Search

GnRH vaccines have been successfully used for the inhibition of gonadotropin secretion and gonadal function. As an alternative to native GnRH, retro-inverso (RI) GnRH might be an improved immunogen. The RI peptides are composed of D-amino acids assembled in the reverse order (C to N terminus) in relation to the parent L peptide. These peptides are immunogenic and can produce high titers of antibodies that bind the parent peptide with high affinity and specificity. We show that RI-GnRH peptides conjugated to ovalbumin as well as unconjugated RI-GnRH elicit high titers of anti-GnRH antibodies in rabbits and mice. Antibodies were affinity purified and shown by ELISA to be selective for mammalian GnRH compared with GnRH II and [Gln(8)]GnRH. The binding kinetics of antibody-peptide interactions was determined using biosensor technology (BIACORE). The purified anti-GnRH antibodies inhibited GnRH-stimulated signal transduction in COS-1 cells expressing the human GnRH receptor. Immunization of mice with unconjugated and conjugated RI-GnRH peptide, in the absence of complete Freund's adjuvant, produced antisera that cross-reacted with mammalian GnRH. As RI peptides are resistant to cleavage by proteolytic enzymes, they are potentially orally active. The ability of RI-GnRH peptides to produce antibodies to GnRH without conjugation and without Freund's complete adjuvant constitutes a novel vaccine with improved properties of potential application in animal management and sex hormone-dependent cancers.

Definitions of “species” and “type species” have been problematic in the taxonomy of different types of biological organisms. For virus taxonomy, the problems have been addressed by the International Committee on Taxonomy of Viruses (ICTV) and the results are compiled in The International Code of Virus Taxonomy and Nomenclature [4, 8]. The Code defines a virus species as a polythetic class of viruses that constitutes a replicating lineage and occupies a particular ecological niche. The application of this definition to virus taxonomy has been reviewed by van Regenmortel [7]. However, although the rules and procedures of virus taxonomy have referred to “type species” in connection with descriptions of genera for a number of years, the virus “type species” has not been explicitly defined. The requirement for virus type species arises from two sections of The Code [4.8]

The ability of a peptide to react specifically with the functional binding site of a complementary antibody is known as its antigenic reactivity or antigenicity. Our understanding of peptide antigenicity has improved considerably in recent years mainly through the X-ray crystallographic analysis of peptide-monoclonal antibody complexes. This knowledge is obtained along reductionist lines by turning the biological question of antigen recognition into the purely chemical phenomenon of protein-peptide interactions described in terms of atomic forces and non-covalent bonds. This makes it possible to improve the degree of steric complementarity between a peptide and a single monoclonal antibody and thus to improve the peptide's antigenicity following structure-based rational design principles. The situation is quite different with immunogenicity which is the ability of the peptide to induce an immune response in a competent host. Whereas antigenicity can be reduced to the level of chemistry, such a reduction is not achievable in the case of immunogenicity which depends on many complex interactions with various elements of the host immune system. These cellular and regulatory mechanisms cannot be controlled by adjusting the structure of the peptide in a predetermined manner. For this reason, it is not possible to develop a synthetic peptide vaccine using molecular design principles.

Elucidating the nature of the relationship between the structure and function of biomolecules remains one of the major challenges in biology. Biomolecules are dynamic entities that possess a variety of structures, and their functions at the molecular, cellular and organismic levels are quite different. Since there is no single causal link between structure and function, the search should be for correlations rather than causal relations. Biosensor instruments based on surface plasmon resonance are widely used for establishing correlations between the chemical structure of binding sites and their binding activity. Mutagenesis studies have shown that only a small percentage of the residues located in a binding site contribute to the binding energy. Since substitutions in residues located far away from the binding site are able to affect binding activity, this greatly complicates the rational design of proteins endowed with improved functions. However, biosensors can be used to determine and predict the influence of the chemical environment and of the structure of a ligand on binding kinetics.

Petunia plants collected in Sao Paulo City, Brazil, showing yellow mosaic, were naturally infected by a virus of the genus Tobamovirus identified according to particle morphology and size, host range, physical properties and cytopathic effects. On the basis of serological properties, amino acid composition and nucleotide sequence of the coat protein gene, the virus isolate was identified as a new strain of Tobacco mosaic virus (TMV-p). A conspicuous feature of this virus infection is the presence of virus-like particles within the mitochondrial matrix. The data from phylogenetic analysis indicate that TMV-p belongs to subgroup 1 of the genus Tobamovirus. Zusammenfassung In Sao Paulo Stadt, Brasilien, wurden Petunia-Pflanzen gesammelt, die Gelbmosaiksymptome zeigten. Es konnte nach der Partikelmorphologie und -grose, sowie dem Wirtsspektrum, den physiologischen Eigenschaften und cytopathologischen Einflussen nachgewiesen werden, dass es sich um eine naturliche Infektion mit einem Tobamovirus handelte. Anhand der serologischen Eigenschaften, der Aminosaurezusammensetzung und der Nukleotidsequenz des Hullproteins wurde das Virus als eine neue Art des Tabak Mosaik Virus (TMV-p) bestimmt. Eine Besonderheit dieser Virusinfektion ist das Vorhandensein von virusahnlichen Partikeln innerhalb der mitochondrialen Matrix. Die Daten der phylogenetischen Analyse deuten darauf hin, dass TMV-p der Subgruppe 1 dem Genus Tobamovirus angehort.

classes like species or families cannot be seen or centrifuged. It certainly is not possible to centrifuge the species Tobacco mosaic virus, the genus Enterovirus or the family Picornaviridae. Some virologists may incorrectly write that they have inoculated the species Nicotiana tabacum with one or other viral species, instead of saying that they inoculated a tobacco plant (a member of the species Nicotiana tabacum) with a member of a viral species. Precision in a scientific publication is, indeed, desirable and can be achieved by referring once, for instance in the Materials and methods section, to the taxonomic placement of the virus under study (species X, genus Y, family Z). Thereafter, vernacular names can be used throughout the publication. However, no conceptual precision would remain if one followed Bos [4] and called the virus that infects a plant an abstraction. An infecting virus is no more an abstraction than an infected plant. Bos is confusing a class Table 3. Examples of current names of vertebrate viruses and their binomial counterparts

The antigenic properties of the tobacco mosaic virus (TMV) have been studied extensively for more than 50 years. Distinct antigenic determinants called neotopes and cryptotopes have been identified at the surface of intact virions and dissociated coat protein subunits, respectively, indicating that the quaternary structure of the virus influences the antigenic properties. A correlation has been found to exist between the location of seven to ten residue–long continuous epitopes in the TMV coat protein and the degree of segmental mobility along the polypeptide chain. Immunoelectron microscopy, using antibodies specific for the bottom surface of the protein subunit, showed that these antibodies reacted with both ends of the stacked disk aggregates of viral protein. This finding indicates that the stacked disks are bipolar and cannot be converted directly into helical viral rods as has been previously assumed. TMV epitopes have been mapped at the surface of coat protein subunits using biosensor technology. The ability of certain monoclonal antibodies to block the co–translational disassembly of virions during the infection process was found to be linked to the precise location of their complementary epitopes and not to their binding affinity. Such blocking antibodies, which act by sterically preventing the interaction between virions and ribosomes may, when expressed in plants, be useful for controlling virus infection.

Most antigenic sites of proteins, known as discontinuous epitopes, are made up of residues on different loops that are brought together by the folding of the polypeptide chain. The individual loops are sometimes able, on their own, to bind to the antibody and they are then known as continuous epitopes. The binding sites of antibodies, known as paratopes, are built up from residues on six hypervariable loops known as complementarity determining regions (CDRs). Peptides corresponding to individual CDR loops are often able to bind the antigen and such peptides may be viewed as continuous paratopes. Using random combinatorial peptide libraries, it is possible to obtain peptides that bind to an antiprotein antibody without showing any sequence similarity with any part of the protein. Such epitope mimics are called mimotopes provided they are able also to elicit antibodies that react with the original antigen. The binding activity of mimotopes may partly be due to the phenomenon of hydropathic compleme...

Peptides corresponding to the immunodominant loop located at residues 135–158 on capsid protein VP1 of foot-and-mouth disease virus (FMDV) generally elicit high levels of anti-peptide and virus-neutralizing antibodies. In some instances, however, the level of neutralizing antibodies is low or even negligible, even though the level of anti-peptide antibodies is high. We have shown previously that the antigenic activity of peptide 141–159 of VP1 of a variant of serotype A can be mimicked by a retro-inverso (all-dretro or retroenantio) peptide analogue. This retro-inverso analogue induced greater and longer-lasting antibody titers than did the correspondingl-peptide. We now show that a single inoculation of the retro-inverso analogue elicits high levels of neutralizing antibodies that persist longer than those induced against the correspondingl-peptide and confer substantial protection in guinea pigs challenged with the cognate virus. In view of the high stability to proteases of retro-inverso peptide analogues and their enhanced immunogenicity, these results have practical relevance in designing potential peptide vaccines.

In an attempt to detect high M r glutenin subunits specifically by immunochemical means, antisera were produced against synthetic peptides corresponding to three N-terminal sequences and to two repetitive motifs of high M r glutenin subunits. The three N-terminal peptides, NT1, NT2 and NT3, differed by a single substitution at the sixth position and correspond, respectively, to the N-termini off Dx subunits, Ax and Bx subunits and By and Dy subunits. The anti-peptide sera did not cross react with gliadins or with low M r glutenin subunits, and differed in their ability to recognise high M r glutenin subunits. The antisera to the repetitive motifs recognised all high M r glutenin subunits, whereas the antisera to the N-terminal peptides detected only some of the subunits. The antiserum directed against the N-terminal peptide from Dx subunits detected these subunits specifically, whereas the antiserum directed against the N-terminal peptide corresponding to y type subunits did not react with the homologous subunits although it did react with Dx or Bx subunits. Antisera were also produced against internal sequences present in the N-terminal domain specific for x and for y-type subunits, but these antisera did not react with the cognate proteins. The failure of some anti-peptide sera to recognise the homologous high M r glutenin subunits may be due to differences in conformation between peptides and the corresponding regions in proteins.

Polyclonal and monoclonal antibodies (Mabs) were produced against the major type of N -terminal amino acid sequence of low M r glutenin subunits. The reactivities of these antibodies were determined using glutenin extracts of several bread wheat cultivars of known allelic composition. Analyses were performed by immunoblotting after one or two-dimensional electrophoresis. One Mab (Mab 6x1) was found to react with low M r glutenin subunits encoded by chromosomes 1B and 1D but not with subunits controlled by chromosome 1A. Only some of the subunits encoded at the Glu-D3 locus were recognised. In contrast, this Mab reacted with all the subunits controlled by the Glu-B3 locus. After single dimension SDS–PAGE, we observed significant differences between immunoblot patterns of cultivars expressing different low M r glutenin subunits from chromosome 1B. Mab6 x1 is a useful reagent for analysing the allelic composition at the Glu-B3 locus.

Summary Eight peptides encompassing neutralization antigenic site 1 of poliovirus type 1 (residues 93–103 of VP1) were synthesized in linear or cyclized form and used to immunize rabbits. The resulting anti-peptide antibodies were tested for their ability to react with linear peptide 95–104, with infectious virus D-particles and heated C-particles and for their capacity to neutralize poliovirus infectivity. A good correlation was observed between the ability of different peptide antisera to immunoprecipitate D-particles and neutralize virus infectivity. The peptides that induced a neutralizing antibody response in the highest number of immunized animals contained flanking residues 104–115 in addition to the 93–103 residues of the epitope. However, a high neutralizing antibody titre was also obtained in two of ten animals immunized with peptide 93–104 cyclized via an amide bond between Asp 93 and Lys 103 . It seems, therefore, that, at least in rabbits, the T-cell epitope recently identified in residues 103–115 of VP1 need not be present in the peptide immunogen in order to obtain poliovirus-specific neutralizing antibodies.

We have previously reported (Muller et al. Vaccine 1990, 8, 308) that two cyclic peptide analogues called D loop and K loop, corresponding to residues 139–147 in site A of the haemagglutinin (HA) of influenza A virus (strain X31), were both able to provide protective immunity to infected OF1 mice when administered in the form of peptide-ovalbumin conjugates. The predicted conformation of the D loop is nearly identical to that of the native loop known from the X-ray structure of HA, while the predicted conformation of the K loop differs significantly from the native one. In this study, the two peptides were conjugated to small unilamellar liposomes, thus creating a chemically defined immunogen, and OF1 mice were immunized with these liposomes containing monophosphoryl lipid A as adjuvant. Compared with protein carrier systems, the liposomal preparations are completely synthetic and avoid the use of Freund's adjuvant. By using liposomes associated with the D loop, we were able to achieve 70% protection of the mice against intranasal challenge with the influenza virus while no protection was obtained with the liposome-associated K loop. The difference in effect between the two liposome and ovalbumin carrier systems may result from the induction of different structures in the peptides when coupled to lipid anchors than when coupled to proteins.

For more than 50 years, tobacco mosaic virus (TMV) has been used as a model system for studying various aspects of virus-antibody interactions. Distinct epitopes called neotopes and cryptotopes have been identified in intact TMV particles and dissociated viral protein respectively and a correlation has been found to exist between the location of continuous epitopes and the extent of segmental mobility along the viral polypeptide chain. The occurrence of bivalent antibody binding was shown to influence the observed affinity of TMV antibodies and kinetic measurements of antibody binding to viral peptides made it possible to analyze the mechanisms of binding of monoclonal antibodies. It seems likely that the TMV model will continue to yield a rich harvest of immunochemical data relevant to many viral systems.

The species concept is applicable in virology because viruses have genomes, replicate, evolve, and occupy particular ecological niches. The following definition of virus species was accepted in 1991 by the International Committee on Taxonomy of Viruses: ‘A virus species is a polythetic class of viruses that constitutes a replicating lineage and occupies a particular ecological niche’. This definition does not provide a list of diagnostic properties for recognizing members of particular virus species. Furthermore, since a virus species is a polythetic class, it is impossible to use a single property such as a certain level of genome homology as defining property of the species. The implications of this new definition of virus species for future virus classification are discussed.

Differences in the affinity of a monoclonal antibody raised against the protein of tobacco mosaic virus for 15 related peptides (residues 134-146) carrying single-residue modifications were investigated using a novel biosensor technology (Pharmacia BIAcore). Analysis of the peptide-antibody interaction in real time allowed fast and reproducible measurements of both association and dissociation rate constants. Out of 15 mutant peptides analyzed, five were not recognized by the antibody at all, and seven were recognized as well as the wild-type peptide. For three of the peptides, the rate constants were different for the mutant and wild-type peptides. The pattern of residue recognition suggests that the epitope is formed by three residues (140, 143, and 144) in a helical conformation that mimics the structure in the protein. Even a minor modification of these residues totally abolishes recognition by the antibody. Modifications of adjacent residues result in small but significant differences in association and/or dissociation rate constants. One of the recognized residues is totally buried in the three-dimensional structure of TMV protein, suggesting that a structural rearrangement next to the helix occurs during protein-antibody interaction.

The binding affinity of a monoclonal antibody to tobacco mosaic virus (TMV) was studied using a Fab'-peroxidase conjugate. Measurement of the enzymatic activity allowed the determination of the amount of free antibody present after ultracentrifugation of virus-antibody complexes at equilibrium. The method was very sensitive and allowed measurements over a 1000-fold range of antibody concns. The calculated affinity constant decreased about 25 fold when the antibody concn used in the binding assay was increased from 30 ng/ml to 35 μg/ml.

Monoclonal antibodies (mAb) capable of reacting with different potyviruses were obtained by immunizing mice with proteolysed tobacco etch virus. The mAb were not equally effective in all ELISA formats and some were specific for different conformational states of the viral coat protein. The mAb also detected antigenic differences between purified virus particles and viral antigen in infected plant sap. In an ELISA format using antigen-coated plates, 5 different potyviruses (out of 7 viruses tested) could be detected in plant sap by one mAb. Different combinations of mAb and polyclonal antiserum could also be used for detecting several potyviruses by ELISA.

SUMMARYThe ability of synthetic peptides encompassing almost the entire sequence of snRNP U1A polypeptide to be recognized in ELISA by sera of autoimmune patients was investigated. Sera from 18 patients with mixed connective tissue disease (MCTD), 145 with systemic lupus erythematosus (SLE) and 120 with other rheumatic autoimmune diseases were tested with 13 overlapping peptides. Among them, peptide 257–282 and, to a lower extent, peptide 1–11 were recognized by MCTD, SLE and Sjogren’s syndrome sera. In contrast, peptide 35–58 was recognized by 94% of MCTD and only 19% of SLE sera. It did not react with any of the other patient sera. The ELISA results were compared with the pattern of reactivity observed in immunoblotting. The results indicate that peptide 35–58 probably contains a major epitope recognized by MCTD autoantibodies. It is noteworthy that in snRNP particles, this region of U1A interacts with RNA and presents only limited homology with the corresponding sequence 32–50 of U2B”.

SUMMARY The C terminal region of a Trypanosoma cruzi ribosomal P protein, encoded by the λgtll JL5 recombinant, defined a major antigenic determinant in chronic Chagas heart disease. Immunopurified anti-JL5 antibodies were tested for anti-human ribosome reactivity by immunoblotting. They recognized the parasite ribosomal P proteins and clearly reacted with the corresponding human P proteins. The peptide R-13, that comprises the 13 C terminal residues of the JL5 recombinant and defines the specificity shared between chronic Chagas heart disease anti-JL5 antibodies and the systemic lupus erythematosus (SLE) anti-P antibodies, was used to study the specificity and the IgG subclass distribution of the anti-R-13 response by ELISA. The R-13 autoepitope is recognized mainly by sera from chagasic patients, but not by sera from malaria patients. Moreover, there was a significant correlation between anti-R-13 antibody levels and anti-T. cruzi antibody titres. The anti-R-13 response was mainly restricted to the IgG1 heavy chain isotype and correlated with the anti-T. cruzi isotype distribution.

The binding affinity of a monoclonal antibody (Mab) to tobacco mosaic virus (TMV) was determined by measuring, in an enzyme-linked immunosorbent assay, the amount of free antibody present after ultracentrifugation of virus-antibody complexes at equilibrium. In antibody excess, univalent binding of Mabs was observed and the affinity constant was K = 3.2 +/- 0.4 10(8) l/mol; in antigen excess, bivalent antibody binding was observed and the antibody avidity was about 15 times higher. In antigen excess, it was imperative to correct experimental data for the presence of 0.55% inactive molecules in the immunopurified antibody preparation. Modelling studies suggest that in the case of antibodies of increasing affinity, it becomes increasingly important to correct for the presence of inactive antibody in the binding assay.

Histone antibodies have been obtained by immunizing rabbits with histones H1, H2A, H2B, H3, H4 and triacetylated H4, uncomplexed to RNA. The reactivity of these antibodies was investigated by ELISA using as antigen isolated histones and chromatin as well as thirty-five different synthetic peptides covering the entire sequence of the four core histones, two peptides of H1 and two acetylated peptides of H4. The binding of these antibodies to histones was also measured in immunoblotting and in microcomplement fixation (MCF) tests. In parallel experiments using the same assays the various antigens were tested with antisera raised against histones complexed with RNA. Antibodies induced in the absence of RNA did not react with histones in MCF tests nor with chromatin in ELISA but reacted with the histones in ELISA, although the antibody titers were somewhat lower than in the case of antisera to histone-RNA complexes. Antibodies to RNA-histone complexes reacted with histones in both ELISA and MCF tests. When they were tested with peptide-coated microtiter plates in a direct binding ELISA format, antibodies induced with uncomplexed histones recognized very few fragments which were mainly located in the N- and C-terminal ends of the histones.

The history of the discovery of viruses by Ivanovsky, Beijerinck, and Loeffler at the end of the nineteenth century is briefly summarized. When tobacco mosaic virus was crystallized by Stanley in 1935, viruses became fascinating entities because they were viewed as living chemicals that possess the ability to reproduce. The question of whether or not viruses should be regarded as living microorganisms is discussed at length. In order to answer this question, it is necessary to clarify the nature of living systems and organisms. Once this is done, it becomes clear that viruses, although they are biological systems, are not living microorganisms. They do not possess many of the essential attributes of living organisms such as the ability to capture and store free energy and they lack the characteristic autonomy that arises from the presence of integrated, metabolic activities. Viruses do not self-replicate themselves but are being replicated through the metabolic activities of the cells they have infected. A virus becomes part of a living system only after its genome has been integrated into an infected host cell, and it is no more alive than cellular constituents such as genes, organelles, and various macromolecules.

The role played by studies of tobacco mosaic virus (TMV) in the development of virology is summarized. TMV was the first virus shown to be able to pass through a bacteria-retaining filter; it was also the first virus to be crystallized, to have its morphology and structure elucidated and its coat protein sequenced. Experiments done with TMV RNA in the 1950s established that viral nucleic acid is the carrier of viral infectivity. The mechanism of self-assembly and disassembly of TMV particles is described and the antigenic properties of virions and dissociated coat protein subunits are discussed. Studies of TMV have also led to various biotechnological applications which are briefly described.

Antigenicity or antigenic reactivity refers to the capacity of viruses to bind to specific antibody molecules. The antigenicity of nonenveloped viruses resides in the antigenic sites or B-cell epitopes of capsid proteins that are recognized by the binding sites of antibodies. Protein epitopes are classified as either continuous or discontinuous depending on whether the amino acids included in the epitope are contiguous in the polypeptide chain or not. Most epitopes are discontinuous and since they consist of surface residues brought together by the folding of the peptide chain, their antigenic reactivity depends on the native conformation of the protein. The quaternary structure of viral capsids gives rise to epitopes known as neotopes. Neotopes arise either through conformational changes in the capsid proteins induced by intersubunit interactions or by the juxtaposition of residues from neighboring subunits forming a novel epitope. Immunogenicity is the ability of a protein to give rise to an immune response in a competent host and it can be defined only in the biological context of the host. Knowledge of the viral antigenic sites recognized by antibodies does not necessarily indicate which immunogenic structure initiated the production of antibodies in the immunized host. Failure to differentiate between antigenicity and immunogenicity is responsible for the lack of success in developing synthetic peptide vaccines against viral diseases.

Comparative immunochemical analysis of different tobamoviruses indicated that the previously reported coat protein sequence of Odontoglossum ringspot virus was likely erroneous. This sequence has been determined again by direct sequencing of the genomic RNA and was found to differ from the previously proposed sequence in 31 of the 157 amino acid residues. The extent of antigenic cross-reactivity between ORSV protein and other tobamovirus proteins was measured by ELISA and found to correlate satisfactorily with the degree of sequence homology.

Two cyclic peptide analogues corresponding to residues 139-146 (site A) of influenza A virus haemagglutinin (strain X31) were synthesized. The ability of these peptides to react with anti-influenza virus antibodies was found to depend on the conformation of the loop and on the orientation in which the peptide was presented to antibodies. Antibodies raised to the peptides were able to bind in ELISA with influenza virus antigen that had been allowed to dry on the microtitre plate. When OF1 mice were immunized with cyclic peptides, approximately 80% of the animals were protected against an intranasal challenge with influenza virus.

SUMMARY The reactivity of autoantibodies present in the serum of patients with systemic lupus erythematosus (SLE) was investigated by ELISA using seven overlapping synthetic peptides representing the entire sequence of the polypeptide D component of ‘Sm antigen’. Of the 165 SLE sera tested, 59% were found to contain IgG antibodies able to bind to peptide 1–20, while 37% of the sera reacted with peptide 44–67. All sera reacting with peptide 44–67 also reacted with peptide 1–20. These two peptides were only seldom recognized by the sera of 187 patients with other rheumatic autoimmune diseases or by 53 sera of normal individuals. In a parallel study using sera that reacted with the D band in immunoblotting, most of the sera recognized peptides 44–67 (89%) and 1–20 (67%), while 33% of them reacted with peptide 97–119. The use of these synthetic peptides in ELISA may be of considerable help for detecting anti Sm autoantibodies.

Peptides prepared by the solid-phase peptide synthesis (SPPS) approach are used increasingly in biological research, for instance to elicit anti-peptide antibodies that will recognize the intact, cognate protein. Recent advances in SPPS are reviewed, including the use of new coupling reagents, new methods for evaluating peptide purity and new techniques of automated and multiple peptide synthesis. Methods for enhancing peptide immunogenicity are discussed such as the use of adjuvants and liposomes, and of synthetic branched polypeptides as carriers.

Monoclonal antibodies to tobacco mosaic virus that bind only to one end of the viral rods have been shown to recognize the surface of the protein subunit designated as the bottom, which contains the right radial and left radial α-helices. The specificity of the antibody binding was established by immunoelectron microscopy of complexes in which the 5′ end of the RNA had been exposed at the bottom of the helical virus particle. These antibodies have been shown to bind to both ends of the stacked disk aggregate of TMV protein, which is therefore bipolar. The observations on the bipolarity of this structure are inconsistent with the presumption that stacked disks are formed by aggregation of polar two-layer disks.

Early work on protein antigenicity led to the common belief that proteins possess a finite number of antigenic determinants located in accessible regions of the molecule's surface. This view is now changing. As a result of extensive studies, seven continuous epitopes have been located on tobacco mosaic virus (TMV) protein by measuring the antigenic activity of short peptides with anti-protein antibodies. The structure of the viral protein has been refined in Cambridge, enabling the Strasbourg workers to correlate the position of these epitopes with regions of high segmental mobility in the protein. Surface accessibility is not a sufficient condition for antigenicity, because six short peptides corresponding to accessible regions of the protein possess no antigenic activity. Recent work in Strasbourg shows that when longer peptides of TMV protein are used, more antigenic determinants are found and these lie in more structured regions of the protein (e.g. helices). In these cases the longer peptide may be folding up in solution to mimic part of the native structure.

The basic concepts used in virus classification are analyzed. A clear distinction is drawn between viruses that are real, concrete objects studied by virologists and virus species that are man-made taxonomic constructions that exist only in the mind. Classical views regarding the nature of biological species are reviewed and the concept of species used in virology is explained. The use of pair-wise sequence comparisons between the members of a virus family for delineating species and genera is reviewed. The difference between the process of virus identification using one or a few diagnostic properties and the process of creating virus taxa using a combination of many properties is emphasized. The names of virus species in current use are discussed as well as a binomial system that may be introduced in the future.