A simple method of determination of antigenic determinants in proteins with known primary structure

A simple and rapid method is proposed for the localization of antigenic determinants in proteins of known primary structure exemplified by human myoglobin. The polypeptide chain of myoglobin was cleaved with BrCN (at Met residues) or with bromosuccinimide (at Trp and Tyr residues) under conditions which on average gave less than one scission per myoglobin molecule. The "single-hit" cleavage products were separated by gel electrophoresis and transferred to nitrocellulose by electroblotting. The peptides containing intact antigenic determinants were vizualized by immuno-peroxidase staining with four monoclonal anti-myoglobin antibodies. Comparison of the lengths of the immuno-reactive peptides with the known positions of methionine, tryptophan and tyrosine residues suggested that the four monoclonal antibodies were bound by myoglobin over the region Trp-14 to Met-55. As compared with other methods of localization, the method proposed is much faster and takes much lesser amount of protein.     

Topographic antigenic determinants recognized by monoclonal antibodies to sperm whale myoglobin

Monoclonal antibodies of high affinity (approximately 10(9) M-1) for sperm whale myoglobin were studied to pinpoint the antigenic determinants with which they interact. None of 6 different monoclonal antibodies tested reacted with any of the 3 CNBr cleavage fragments which encompass the whole sequence of myoglobin, an indication that they react with determinants present only on the native structure. To identify these sites, we compared the affinities of each antibody for a series of 14 mammalian myoglobins of known sequence and similar tertiary structure. Correlation of sequence differences with relative affinities allowed us, thus far, to identify critical antigenic residues recognized by 3 of the antibodies. Two of these antibodies recognize groups of residues which are far apart in primary structure but close together in the 3-dimensional structure of the native myoglobin molecule, i.e. topographic determinants. The third antibody distinguishes 140 Lys leads to Asn plus, probably, surface residues nearby. These determinants differ from previously reported antigenic sites on sperm whale myoglobin both in that they are topographic, rather than sequential, and in that almost all the critical residues recognized by these antibodies are outside the previously reported sites. Monoclonal antibodies are sensitive to subtle changes, e.g. Glu leads to Asp, in the antigenic site.     

Hydrophilicity and antigenicity of proteins — A case study of myoglobin and hemoglobin

Hydrophilicity index is used to locate antigenic determinants on two related groups of proteins-myoglobin and hemoglobin. The data on 41 species (including 34 mammals) of myoglobin show that average hydrophilicity for the complete myoglobin molecules as well as the average hydrophilicity for all hydrophilic regions put together seem to remain constant; the variation in the size and location of the antigenic determinants in these species is very small indicating that the antigenic sites are not shifted during evolution. In the case of both the proteins there is a good agreement between the antigenic sites picked up by using hydrophilicity index and the experimentally determined antigenic sites. The data on 56 species of hemoglobin α-chains and 44 species of hemoglobinβ-chains showed that although there are few sites on hemoglobin which have remained invariant during evolution, there is a significant variation in other sites in terms of either a splitting of a site, or a drastic change in the hydrophilicity values and/or a length of the site. Comparison of the hydrophilicity data on these two groups of proteins suggests that hemoglobins which perform a variety of functions as compared to myoglobins are evolving faster than myoglobins supporting the contention of earlier workers.     

Prediction of strong antigenic determinant of seminalplasmin and ribonuclease from the amino acid sequence

It has been reported earlier [Hopp, T.P. & Woods, K.R. (1981) Proc. Natl. Acad. Sci. US 78, 3824-3828] that the antigenic determinants of a protein can be delineated by examining the average local hydrophilicity values along the peptide chain. I have used this method to predict the strong antigenic determinants of two proteins, seminalplasmin and ribonuclease, of known sequence. In the former case, the N-terminal segment 1-14, and in the latter case the segments 27-38 and 80-86, are predicted to possess the antigenic determinants of the two proteins. Experimental verification already exists for the former case.     

Beneath the surface: Amino acid variation underlying two decades of dengue virus antigenic dynamics in Bangkok,Thailand

Neutralizing antibodies are important correlates of protection against dengue. Yet, determinants of variation in neutralization across strains within the four dengue virus serotypes (DENV1-4) is imperfectly understood. Studies focus on structural DENV proteins, especially the envelope (E), the primary target of anti-DENV antibodies. Although changes in immune recognition (antigenicity) are often attributed to variation in epitope residues, viral processes influencing conformation and epitope accessibility also affect neutralizability, suggesting possible modulating roles of nonstructural proteins. We estimated effects of residue changes in all 10 DENV proteins on antigenic distances between 348 DENV collected from individuals living in Bangkok, Thailand (1994-2014). Antigenic distances were derived from response of each virus to a panel of twenty non-human primate antisera. Across 100 estimations, excluding 10% of virus pairs each time, 77 of 295 positions with residue variability in E consistently conferred antigenic effects; 52 were within ±3 sites of known binding sites of neutralizing human monoclonal antibodies, exceeding expectations from random assignments of effects to sites (p = 0.037). Effects were also identified for 16 sites on the stem/anchor of E which were only recently shown to become exposed under physiological conditions. For all proteins, except nonstructural protein 2A (NS2A), root-mean-squared-error (RMSE) in predicting distances between pairs held out in each estimation did not outperform sequences of equal length derived from all proteins or E, suggesting that antigenic signals present were likely through linkage with E. Adjusted for E, we identified 62/219 sites embedding the excess signals in NS2A. Concatenating these sites to E additionally explained 3.4% to 4.0% of observed variance in antigenic distances compared to E alone (50.5% to 50.8%); RMSE outperformed concatenating E with sites from any protein of the virus (ΔRMSE, 95%IQR: 0.01, 0.05). Our results support examining antigenic determinants beyond the DENV surface.     

Prediction, conservation analysis, and structural characterization of mammalian mucin-type O-glycosylation sites

O-GalNAc-glycosylation is one of the main types of glycosylation in mammalian cells. No consensus recognition sequence for the O-glycosyltransferases is known, making prediction methods necessary to bridge the gap between the large number of known protein sequences and the small number of proteins experimentally investigated with regard to glycosylation status. From O-GLYCBASE a total of 86 mammalian proteins experimentally investigated for in vivo O-GalNAc sites were extracted. Mammalian protein homolog comparisons showed that a glycosylated serine or threonine is less likely to be precisely conserved than a nonglycosylated one. The Protein Data Bank was analyzed for structural information, and 12 glycosylated structures were obtained. All positive sites were found in coil or turn regions. A method for predicting the location for mucin-type glycosylation sites was trained using a neural network approach. The best overall network used as input amino acid composition, averaged surface accessibility predictions together with substitution matrix profile encoding of the sequence. To improve prediction on isolated (single) sites, networks were trained on isolated sites only. The final method combines predictions from the best overall network and the best isolated site network; this prediction method correctly predicted 76% of the glycosylated residues and 93% of the nonglycosylated residues. NetOGlyc 3.1 can predict sites for completely new proteins without losing its performance. The fact that the sites could be predicted from averaged properties together with the fact that glycosylation sites are not precisely conserved indicates that mucin-type glycosylation in most cases is a bulk property and not a very site-specific one. NetOGlyc 3.1 is made available at www.cbs.dtu.dk/services/netoglyc.     

New hydrophilicity scale derived from high-performance liquid chromatography peptide retention data: correlation of predicted surface residues with antigenicity and X-ray-derived accessible sites

A new set of hydrophilicity high-performance liquid chromatography (HPLC) parameters is presented. These parameters were derived from the retention times of 20 model synthetic peptides, Ac-Gly-X-X-(Leu)3-(Lys)2-amide, where X was substituted with the 20 amino acids found in proteins. Since hydrophilicity parameters have been used extensively in algorithms to predict which amino acid residues are antigenic, we have compared the profiles generated by our new set of hydrophilic HPLC parameters on the same scale as nine other sets of parameters. Generally, it was found that the HPLC parameters obtained in this study correlated best with antigenicity. In addition, it was shown that a combination of the three best parameters for predicting antigenicity further improved the predictions. These predicted surface sites or, in other words, the hydrophilic, accessible, or mobile regions were then correlated to the known antigenic sites from immunological studies and accessible sites determined by X-ray crystallographic data for several proteins.     

The nature of amino acid substitution in antigenic drift of hemagglutinin N3 and neuraminidase N2 from the influenza virus

The nature of amino acid replacements in 16 drift variants of hemagglutinin H3 subtype and 5 drift variants of neuraminidase N2 subtype of the influenza A virus were studied. The dependences of relative replacement frequencies and relative quantities of frequent replacements upon differences of properties of substituted residues are plotted. In contrast to most of the known proteins, amino acid replacements in hemagglutinin and neuraminidase depend weakly on the physico-chemical parameters of amino acids. For the antigenic determinants studied the replacement frequencies were compared to those calculated according to two models: one for conservative replacements and the other for accidental mutation of the genetic code. The differences in the nature of amino acid replacements are found in four antigenic determinants of hemagglutinin. The replacements in experimentally selected proteins are shown to go beyond limitations of natural variants. The explanations of the reasons of low epidemicity of some strains and ineffective attempt to imitate the natural antigenic drift of viruses by using experimental selection are proposed. The causes of time-limited circulation of H3N2 influenza virus subtype are discussed.     

Genetic analysis of type-specific antigenic determinants of herpes simplex virus glycoprotein C.

Herpes simplex virus type 1 (HSV-1) glycoprotein C (gC-1) elicits a largely serotype-specific immune response directed against previously described determinants designated antigenic sites I and II. To more precisely define these two immunodominant antigenic regions of gC-1 and to determine whether the homologous HSV-2 glycoprotein (gC-2) has similarly situated antigenic determinants, viral recombinants containing gC chimeric genes which join site I and site II of the two serotypes were constructed. The antigenic structure of the hybrid proteins encoded by these chimeric genes was studied by using gC-1- and gC-2-specific monoclonal antibodies (MAbs) in radioimmunoprecipitation, neutralization, and flow cytometry assays. The results of these analyses showed that the reactivity patterns of the MAbs were consistent among the three assays, and on this basis, they could be categorized as recognizing type-specific epitopes within the C-terminal or N-terminal half of gC-1 or gC-2. All MAbs were able to bind to only one or the other of the two hybrid proteins, demonstrating that gC-2, like gC-1, contains at least two antigenic sites located in the two halves of the molecule and that the structures of the antigenic sites in both molecules are independent and rely on limited type-specific regions of the molecule to maintain epitope structure. To fine map amino acid residues which are recognized by site I type-specific MAbs, point mutations were introduced into site I of the gC-1 or gC-2 gene, which resulted in recombinant mutant glycoproteins containing one or several residues from the heterotypic serotype in an otherwise homotypic site I background. The recognition patterns of the MAbs for these mutant molecules demonstrated that (i) single amino acids are responsible for the type-specific nature of individual epitopes and (ii) epitopes are localized to regions of the molecule which contain both shared and unshared amino acids. Taken together, the data described herein established the existence of at least two distinct and structurally independent antigenic sites in gC-1 and gC-2 and identified subtle amino acid sequence differences which contribute to type specificity in antigenic site I of gC.     

Identification of discontinuous antigenic determinants on proteins based on shape complementarities

Diverse procedures for identifying antigenic determinants on proteins have been developed, including experimental as well as computational approaches. However, most of these techniques focus on continuous epitopes, whereas fast and reliable identification and verification of discontinuous epitopes remains barely amenable. In this paper, we describe a computational workflow for the detection of discontinuous epitopes on proteins. The workflow uses a given protein 3D structure as input, and combines a per residue solvent accessibility constraint with epitope to paratope shape complementarity measures and binding energies for assigning antigenic determinants in the conformational context. We have developed the procedure on a given set of 26 antigen-antibody complexes with a known structure, and have further expanded the available paratope shapes by generating a virtual paratope library in order to improve the screening for candidate residues constituting discontinuous epitopes. Applying the workflow on the 26 given antigens with known discontinuous epitopes resulted in the correct identification of the spatial proximity of 12 antigen-antibody interaction sites. Combining solvent accessibility, shape complementarity and binding energies towards the identification of discontinuous epitopes clearly outperforms approaches solely considering accessibility and residue distance constraints.     

Topographic determinants on cytochrome c. I. The complete antigenic structures of rabbit, mouse, and guanaco cytochromes c in rabbits and mice1.

Rabbit, mouse, and guanaco cytochromes c differ from each other by only two amino acid residues. The identification is described of all of the antigenic determinants of mouse and guanaco cytochrome c that elicit an antibody response in rabbits, and those of the rabbit and guanaco proteins that elicity antibodies in the mouse. All except one of these sites center around single amino acid residue differences between the antigen and the host cytochrome c. The corresponding antibody popylations bind only to the areas of the protein in which the substitutions occur. Such antigenic determinants manifested in rabbits by quanaco and mouse cytochromes c are centered around residues 62 and 89, and residues 44 and 89, respectively. Similarly, the mouse recognizes sites containing residues 44 and 62 in guanaco cytochrome c, and residues 44 and 89 in rabbit cytochrome c. In none of these instances has a change in sequence failed to produce an antibody response. Each of these determinants appears to elicit and bind to its antibody, independently of other determinants present on the protein. In addition, two different autoantigenic responses have been detected. The antibodies produced against the determinant formed by glutamyl residue 62 of the guanaco protein in both rabbits and mice, the cytochromes c of which carry an aspartyl residue in that position, also bind to the aspartyl-containing region but with lower affinity. However, mouse and rabbit cytochrome c also elicit antibodies to the area of residue 62 in rabbits and mice, respectively, and these antibodies still bind more strongly to the glutamyl-than to the aspartyl-containing determinant. This last response occurs only when there are residue substitutions elsewhere in the molecule, because mice and rabbits fail to respond to their own cytochrome c. Antibodies produced in mice against the change from alanyl to valyl residue 44 by rabbit and guanaco cytochromes c also bind to the alanyl-containing determinant, except less tightly than to the valyl region. Conversely, antibodies raised in rabbits against the change from valyl to alanyl residue 44 only bind to this region when it carries an alanine. It is suggested that antigenic determinants that arise as a result of amino acid residue substitutions between the immunizing and the corresponding host protein, without a change in the spatial arrangement of the polypeptide backbone, be termed topographic determinants.     

The location and characterisation of the O-linked glycans of the human insulin receptor

O-linked glycosylation is a post-translational and post-folding event involving exposed S/T residues at beta-turns or in regions with extended conformation. O-linked sites are difficult to predict from sequence analyses compared to N-linked sites. Here we compare the results of chemical analyses of isolated glycopeptides with the prediction using the neural network prediction method NetOGlyc3.1, a procedure that has been reported to correctly predict 76% of O-glycosylated residues in proteins. Using the heavily glycosylated human insulin receptor as the test protein six sites of mucin-type O-glycosylation were found at residues T744, T749, S757, S758, T759, and T763 compared to the three sites (T759 and T763- correctly, T756- incorrectly) predicted by the neural network method. These six sites occur in a 20 residue segment that begins nine residues downstream from the start of the insulin receptor beta-chain. This region which also includes N-linked glycosylation sites at N742 and N755, is predicted to lack secondary structure and is followed by residues 765-770, the known linear epitope for the monoclonal antibody 18-44.     

Determinants of protein function revealed by combinatorial entropy optimization

We use a new algorithm (combinatorial entropy optimization [CEO]) to identify specificity residues and functional subfamilies in sets of proteins related by evolution. Specificity residues are conserved within a subfamily but differ between subfamilies, and they typically encode functional diversity. We obtain good agreement between predicted specificity residues and experimentally known functional residues in protein interfaces. Such predicted functional determinants are useful for interpreting the functional consequences of mutations in natural evolution and disease.     

Characterization of the antigenic and adhesive properties of FaeG, the major subunit of K88 fimbriae

The two K88 serotypes, K88ab and K88ac, differ in terms of antigenic and adhesive properties. The structural determinants of the serotype-specific epitopes and the identify of the amino acid residues involved in fimbriae-receptor interaction were studied by the construction and analysis of K88 hybrid proteins in which various parts of the K88ab and K88ac fimbrial subunit FaeG were exchanged, and by in vitro mutagenesis of non-conserved amino acid residues. Using a set of monoclonal antibodies, several regions or amino acid residues involved in the formation of serotype-specific antigenic determinants were located. The haemagglutinating activity of the hybrid and mutant proteins revealed several amino acid residues involved in the formation of the receptor binding site. A clear correlation was found between the receptor binding site and the serotype-specific antigenic determinants.     

Scanning the available Dictyostelium discoideum proteome for O-linked GlcNAc glycosylation sites using neural networks.

Dictyostelium discoideum has been suggested as a eukaryotic model organism for glycobiology studies. Presently, the characteristics of acceptor sites for the N-acetylglucosaminyl-transferases in Dictyostelium discoideum, which link GlcNAc in an alpha linkage to hydroxyl residues, are largely unknown. This motivates the development of a species specific method for prediction of O-linked GlcNAc glycosylation sites in secreted and membrane proteins of D. discoideum. The method presented here employs a jury of artificial neural networks. These networks were trained to recognize the sequence context and protein surface accessibility in 39 experimentally determined O-alpha-GlcNAc sites found in D. discoideum glycoproteins expressed in vivo. Cross-validation of the data revealed a correlation in which 97% of the glycosylated and nonglycosylated sites were correctly identified. Based on the currently limited data set, an abundant periodicity of two (positions-3, -1, +1, +3, etc.) in Proline residues alternating with hydroxyl amino acids was observed upstream and downstream of the acceptor site. This was a consequence of the spacing of the glycosylated residues themselves which were peculiarly found to be situated only at even positions with respect to each other, indicating that these may be located within beta-strands. The method has been used for a rapid and ranked scan of the fraction of the Dictyostelium proteome available in public databases, remarkably 25-30% of which were predicted glycosylated. The scan revealed acceptor sites in several proteins known experimentally to be O-glycosylated at unmapped sites. The available proteome was classified into functional and cellular compartments to study any preferential patterns of glycosylation. A sequence based prediction server for GlcNAc O-glycosylations in D. discoideum proteins has been made available through the WWW at http://www.cbs.dtu.dk/services/DictyOGlyc/ and via E-mail to DictyOGlyc@cbs.dtu.dk.     

Use of the averaged mutation rate in pieces of protein sequences to predict the location of antigenic determinants

Antigenic determinants in proteins show properties, which can be used to their prediction: high degree of accessibility, mobility, polarity etc. Furthermore, there is a remarkable correlation between the location of antigenic determinants and regions with high frequency of accepted mutations during the evolution. Consequently, it is possible to predict successfully the location of antigenic epitopes in proteins using the averaged mutation rate of protein sequence pieces.     

Proteome-wide prediction of overlapping small molecule and protein binding sites using structure

Small molecules that modulate protein-protein interactions are of great interest for chemical biology and therapeutics. Here I present a structure-based approach to predict 'bi-functional' sites able to bind both small molecule ligands and proteins, in proteins of unknown structure. First, I develop a homology-based annotation method that transfers binding sites of known three-dimensional structure onto protein sequences, predicting residues in ligand and protein binding sites with estimated true positive rates of 98% and 88%, respectively, at 1% false positive rates. Applying this method to the human proteome predicts 8463 proteins with bi-functional residues and correctly recovers the targets of known interaction modulators. Proteins with significantly (p < 0.01) more bi-functional residues than expected were found to be enriched in regulatory and depleted in metabolism functions. Finally, I demonstrate the utility of the method by describing examples of predicted overlap and evidence of their biological and therapeutic relevance. The results suggest that combining the structures of known binding sites with established fold detection algorithms can predict regions of protein-protein interfaces that are amenable to small molecule modulation. Open-source software and the results for several complete proteomes are available at http://pibase.janelia.org/homolobind.     

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).     

The application of an improved solid phase synthetic technique to the delineation of an antigenic site of apolipoprotein A-II

Previous studies have shown that the antigenic sites of human plasma high-density apolipoprotein A-II (apoA-II) are separate from their lipid-binding determinants in human high density lipoproteins (HDL). A specific radioimmunoassay has shown that three distinct antigenic sites are located in residues 4-23, 31-46, and 56-77; these studies suggested that an antigenic site might be restricted to residues 60-77 in the 56-77 fragment. To further delineate this site, we have developed a solid phase radioimmunoassay technique using an improved solid support on which selected sequences of peptides were synthesized, deprotected with HF, and the resulting peptidyl-resins tested for their capability of binding purified 125I-anti-apoA-II antibodies. Amino acid analyses and solid phase sequence analyses were performed to verify the sequence of the synthetic peptide on the solid support. Using this technique, 125I-anti-apoA-II antibodies had achieved 50% of maximal binding when residues 61-77 were attached to the solid support. The maximal binding was achieved by the addition of one more residue, Leu60, thus confirming our suggestion that a major antigenic site is located in residues 60-77. The binding to the peptidyl-resin was inhibited by a synthetic fragment corresponding to residues 60-77 indicating that the antibodies were specifically bound to the resin.     

Genetic mapping of antigenic determinants on a membrane protein

The antigenic determinants recognized by two monoclonal antibodies were mapped on LamB, an outer membrane protein of Escherichia coli. The procedure consisted of performing immunoprecipitation experiments with extracts of strains which produced truncated fragments of LamB, either in a free form (deletion and nonsense mutants) or fused to another polypeptide (malK-lamB and lamB-lacZ fusion strains). The conclusion is that the two antigenic determinants are located within 70 residues from the COOH-terminal end of LamB, which contains a total of 421 amino acids. Since these two antigenic sites were previously demonstrated to be exposed at the cell surface, it follows that a COOH-terminal portion of LamB must be located on the outer surface of the outer membrane.