogens that cause acute hemorrhagic conjunctivitis, hand-foot-and-mouth disease, and epidemic severe central nervous system disease (10, 13, 20, 21), we also developed one EV70-specific MAb and one EV71/Cox A16-bispecific MAb using VP1 of EV70 and EV71, respectively. These two MAbs were added to the two pan-EV MAbs to generate a pan-EV MAb mix and successfully stained 40 prototype EVs (Table (Table3)3) and 219 archived clinical isolates (data not shown), including most, if not all, of the highly prevalent EV serotypes in humans during the past 2 decades in United States (3-5). The pan-EV MAb mix did not react with EV68 in this study; however, EV68 is not a typical EV due to its high genetic and antigenic similarity to human rhinovirus 87. Also based on its acid sensitivity, it has been suggested that EV68 could be classified as a rhinovirus (1, 11). Compared with two other commercially available EV diagnostic reagents from Millipore and Dako, our pan-EV MAb mix exhibited higher specificity than the Millipore EV MAb blend and broader spectrum EV reactivity than the Dako 5-D8/1 MAb (Tables (Tables33 and and4).4). Although the Dako MAb recognizes the VP-1 protein (17), it was generated using heat-inactivated purified Cox B5 (23). A previous study has also reported that the Dako MAb failed to react with human EV group A and EV71 (19). Since the epitopes recognized by the Millipore EV MAb blend are unknown to us, the reason(s) for the cross-reactivity to non-EVs is uncertain. The high specificity and reactivity of our pan-EV MAb mix are encouraging; however, many human viruses have yet to be examined, and in particular, only three types of rhinovirus were available for testing. Therefore, the performance of the pan-EV MAb mix will have to be assessed in the field in the coming years.
Based on the present study, we conclude that full-length VP1 proteins of Polio 1 and Cox B3 viruses are effective immunogens for generating pan-EV MAbs. Two MAbs that recognize the conserved region at the N terminus of VP1 exhibit high specificity and a broad spectrum of reactivity to most EVs. We blended these two pan-EV MAbs with two specific EV MAbs to generate a pan-EV MAb mix that is more promising than the other two commercially available EV tests and can be used as a diagnostic tool for identifying a wide range of EVs isolated from infected patients.
We thank David Scholl for his support of this project, Paul Olivo for critical review of the manuscript, and Nicole Barton for preparation of the images of EV-infected cells.
Published ahead of print on 6 August 2009.