Prediction and conformation by synthesis of two antigenic sites in human haemoglobin by extrapolation from the known antigenic structure of sperm-whale myoglobin.

The complete antigenic structure of sperm-whale myoglobin was previously determined in our laboratory. By structural analogy with myoglobin, two regions in human haemoglobin were predicted to comprise antigenic sites. One region was on the alpha-chain [alpha-(15-23)] and the other on the beta-chain [beta-(16-23)]. These two regions were synthesized, purified and characterized, and their immunochemistry was studied. Each peptide was able specifically to bind considerable amounts of haemoglobin antibodies. In a set of homologous proteins, barring any drastic conformational or electrostatic inductive effects exerted by the substitutions, and allowing for obstruction due to subunit interaction, the determination of the antigenic structure of one protein may serve as a useful starting model for the others.     

Antigenic structure of human haemoglobin. Localization of the antigenic sites of the beta-chain in three host species by synthetic overlapping peptides representing the entire chain.

A comprehensive synthetic approach is applied here to localize the continuous antigenic sites of the beta-chain of haemoglobin. The approach was based on the synthesis and purification of the following consecutive 15-residue peptides (each overlapping by five residues at both ends with the peptides preceding it and following it in the sequence): 1-15, 11-25 etc. Quantitative radiometric titrations of protein and peptide adsorbents were performed with 125I-labelled anti-haemoglobin antibodies from three different host species. The specificity of antibody binding to peptide adsorbents was confirmed by inhibition studies and by the binding specificity of antibodies isolated from peptide adsorbents. These studies established the full profile of antigenic beta-chain regions, which was found to be independent of the host species. Five major antigenic sites were localized, and their three-dimensional and structural characteristics are discussed in relation to the immune recognition of haemoglobin and other proteins.     

Molecular cartography of globular proteins with application to antigenic sites

A method, molecular cartography, is introduced as a way to quantitate the topographic structure of a protein surface. The method is applied to the problem of antigenic determinants, and it is used to examine local and global topography of reported antigenic regions on the surface of myoglobin and lysozyme. In nine antigenic sites taken from the literature and studied in detail, no local property was found in sites that was not also found in remaining regions of the surface. However, a strong correlation was found between antigenic sites and regions of the surface that are globally exposed. This finding suggests that global exposure of the protein surface may play a primary role in determining the antigenic structure of the protein. Molecular cartography may be useful in other instances of protein–protein interactions such as those between proteolytic enzymes and their substrates.     

Haemoglobin binding with haptoglobin. Localization of the haptoglobin-binding sites on the beta-chain of human haemoglobin by synthetic overlapping peptides encompassing the entire chain.

A synthetic approach was employed to identify the haptoglobin-binding sites on the beta-chain of human haemoglobin. This approach consists of the synthesis of a series of consecutive overlapping peptides that, together, systematically represent the entire protein chain. Fourteen synthetic peptides (beta 1-15, beta 11-25 etc.) were examined for their ability to bind human haptoglobin by quantitative solid-phase radiometric titrations of 125I-labelled haptoglobin. Of these 14 peptides only peptides beta 11-25 and beta 131-146 bound haptoglobin significantly; peptide beta 21-35 exhibited a small binding activity as a consequence of the overlap with peptide beta 11-25. On this basis and by examination of the three-dimensional structure of haemoglobin, it was concluded that the beta-chain of haemoglobin has two binding sites for haptoglobin that reside in, but do not necessarily encompass all of, the regions beta 11-25 and beta 131-146.     

Predicting antigenic sites on proteins.

The ability to predict antigenic sites on proteins is of major importance for the production of synthetic peptide vaccines and synthetic peptide probes of antibody structure. Many predictive methods, based on various assumptions about the nature of the antigenic response have been proposed and tested. This review will discuss the principles underlying the various approaches to predicting antigenic sites and will attempt to answer the question of how well they work.     

Subunit interacting surfaces of human haemoglobin. Localization of the alpha-subunit-beta-subunit interacting surfaces on the beta-chain by a comprehensive synthetic strategy.

A synthetic approach is introduced for localization of subunit interacting surfaces in oligomeric proteins. It consists of studying the binding activity of consecutive uniform overlapping peptides encompassing an entire subunit to the other, radiolabelled, subunit. This permits the establishment of the full profile of peptides that bind the other intact subunit. This approach has been demonstrated with haemoglobin, and its application here with the beta-chain peptides has enabled the localization on the beta-chain of the submolecular regions responsible for its binding to alpha-chain in solution. There was good agreement between the binding surfaces found here in solution and those expected from the crystal structure. There were also, however, some significant differences in the levels of binding found in solution and those expected from the crystal. Peptide 21-35 possessed much higher binding activity than would be expected from its contribution to subunit association in the crystal. Conversely, other regions expected to possess considerable binding capacity for alpha-chain either showed low (peptides 111-125 and 121-135) or almost no binding (peptides 91-105 and 101-115) capacity. On the other hand, two interacting surfaces (within peptides 11-25 and 71-85) that make a contribution in solution do not appear to play a role in the crystal. It is concluded that the regions of subunit association in solution are close to, but not identical with, those in the crystal. The approach should serve as an effective method for localization of subunit interacting surfaces of unknown proteins, even those that can be isolated only in traces.     

Localization and verification by synthesis of five antigenic sites of bovine serum albumin.

Recently we have shown that the major antigenic sites of bovine serum albumin exhibit functional equivalence progessively increasing with the time at which antibodies are obtained after the first immunization. Analysis of our recent immunochemical findings and the known covalent structure of bovine serum albumin have enabled us to predict the locations of five antigenic sites of bovine serum albumin. The predicted locations were synthesized, and immunochemical studies with late-course antisera showed them to constitute antigenic sites of native bovine serum albumin.     

The antibody response to myoglobin is independent of the immunized species. Analysis in terms of replacements in the antigenic sites and in environmental residues of the cross-reactions of fifteen myoglobins with sperm-whale myoglobin antisera raised in different species.

The recent determination of the entire antigenic structure of sperm-whale myoglobin with rabbit and goat antisera has permitted the examination of whether the antigenic structure recognized by antibodies depends on the species in which the antisera are raised. Also, by knowledge of the antigenic structure, the molecular factors that determine and influence antigenicity can be better understood in terms of the effects of amino acid substitutions occurring in the antigenic sites and in the environmental residues of the sites. In the present work, the myoglobins from finback whale, killer whale, horse, chimpanzee, sheep, goat, bovine, echidna, viscacha, rabbit, dog, cape fox, mouse and chicken were examined for their ability to cross-react with antisera to sperm-whale myoglobin. By immunoadsorbent titration studies with radioiodinated antibodies, each of these myoglobins was able to bind antibodies to sperm-whale myoglobin raised in goat, rabbit, chicken, cat, pig and outbred mouse. It was found that the extent of cross-reaction of a given myoglobin was not dependent on the species in which the antisera were raised. This indicated that the antibody response to sperm-whale myoglobin (i.e. its antigenic structure) is independent of the species in which the antisera are raised and is not directed to regions of sequence differences between the injected myoglobin and the myoglobin of the immunized host. Indeed, in each antiserum from a given species examined, that antiserum reacted with the myoglobin of that species. The extent of this auto-reactivity for a given myoglobin was comparable with the general extent of cross-reactivity shown by that myoglobin with antisera raised in other species. The cross-reactivities and auto-reactivities (both of which are of similar extents for a given myoglobin) can be reasonably rationalized in terms of the effects of amino acid substitutions within the antigenic sites and within the residues close to these sites. These findings confirm that the antigenicity of the sites is inherent in their three-dimensional locations.     

Delineation and conformational analysis of two synthetic peptide models of antigenic sites on rodent cytochrome c

Two regions of rodent cytochrome c, one within the first four residues of the molecule, which is N-acetylated, and one at a beta bend around residue 44, are known to be immunogenic and antigenic in rabbits. Using sequential peptide synthesis, we have determined the residues required for linear synthetic peptides within these sequences to bind to antibody raised in rabbits to intact rat cytochrome c. The residues that were important in binding the N-terminal peptides were N-acetylglycine at position 1 and valine at position 3. The smallest peptide sequence around residue 44 that would bind to antibodies was Gln-Ala-Ala-Gly-Phe. A theoretical conformational analysis of these peptides showed that the amino-terminal tetrapeptide adopts a wide statistical ensemble of conformational states and that the addition of residues beyond 41 and 45 in the other sequence does not appear to stabilize longer peptides in the native beta-bend conformation. Thus, the antigenicity conferred by Phe-46 and Gln-42 in this peptide is most likely due to the direct interaction of the side chains of these residues with the antibody binding site. The demonstration here that native conformation is not essential for antigenic peptides to bind to antibodies raised against the whole protein indicates that the association energy between antigen and antibody can be sufficient to induce conformation in conformationally flexible peptides. This supports the concept that anti-protein and anti-peptide antibodies may invoke conformational changes in cross-reactive protein antigens and may explain why longer peptides, which may adopt stable nonnative secondary structure, often do not bind to antibodies raised to the whole molecule.     

Identification of antigenic sites on staphylococcal enterotoxin B and toxoid

The staphylococcal enterotoxins (SEs) are capable of causing both food poisoning and a toxic shock-like illness in man. In addition, SEs are known to act as superantigens, stimulating T-cells according to their T-cell receptor Vβ type. Relatively little is known of their antigenic determinants and how these may relate to the structure and function of the toxins. As a step in the study of these relationships, the entire molecule of SEB was synthesized in duplicate as a series of octapeptides overlapping by seven residues. This series thus represented all the potential linear epitopes of eight residues or less. The reactivity of the octapeptide series with antisera raised to purified SEB and to formaldehyde-inactivated SEB has been used to locate several antigenic sites on native SEB and to identify antigenic differences in the toxoid. Three antigenic peptides identified from the antigenic profile were synthesized and characterized. These represented amino acids 21–32, 93–107 and 202–217 of SEB. None of these peptides affected SEB-induced T-cell proliferation. However, the occurrence or absence of cross-reactivity of these peptides with antibodies to native SEB corresponds to the degree of exposure and/or the rigidity of these regions within SEB.     

Prediction of immunodominant helper T cell antigenic sites from the primary sequence

We have used a data base of 23 known immunodominant helper T cell antigenic sites located on 12 proteins to systematically develop an optimized algorithm for predicting T cell antigenic sites. The algorithm is based on the amphipathic helix model in which antigenic sites are postulated to be helices with one face predominantly polar and the opposite face predominantly apolar. Such amphipathic structures can form when the polarity of residues along the sequence varies with a more or less regular period. Hence they can be identified by methods (so called power spectrum procedures) that detect periodic variations in properties of a sequence. The choice of power spectrum procedure, hydrophobicity scale, and model parameters are examined. An algorithm is tested by comparing the predicted amphipathic segments with the locations of the known T cell sites, counting the number of matches, and calculating the probability of getting this number by chance alone. The optimum algorithm, which predicts the largest number of sites with the lowest chance probability, uses the Fauchere-Pliska hydrophobicity scale and a least squares fit of a sinusoid as its power spectrum procedure. By applying this algorithm, 18 of the 23 known sites are identified (75% sensitivity) with a high degree of significance (p less than 0.001). The success of the algorithm supports the hypothesis that stable amphipathic helices are fundamentally important in determining immunodominance. This approach may be of practical value in designing synthetic vaccines aimed at T cell immunity.     

Localization of the alpha-chain cross-link acceptor sites of human fibrin

The potential cross-link acceptor sites of fibrin were specifically labeled with the fluorescent, substitute cross-link donor monodansyl cadaverine (MDC). Several fluorescent alpha-chain peptides generated from enzymatic and cyanogen bromide (CNBr) cleavage of the labeled fibrin were identified by sodium dodecyl sulfate disc gel electrophoresis; they were isolated and then characterized by amino acid analysis, NH2-terminal sequence analysis, and chromatographic and electrophoretic analyses of their digestion products. Ancrod cleavage of MDC-labeled fibrin produced a series of six alpha-chain peptides of molecular weights 34,000 to 12,000, each of which contained an MDC-labeled acceptor site, and an NH2-terminal alpha-chain derivative of molecular weight 37,500. The latter remains disulfide bound in the residual fibrin and has two MDC-labeled sit-s which are separable by CNBr cleavage. Mild plasmin digestion of MDC-labeled fibrin generated fluorescent alpha-chain peptides of molecular weights 45,000, 42,000, 35,000, 23,000, 21,000, and 2,500 in the supernatant and a nonfluorescent NH2-terminal alpha-chain derivative of molecular weight 25,000 which remained in the insoluble residual fibrin. The alignment of these plasmic supernatant peptides was determined from NH2-terminal sequence analyses which indicated that an MDC acceptor site was located at approximately residue 255 of the Aalpha-chain. Cleavage of the MDC-labeled alpha-chain by CNBr, however, localized most of its fluorescence (approximately 80%) to a fragment of molecular weight 29,000 which had the same NH2-terminal sequence as the labeled plasmic peptide of molecular weight 21,000. Both peptides were cleaved by ancrod into two acceptor site-containing peptides of approximately equal fluorescence. The preliminary NH2-terminal sequence analyses of these peptides, when combined with the above findings, indicated that these two other cross-link acceptor sites are in a peptide segment which comprises the middle 17% of the Aalpha-chain.     

Prediction of sequential antigenic regions in proteins

Prediction of antigenic regions in a protein will be helpful for a rational approach to the synthesis of peptides which may elicit antibodies reactive with the intact protein. Earlier methods are based on the assumption that antigenic regions are primarily hydrophilic regions at the surface of the protein molecule. The method presented here is based on the amino acid composition of known antigenic regions in 20 proteins which is compared with that of 314 proteins [(1978) Atlas of Protein Sequence and Structure, vol. 5, suppl. 3, 363-373]. Antigenicity values were derived from the differences between the two data sets. The method was applied to bovine ribonuclease, the B-subunit of cholera toxin and herpes simplex virus type 1 glycoprotein D. There was a good correlation between the predicted regions and previously determined antigenic regions.     

A model for the hepatitis B virus core protein: prediction of antigenic sites and relationship to RNA virus capsid proteins.

The sequences of the core proteins from several serotypes of human hepatitis B virus and related mammalian and avian hepadnaviruses are aligned with the vp3 capsid protein of mengo virus, a picornavirus. The homology indicates an eight-stranded antiparallel beta-barrel fold for the hepatitis protein, as observed in the tertiary structure of the picornavirus protein. The locations of known antigenic sites and other modifications are consistent with this structure for the core protein. The predicted folding suggests additional exposed antigenic sites and supports an evolutionary relationship between this family of enveloped DNA viruses and enveloped and non-enveloped RNA viruses.     

Distance calculation of residues neighbouring to lysozyme antigenic sites. Site-neighbouring residues whose evolutionary substitution can modify the characteristics and binding energy of the sites.

By using the antigenic structure of lysozyme determined in this laboratory and the X-ray co-ordinates we have calculated the closest-atom distances between each of the residues in the three antigenic sites and all the other amino acids of the lysozyme molecule. These calculations enabled us to identify the nearest neighbours to each of the site residues. Thus the immediate environment of each site residue is described. For the three antigenic sites there is a total of 71 neighbouring residues. The effects of evolutionary amino acid substitutions in site-neighbouring residues on the binding capacity of protein binding sites in general and on protein antigenic sites in particular are discussed. These, together with the direct replacements in site residues, will acount for the major effects. However, the limitations of this treatment are stressed. The smaller effects on antigenic sites of replacements at once-removed and even at more distant locations, which, when they become cumulative, could be considerable, are brought to attention, together with any influences of conformational readjustments that can take place as a result of evolutionary amino acid replacements.     

Identification of three continuous antigenic sites in horse muscle acylphosphatase

The main antibody-combining sites of horse skeletal muscle acylphosphatase were mapped by preparing and purifying CNBr, tryptic and peptic peptides from the pure enzyme, and looking for the immunoreactivity of each peptide by the dot-immunobinding assay using specific polyclonal antienzyme antibodies previously purified by immunoaffinity chromatography. The immunoreactive peptides were identified on the basis of either their elution times in the fingerprint analysis or amino acid composition, or both, by comparison with the known enzyme amino acid sequence. All the CNBr as well as two tryptic and two peptic peptides were immunopositive, leading to identification of three main continuous antigenic sites on the enzyme molecule. The strong inhibition (92%) of the antigen-antibody reaction carried out in the presence of antibodies previously incubated with the immunoreactive peptide mixture supports the possibility that, at our experimental condition, the three identified antigenic domains contain the main antigenic determinants of the enzyme. The relationship between structure and antigenicity of the immunoreactive peptides is discussed in detail.     

Use of synthetic peptides to map the antigenic determinants of glycoprotein D of herpes simplex virus.

The predictive algorithm Surfaceplot (J.M.R. Parker, D. Guo, and R.S. Hodges, Biochemistry 25:5425-5432, 1986) was used to examine glycoprotein D of herpes simplex virus type 1 (HSV-1) for amino acid residues with a high probability of being exposed on the molecular surface. Based on these data, 11 different peptides corresponding to 10-residue segments in the primary sequence of glycoprotein D and one 20-residue segment were synthesized, conjugated to carrier proteins, and used to generate specific antisera in rabbits. Two synthetic peptides predicted not to be on the surface of glycoprotein D were included as negative controls. The polyclonal antisera against individual synthetic peptide conjugates were in turn evaluated for their ability to recognize both isolated glycoprotein D and intact HSV-1 virions in an enzyme-linked immunosorbent assay. Based on Surfaceplot predictions, eight linear antigenic sites on glycoprotein D were thereby defined from the 12 antipeptide antisera prepared. Four of these sites contained epitopes to which complement-independent neutralizing antibodies could be generated. The latter sites corresponded to sequences 12 to 21, 267 to 276, 288 to 297, and 314 to 323 of the mature protein. An additional peptide sequence, 2 to 21, was found to generate antisera which had potent virus-neutralizing capacity in the presence of complement. Identification of a neutralizing epitope in the sequence 314 to 323 makes it likely that the membrane-spanning region of glycoprotein D is within the subsequent sequence, 323 to 339. Antipeptide antisera prepared in this study from 12 synthetic peptides contained 13 surface sites predicted by Surfaceplot, of which 7 were not predicted by the parameters of Hopp and Woods (Proc. Natl. Acad. Sci. USA 78:3824-3828, 1981). Of these seven sites not predicted by the Hopp and Woods plot, all generated antipeptide antibodies that bound to HSV-1 virions and three of these seven sites generated neutralizing antibodies. In total, 8 of 12 synthetic peptides containing surface regions produced antipeptide antibodies that bound to HSV-1 virions and 5 of these generated neutralizing antibodies. These results suggest the advantages of Surfaceplot in mapping antigenic determinants in proteins.     

Nearest-neighbour analysis of myoglobin antigenic sites. Nearest-neighbour residues whose replacement can alter the environment of binding-site residue(s) and thus change their characteristics and binding capability.

By using the known antigenic structure of sperm-whale myoglobin previously determined in this laboratory and the X-ray co-ordinates for the myoglobin molecule, we have calculated the nearest-atom distances between each of the residues of the antigenic sites and all the other amino acids of the myoglobin molecule. These calculations have enabled us to identify the nearest-neighbour residues to each of the residues in the five antigenic sites, and which thus describe the immediate molecular environment of the sites. The influences of chemical changes or replacements in these environmental residues on the binding capacity of an antigenic site, when considered together with replacements directly in the antigenic sites, are expected to account for the major effects and will be extremely useful in explaining the cross-reactions of myoglobins from various species. However, it is stressed that the analysis has limitations due to the qualitative estimates of the effects, the influences of substitutions of once-removed or even at more distant locations (especially when they are cumulative) and finally the influences of any conformational re-adjustments when these occur as a result of the replacement(s).     

Plasma desorption mass spectrometry of haemoglobin tryptic peptides for the characterization of a Hungarian alpha-chain variant.

S-Aminoethylated-alpha A and -beta A globin tryptic peptides separated by reversed-phase high-performance liquid chromatography have been analysed by plasma desorption mass spectrometry. Almost all the expected alpha A and beta A tryptic fragments were tentatively assigned relative to the known globin chain sequences based on the molecular weight obtained by plasma desorption mass spectrometric analysis of the purified peptides. The application of plasma desorption mass spectrometry for structure elucidation of a haemoglobin alpha-chain variant revealed the first case of Hb Hasharon in Hungary.     

The existence of two completely distinct antigenic sites within a decapeptide.

A decapeptide (1182-1191) derived from the bovine interphotoreceptor retinoid-binding protein (IRBP) was found to contain two completely distinct antigenic sites when tested in Lewis rats. One site, localized in sequence 1182-1191, is the core of the immunodominant and highly uveitogenic determinant of IRBP. The second site localizes within sequence 1183-1191 and becomes detectable only when tryptophan at position 1182 is deleted. Lymphocytes sensitized against the first, larger site recognized all longer peptides within sequence 1169-1191, as well as whole IRBP. In contrast, lymphocytes sensitized against the second, short epitope recognized only two peptides, 1184-1191 and (to a lesser degree) 1183-1191. The responses to both sites were restricted by the same major histocompatibility complex (MHC) product (I-A), as shown by monoclonal antibody blocking and by the finding that the lymphocyte response to 1184-1191 was competitively inhibited by peptide 1181-1191. The unique finding of two completely distinct antigenic sites within a decapeptide could be explained by the hypothesis that peptides of the two sites combine with the MHC molecule on antigen-presenting cells by different configurations, thus forming two distinct antigenicities.