Identification of a new neutralization antigenic site on poliovirus coat protein VP2

The Picornaviridae Family: Knowledge Gaps, Animal Models, Countermeasures, and Prototype Pathogens

Picornaviruses are nonenveloped particles with a single-stranded RNA genome of positive polarity. This virus family includes poliovirus, hepatitis A virus, rhinoviruses, and Coxsackieviruses. Picornaviruses are common human pathogens, and infection can result in a spectrum of serious illnesses, including acute flaccid myelitis, severe respiratory complications, and hand-foot-mouth disease. Despite research on poliovirus establishing many fundamental principles of RNA virus biology and the first transgenic animal model of disease for infection by a human virus, picornaviruses are understudied. Existing knowledge gaps include, identification of molecules required for virus entry, understanding cellular and humoral immune responses elicited during virus infection, and establishment of immune-competent animal models of virus pathogenesis. Such knowledge is necessary for development of pan-picornavirus countermeasures. Defining enterovirus A71 and D68, human rhinovirus C, and echoviruses 29 as prototype pathogens of this virus family may provide insight into picornavirus biology needed to establish public health strategies necessary for pandemic preparedness.

Identification of the epitope in human poliovirus type 1 Sabin strain recognized by the monoclonal antibody 1G10 using mimotope strategy

Following the recommended use of the inactivated poliovirus vaccine from Sabin strains (sIPV) by the WHO, a D antigen-specific neutralizing monoclonal antibody (mAb) 1G10 that recognized the human poliovirus type 1 Sabin strain (PV-I Sabin) was produced for D-antigen potency evaluation on sIPV. Study of the mAb 1G10 showed that it recognized a discontinuous conformational epitope of PV-I Sabin antigen. To identify this epitope quickly, easily and cost-effectively, clues to the epitope's identity were first obtained by employing a novel mimotope strategy based on a phage display library and "in silico" prediction. Then, the conformation of the epitope region, including the amino acid sequence, neutralizing sites, and location of this epitope, was identified using several classic epitope-mapping methods such as synthesized peptides analysis, neutralization assay and site-directed mutagenesis. The mimotope strategy may offer some guidance for achieving epitope identification in a more feasible and effective way. This mAb could be used in an in-house or national and international standard IPV D-antigen potency ELISA kit in the future.

Antibody recognition of picornaviruses and escape from neutralization: a structural view

Escape of picornaviruses from neutralization by monoclonal antibodies is mediated by substitutions of very few, defined amino acid residues of the capsid, generally located on the tip of some surface-exposed loops. Substitutions at the same positions are possibly of major relevance to antigenic variation of picornaviruses in the field. Such residues tend to cluster in discrete areas, termed antigenic sites. The structure of virus-antibody and peptide-antibody complexes, determined by cryoelectron microscopy and X-ray crystallography, combined with studies using site-directed mutagenesis, are beginning to reveal new features of picornavirus epitopes. This information complements and expands the view on picornavirus antigenicity previously provided by analyses of antibody-escape mutants. In addition to amino acids found replaced in escape mutants, other surface residues which remain invariant in spite of immune pressure also participate in contacts with the antibody molecule. Some invariant residues are even critical for the antigen-antibody interaction. Escape mutations occur at the subset of antigenically critical residues which are tolerant to change because they are not essentially involved in capsid structure or function. Restrictions to variation differ among epitopes; this may contribute to explain the different number of serotypes among picornaviruses, and the frequency at which antigenically highly divergent variants occur in the field.

Characterization of neutralizing antibodies to bovine enterovirus elicited by synthetic peptides

Six synthetic peptides corresponding to regions of bovine enterovirus (BEV), strain VG-5-27, elicited antibodies in mice which reacted with the virus in various assays. These antibodies have been characterised on the basis of their ability to (1) neutralize the virus, (2) bind to the intact virus particle in an immunoprecipitation test, (3) react with the denatured viral proteins, and (4) give immunofluorescent staining of virus infected cells. We have also determined the proportion of antipeptide antibody which binds to the virus in each case. All of the sera immunoprecipitated the virus and neutralized its activity to varying extents. Two of the sera specific for VP 1 sequences failed to react with denatured VP 1 whereas all the other antisera reacted with their respective parental proteins. All of the sera reacted with VG-5-27 infected cells in an immunofluorescence test. The proportion of antibodies to each peptide recognizing intact virus was variable and did not appear to correlate with neutralizing activity. In addition, the ability of each of the sera to react with and neutralize three other strains of the virus was analysed. With one of these strains significant cross-neutralization was observed.

Antigenic regions of poliovirus type 3/Sabin capsid proteins recognized by human sera in the peptide scanning technique

We used the peptide scanning technique to identify regions of poliovirus type 3/Sabin capsid proteins that bind antibodies from human immune sera. Several reactive regions were seen in VP1, VP2, and VP3 while peptides resembling VP4 did not bind antibodies. Peptides derived from sequences of the previously known antigenic sites 1 and 3 were recognized to a moderate degree. Peptides imitating the four loops in the closed ends of the beta barrels or the alpha helical CD insertions of VP1, VP2 or VP3, whether exposed in the crystal structure or not, all represented major reactivity in the scans. In VP1 several additional reactive regions were found in the amino terminal quarter of the protein, which is buried in the crystal structure, and in a partially exposed region close to but separated from the carboxy terminus. In VP2 the nonexposed peak activities clustered in a bridge-like structure spanning from the outer to the inner surface of the capsid shell. Likewise, most of the novel antigenic regions of VP3 clustered in an internal location and partially composed of beta sheets with a conserved amino acid sequence. Whether any of the novel antigenic sites is capable of inducing neutralizing antibodies is not known.

Poliovirus vaccines. A continuing challenge

The control of poliomyelitis remains a provocative challenge. Alternative vaccination schedules, continuing research toward better vaccines, and ongoing international scientific, epidemiologic, and economic collaboration may make it possible to provide effective immunization for all children of the world and eventually may eradicate poliomyelitis worldwide, a goal set forward by the Expanded Programme on Immunization of the World Health Organization.

Molecular basis for linkage of a continuous and discontinuous neutralization epitope on the structural polypeptide VP2 of poliovirus type 1

We obtained neutralizing monoclonal antibodies against a continuous neutralization epitope on VP2 of poliovirus type 1 strain Mahoney by using a combined in vivo-in vitro immunization procedure. The antibody-binding site was mapped to amino acid residues within the peptide segment (residues 164 through 170) of VP2 by competition with synthetic peptide and sequencing of resistant mutants. Cross-neutralization of these mutants with another neutralizing monoclonal antibody revealed a linkage of the continuous epitope and a discontinuous neutralization epitope involving both loops of the double-loop structure of VP2 at the twofold axis on the surface of the virion.

Construction of a poliovirus type 1/type 2 antigenic hybrid by manipulation of neutralization antigenic site II

There are three serotypes of poliovirus, poliovirus type 1 (PV-1), PV-2, and PV-3. These viruses each display four distinct neutralization antigenic sites, designated N-AgI, N-AgII, N-AgIIIA, and N-AgIIIB. It has been demonstrated previously that part of N-AgI can be replaced with heterogeneous amino acid sequences, resulting in hybrid viruses expressing heterogeneous antigenic determinants. To study whether hybrid viruses could be constructed by modifying another antigenic site, a part of N-AgII (amino acids 158 to 173 of VP2) of PV-1(Mahoney) was replaced with the equivalent sequence from PV-2(Lansing). The resulting hybrid was viable and expressed both PV-1 and PV-2 antigenic determinants. When inoculated into rabbits, the hybrid induced neutralizing antibodies against both PV-1 and PV-2, showing that amino acids 158 to 173 of VP2 are able to function as an antigenic site independent of the rest of N-AgII. Manipulation of N-AgII represents a useful alternative method for the production of hybrid polioviruses.

The antigenic structure of poliovirus

We have solved the structure of the Mahoney strain of type 1 and the Sabin (attenuated vaccine) strain of type 3 poliovirus by X-ray crystallographic methods. By providing a three-dimensional framework for the interpretation of a wealth of experimental data, the structures have yielded insight into the architecture and assembly of the virus particle, have provided information regarding the entry of virus into susceptible cells, and defined the sites on the virus particle that are recognized by neutralizing monoclonal antibodies. Thus locating mutations in variants selected for resistance to neutralizing monoclonal antibodies has defined three antigenic sites of the surface of the virion, and provided clues as to the mechanisms by which viruses escape neutralization. Finally, comparison of the structures of the two strains, together with analysis of sequences of many poliovirus strains, have begun to define the structural changes associated with serotypic differences between polioviruses.

Three-dimensional model of the capsid proteins of two biologically different Theiler virus strains: clustering of amino acid difference identifies possible locations of immunogenic sites on the virion

To explore structural features of the Theiler murine encephalomyelitis virion, we have constructed a three-dimensional model of the capsid proteins (VP1, VP2, and VP3) of the BeAn strain based on the atomic coordinates of the closely related Mengo virus. By superimposition of amino acid differences between BeAn virus and another Theiler virus strain, GDVII, on the three-dimensional model, clusters of differences were found in four distinct sites; the VP1 third corner, the VP2 "puff," and the VP3 first corner and "knob." These clusters, which are found on the surface of the virion, may represent neutralizing immunogenic sites that have come under selective pressure from neutralizing antibodies. Furthermore, the putative viral receptor binding site ("pit") of the two Theiler virus strains was found to be markedly conserved.

Different rhinovirus serotypes neutralized by antipeptide antibodies

Recently, Rossman et al. have described the three-dimensional structure of a human rhinovirus. A possible host cell surface receptor binding site was identified with a cleft on each icosahedral face. Two highly conserved amino-acid sequences found in rhino-, polio-, and foot-and-mouth disease (FMD) viruses are located near the base of this site and could be important in maintaining its topology. We have prepared site-specific antibodies to two synthetic peptides which include these sequences. The antibodies bind to the predicted capsid proteins of rhinovirus and neutralize approximately 60% of 48 rhinovirus serotypes tested. These results could provide a route to a rhinovirus vaccine effective against most of the numerous serotypes of this virus.

Epitopes on foot-and-mouth disease virus particles. I. Topology

Monoclonal antibodies (MAb) against an O1 Suisse isolate of FMDV were used to identify epitopes on the virus particle and to determine their relative function. Six major antigenic sites containing one or more epitopes were identified using competition ELISA. An epitope relationship is proposed consisting of a trypsin-sensitive sequential site, termed B2/D9, from the codings for the MAb which reacted with it, which was associated with virus infectivity and is probably at or near to the cell-binding site of the virion; a trypsin-resistant, conformational site 1C6/4C9 MAb reaction at which also resulted in neutralization of virus infectivity; a second trypsin-resistant, conformational site 3C8, where again MAb reaction neutralised virus infectivity; a third trypsin-resistant, conformational site 6C3/2G5, at which MAb-dependent neutralisation of virus infectivity was inefficient; a site 3G4, the expression of which was impaired but not destroyed by trypsin treatment, and was not related to virus infectivity; an internal site A8, which appears to be a "12S subunit-specific" site. This work clearly demonstrates for the first time that both trypsin-sensitive and trypsin-resistant neutralisable (infectivity-associated) sites exist on the FMDV particle, and only one of these can be related to the sequential site used to formulate current FMDV peptide vaccines.

Binding site of neutralizing monoclonal antibodies obtained after in vivo priming with purified VP1 of poliovirus type 1 is located between amino acid residues 93 and 104 of VP1

Three hybridomas obtained after in vitro stimulation of spleen cells of mice primed in vivo with purified VP1 of poliovirus type 1 (Mahoney) with the homologous virus produced antibodies which reacted with VP1 and immunoprecipitated and neutralized only the homologous virus. Evidence for the location of their binding sites was obtained by inhibition of virus neutralization and virus binding by a synthetic peptide comprising the amino acid sequence 93-104 of VP1 of poliovirus type 1 (Mahoney).

Neutralization of poliovirus by polyclonal antibodies requires binding of a single IgG molecule per virion

Neutralization of poliovirus type 1 was studied using radioactively labelled polyclonal IgG. With nonsaturating antibody concentrations various virus-antibody complexes were produced which were isolated by sucrose gradient centrifugation and identified by electron microscopy as virus monomers, dimers, trimers, tetramers and pentamers. The neutralization rate (n.r.) of each of the virus-antibody complexes relative to non-neutralized virus and the stoichiometry have been estimated. The monomer fraction showed that about every fifth virion was associated with one IgG molecule and neutralized. The antibody was bivalently attached. The majority of virus particles formed aggregates of different sizes, which were cross-linked by antibodies. The following neutralization rates and ratios of IgG to virus (IgG/V) were determined for the oligomers: dimers, 59.2 per cent n.r. and 0.55 IgG/V; trimers, 67.3 per cent n.r. and 0.66 IgG/V; tetramers, 79.0 per cent n.r. and 0.75 IgG/V; pentamers, 86.3 per cent n.r. and 0.98 IgG/V. Two different mechanisms of neutralization are proposed: i) an antibody-mediated mechanism specifically inhibits infectivity of the monomer virus-antibody complexes and ii) reduction of infectivity of oligomer virus-antibody complexes is caused simply by reduction of the actual number of infectious units. Immunoprecipitation of the denatured capsid proteins showed that only VP 1 was recognized by the polyclonal IgGs.

Mutations conferring resistance to neutralization with monoclonal antibodies in type 1 poliovirus can be located outside or inside the antibody-binding site

Antigenic variants resistant to eight neutralizing monoclonal antibodies were selected from wild (Mahoney) and attenuated (Sabin) type 1 infectious poliovirions. Cross-immunoprecipitation revealed interrelationships between epitopes which were not detected by cross-neutralization. Operational analysis of antigenic variants showed that seven of eight neutralization epitopes studied were interrelated. Only one neutralization epitope, named Kc, varied independently from all the others. This latter, recognized by C3 neutralizing monoclonal antibody, was present not only on infectious virions but also on heat-denatured (C-antigenic) particles and on isolated capsid protein VP1. Loss of the neutralization function of an epitope did not necessary result from the loss of its antibody-binding capacity. Such potential, but not functional, neutralization epitopes exist naturally on Mahoney and Sabin 1 viruses. Their antibody-binding property could be disrupted by isolating antigenic variants in the presence of the nonneutralizing monoclonal antibody and anti-mouse immunoglobulin antibodies. Single-point mutations responsible for the acquisition of resistance to neutralization in the antigenic variants were located by sequence analyses of their genomes. Mutants selected in the presence of C3 neutralizing monoclonal antibody always had the mutation located inside the antibody-binding site (residues 93 through 103 of VP1) at the amino acid position 100 of VP1. On the contrary, antigenic variants selected in the presence of neutralizing monoclonal antibodies reacting only with D-antigenic particles had mutations situated in VP3, outside the antibody-binding site (residues 93 through 103 of VP1). The complete conversion of the Mahoney to the Sabin 1 epitope map resulted from a threonine-to-lysine substitution at position 60 of VP3.

Use of monoclonal antibodies to identify four neutralization immunogens on a common cold picornavirus, human rhinovirus 14

A collection of 35 mouse monoclonal antibodies, raised against human rhinovirus 14 (HRV-14), was used to isolate 62 neutralization-resistant mutants. When cross-tested against the antibodies in a neutralization assay, the mutants fell into four antigenic groups, here called neutralization immunogens: NIm-IA, -IB, -II, and -III. Sequencing the mutant RNA in segments corresponding to serotype-variable regions revealed that the amino acid substitutions segregated into clusters, which correlated exactly with the immunogenic groups (NIm-IA mutants at VP1 amino acid residue 91 or 95; NIm-II mutants at VP2 residue 158, 159, 161, or 162; NIm-III mutants at VP3 residue 72, 75, or 78; and NIm-IB mutants at two sites, either VP1 residue 83 or 85, or residue 138 or 139). Examination of the three-dimensional structure of the virus (M. G. Rossmann, E. Arnold, J. W. Erickson, E. A. Frankenberger, J. P. Griffith, H.-J. Hecht, J. E. Johnson, G. Kamer, M. Luo, A. G. Mosser, R. R. Rueckert, B. Sherry, and G. Vriend, Nature [London], 317:145-153, 1985) revealed that each of the substitution clusters formed a protrusion from the virus surface, and the side chains of the substituted amino acids pointed outward. Moreover, four of the amino acid substitutions, which initially appeared to be anomalous because they were encoded well outside the cluster groups, could be traced to surface positions immediately adjacent to the appropriate viral protrusions. We conclude that three of the four antigens, NIm-IB, -II, and -III, are discontinuous. Thus, the amino acid substitutions in all 62 mutants fell within the proposed immunogenic sites; there was no evidence for alteration of any antigenic site by a distal mutation.

Poliovirus genome RNA hybridizes specifically to higher eukaryotic rRNAs

The RNA genome of poliovirus hybridizes to 28S and 18S rRNAs of higher eukaryotes under stringent conditions. The hybridization detected by Northern blot analyses is specific since little or no signal was detected for yeast or prokaryotic rRNAs or other major cellular RNAs. Southern blot analysis of DNA clones of mouse rRNA genes leads us to conclude that several regions of 28S rRNA, and at least one region in 18S rRNA, are involved in the hybridization to polio RNA, and that G/C regions are not responsible for this phenomenon. We have precisely mapped one of these hybridizing regions in both molecules. Computer analysis confirms that extensive intermolecular base-pairing (81 out of 104 contiguous bases in the rRNA strand) could be responsible for this one particular site of interaction (polio genome, bases 5075-5250; 28S rRNA, bases 1097-1200). We discuss the possible functional and/or evolutionary significance of this novel type of interaction.

Three-dimensional structure of poliovirus at 2.9 A resolution

The three-dimensional structure of poliovirus has been determined at 2.9 A resolution by x-ray crystallographic methods. Each of the three major capsid proteins (VP1, VP2, and VP3) contains a "core" consisting of an eight-stranded antiparallel beta barrel with two flanking helices. The arrangement of beta strands and helices is structurally similar and topologically identical to the folding pattern of the capsid proteins of several icosahedral plant viruses. In each of the major capsid proteins, the "connecting loops" and NH2- and COOH-terminal extensions are structurally dissimilar. The packing of the subunit "cores" to form the virion shell is reminiscent of the packing in the T = 3 plant viruses, but is significantly different in detail. Differences in the orientations of the subunits cause dissimilar contacts at protein-protein interfaces, and are also responsible for two major surface features of the poliovirion: prominent peaks at the fivefold and threefold axes of the particle. The positions and interactions of the NH2- and COOH-terminal strands of the capsid proteins have important implications for virion assembly. Several of the "connecting loops" and COOH-terminal strands form prominent radial projections which are the antigenic sites of the virion.

Antigenic variation and resistance to neutralization in poliovirus type 1

Mutations have been identified in variants of poliovirus, type 1 (Mahoney) on the basis of their resistance to neutralization by individual monoclonal antibodies. The phenotypes of these variants were defined in terms of antibody binding; the pattern of epitopes expressed or able to be exploited for neutralization were complex. Single amino acid changes can have distant (in terms of linear sequence) and generalized effects on the antigenic structure of poliovirus and similarly constituted virions.

Trypsin sensitivity of the Sabin strain of type 1 poliovirus: cleavage sites in virions and related particles

Treatment of the Sabin strain of type 1 poliovirus with trypsin produced two stable fragments of capsid protein VP1 which remained associated with the virions. Trypsinized virus was fully infectious and was neutralized by type-specific antisera. The susceptible site in the Sabin 1 strain was between the lysine at position 99 and the asparagine at position 100. A similar tryptic cleavage occurred in the Leon and Sabin strains of type 3 poliovirus, probably at the arginine at position 100, but not in the type 1 Mahoney strain, which lacks a basic residue at either position 99 or position 100. Tryptic treatment of heat-treated virus and 14S assembly intermediates produced unique stable fragments which were different from those produced in virions. The implications of our results for future characterization of the surface structures of these particles and structural rearrangements in the poliovirus capsid are discussed.

Natural variants of the Sabin type 1 vaccine strain of poliovirus and correlation with a poliovirus neutralization site

Independent substitution mutations have been detected in capsid polypeptide VP1 of the type 1 oral poliovirus vaccine isolated from normal infant vaccine recipients. These mutations map at amino acid residues 142 and 147 of VP1, a region only minimally hydrophilic. A synthetic peptide, corresponding to residues 141 to 147 of VP1 was synthesized, conjugated to a carrier polypeptide of bovine serum albumin. The conjugate was found to elicit a weak poliovirus neutralizing antibody response. It was also capable of priming the immune system for the production of IgG-type antibodies able to neutralize greater than 99.999% of infectious type 1 virus. It is suggested that region 141 to 147 of VP1 may be involved in neutralization of the virus and that the mutants may have accumulated by antibody selection.

Molecular cloning and complete sequence determination of RNA genome of human rhinovirus type 14

The genomic RNA of human rhinovirus type 14 was cloned in Escherichia coli and the complete nucleotide sequence was determined. The RNA genome is 7212 nucleotides long. A single large open reading frame of 6536 nucleotides was identified, which starts at nucleotide 678 and ends 47 nucleotides from the 3' end of the RNA genome. Comparisons of the specified proteins with those of other picornaviruses showed a striking homology (44-65%) between rhinovirus and poliovirus. The rhinovirus genomic RNA is rich in adenosine (32.1%) and strongly favors an adenosine or uridine in the third position of codons. The predicted map locations of all the rhinovirus structural and non-structural proteins and their proposed proteolytic cleavage sites are described.