Immunological properties of oxygen-transport proteins: hemoglobin, hemocyanin and hemerythrin.

It is now well documented that peptides with enhanced or alternative functionality (termed cryptides) can be liberated from larger, and sometimes inactive, proteins. A primary example of this phenomenon is the oxygen-transport protein hemoglobin. Aside from respiration, hemoglobin and hemoglobin-derived peptides have been associated with immune modulation, hematopoiesis, signal transduction and microbicidal activities in metazoans. Likewise, the functional equivalents to hemoglobin in invertebrates, namely hemocyanin and hemerythrin, act as potent immune effectors under certain physiological conditions. The purpose of this review is to evaluate the true extent of oxygen-transport protein dynamics in innate immunity, and to impress upon the reader the multi-functionality of these ancient proteins on the basis of their structures. In this context, erythrocyte-pathogen antibiosis and the immune competences of various erythroid cells are compared across diverse taxa.

Chemistry of antibody binding to a protein.

The chemistry of antibody recognition was studied by mapping the antigenicity of the protein myohemerythrin with peptide homologs of the protein sequence. The results suggest that the entire protein surface is antigenic, but the probability of there being antibodies to a given site is influenced by local stereochemistry. Although accessible to an antibody binding domain, the least reactive positions cluster in the most tightly packed and least mobile regions and are closely associated with narrow, concave grooves in the molecular surface containing bound water molecules. The most frequently recognized sites form three-dimensional superassemblies characterized by high local mobility, convex surface shape, and often by negative electrostatic potential.

Mechanisms of antibody binding to a protein.

The mechanisms of antibody binding to a protein were studied by an analysis of specific amino acid residues critical to nine antigenic sites on myohemerythrin. Rabbit antisera to the whole protein were assayed for binding to more than 1500 distinct peptide analogs differing from the protein sequence by single amino acid replacements. The results, combined with information from the three-dimensional crystallographic structure, were used to evaluate probable mechanisms of antibody binding at individual sites. The data from all sites examined indicate that initial binding to solvent-exposed amino acid residues may promote local side-chain displacements and thereby allow the participation of other, previously buried, residues.

Crystal structures of an antibody to a peptide and its complex with peptide antigen at 2.8 A.

The three-dimensional structures of an antibody to a peptide and its complex with the peptide antigen have been determined at 2.8 A resolution. The antigen is a synthetic 19-amino acid peptide homolog of the C helix of myohemerythrin (Mhr). The unliganded Fab' crystals are orthorhombic with two molecules per asymmetric unit, whereas the complex crystals are hexagonal with one molecule per asymmetric unit. The Fab' and the Fab'-peptide complex structures have been solved independently by molecular replacement methods and have crystallographic R factors of 0.197 and 0.215, respectively, with no water molecules included. The amino-terminal portion of the peptide sequence (NH2-Glu-Val-Val-Pro-His-Lys-Lys) is clearly interpretable in the electron density map of the Fab'-peptide complex and adopts a well-defined type II beta-turn in the concave antigen binding pocket. This same peptide amino acid sequence in native Mhr is alpha-helical. The peptide conformation when bound to the Fab' is inconsistent with binding of the Fab' to native Mhr, and suggests that binding of the Fab' to conformationally altered forms of the native Mhr or to apo-Mhr. Immunological mapping previously identified this sequence as the peptide epitope, and its fine specificity correlates well with the structural analysis. The binding pocket includes a large percentage of hydrophobic residues. The buried surfaces of the peptide and the antibody are complementary in shape and cover 460 A2 and 540 A2, respectively. These two structures now enable a comparison of a specific monoclonal Fab' both in its free and antigen complexed state. While no major changes in the antibody were observed when peptide was bound, there were some small but significant side chain and main chain rearrangements.

Searching for peptide ligands with an epitope library.

Tens of millions of short peptides can be easily surveyed for tight binding to an antibody, receptor or other binding protein using an "epitope library." The library is a vast mixture of filamentous phage clones, each displaying one peptide sequence on the virion surface. The survey is accomplished by using the binding protein to affinity-purify phage that display tight-binding peptides and propagating the purified phage in Escherichia coli. The amino acid sequences of the peptides displayed on the phage are then determined by sequencing the corresponding coding region in the viral DNA's. Potential applications of the epitope library include investigation of the specificity of antibodies and discovery of mimetic drug candidates.

Folding of immunogenic peptide fragments of proteins in water solution. II. The nascent helix.

1H nuclear magnetic resonance experiments indicate formation of secondary structures in water solutions of a synthetic immunogenic peptide of sequence EVVPHKKMHKDFLEKIGGL corresponding to the C-helix (residues 69 to 87) of myohemerythrin. The conformational ensemble consists of a set of turn-like structures, distributed over the C-terminal half of the peptide and rapidly interconverting by way of unfolded states. These structures, termed nascent helix, are stabilized into helical structure with long-range order in water/trifluorethanol mixtures. Circular dichroism measurements confirm the presence of 50% helix in water/trifluoroethanol but show no evidence of helicity in water solutions of the peptide. It is apparent that no one member of the transient set of helical conformations which constitutes the nascent helix is sufficiently long to be detectable by circular dichroism experiments. No preferred conformations could be detected by nuclear magnetic resonance in the N-terminal half of the peptide, either in water or water/trifluoroethanol mixtures. This region of the peptide is stabilized in helix by long-range interactions in the folded protein. The possible role of nascent secondary structure in induction of antipeptide antibodies and in initiation of protein folding is discussed.

Synthesis and characterization of a recombinant myohemerythrin protein encoded by a synthetic gene.

The antigenic epitopes of the myohemerythrin (MHr) molecule have been studied extensively. The critical amino acid residues responsible for its immune recognition have been identified by using synthetic peptides and the technique of epitope scanning. To assess the true relevance of these techniques for determining the molecular mechanism of antigenic recognition and immunogenicity, the results obtained with isolated peptides should be tested in the context of the folded protein. To this end, we have designed and constructed a synthetic MHr gene, in modular form, which will allow subsequent alterations of nucleotide sequence encoding epitopes of interest. We have produced the recombinant protein at high level, and have shown by several criteria that it possesses the chemical, physical and immunological properties of the native worm protein. Thus, we have developed a valuable system for detailed immunological studies of the structure and chemistry required for antibody binding to protein.

Strategies for epitope analysis using peptide synthesis.

A recently developed approach to the synthesis and ELISA screening of large numbers of peptides is described. The method has created the opportunity to tackle questions about the sites and specificity of antigenic determinants which were formerly thought to be too difficult to answer. The various strategies for application of this method are described along with examples of their successful use. They include a procedure for locating all the continuous antigenic peptides of a protein antigen, and the identification of non-replaceable amino acid residues within an antigenic peptide. An approach to the determination of amino acid residues involved in the epitope for any monoclonal antibody is also described. These strategies open up the prospect of rapid mapping of the antigenic properties of hitherto poorly understood antigens.

Systematic fractionation of serum antibodies using multiple antigen homologous peptides as affinity ligands.

The fractionation of polyclonal antibodies on multiple peptide ligands is described. The method is an application of a procedure for the synthesis of large numbers of peptides on individual polyethylene pins (Geysen et al., 1987). In this application, each pin-bound peptide is used as an affinity support. Antibodies bound to the peptides are then eluted, using buffers of either high or low pH. Each eluted antibody is then tested for specific binding to peptides or proteins, using ELISA procedures. A rabbit antiserum raised to gonococcal pilin was fractionated on a complete set of octapeptides homologous with the sequence of the pilin protein. Antibodies eluted from some of the peptides bound to pilin in solution. In a second example three hyperimmune sera raised to three different potyviruses were fractionated on their respective homologous peptide sequences. Testing the eluted antibodies on the three virus coat proteins revealed peptides which bound cross-reacting antibodies. Thus the method can be used to confirm direct peptide binding evidence for sequential epitopes. These peptides can then be used in affinity chromatography to increase the specificity of polyclonal sera. This can be achieved either by elution of the specific antibody from the peptide or by removal of cross-reacting antibodies from the whole serum by absorption on peptide.

An assessment of prediction methods for locating continuous epitopes in proteins.

Many attempts have been made to predict the position of antigenic sites in proteins from certain features of their primary structure. Parameters such as the hydrophilicity, static accessibility and mobility of short segments of polypeptide chains have been correlated with the location of continuous epitopes in proteins. Relative scales describing the structural propensity of each of the 20 amino acids have been derived and these are commonly used for constructing structural prediction profiles and for locating the position of epitopes. The predictive value of algorithms based on eight such scales has been compared in the present study, using as antigenicity data base the location of 29 continuous epitopes in four model proteins. A chi 2 statistical analysis showed that a segmental mobility scale and a hydrophilicity scale based on peptide retention times during chromatography gave the highest level of correct predictions.

Antigenic determinants in proteins coincide with surface regions accessible to large probes (antibody domains).

We evaluated surface areas on proteins that would be accessible to contacts with large (1-nm radius) spherical probes. Such spheres are comparable in size to antibody domains that contain antigen-combining sites. We found that all the reported antigenic sites correspond to segments particularly accessible to a large sphere. The antigenic sites were also evident as the most prominently exposed regions (hills and ridges) in contour maps of the solvent-accessible (small-probe) surface. In myoglobin and cytochrome c, virtually all of the van der Waals surface is accessible to the large probe and therefore potentially antigenic; in myohemerythrin, distinct large-probe-inaccessible, and nonantigenic, surface regions are apparent. The correlation between large-sphere-accessibility and antigenicity in myoglobin, lysozyme, and cytochrome c appears to be better than that reported to exist between antigenicity and segmental flexibility; that is, surface regions that are rigid often constitute antigenic epitopes, whereas some of the flexible parts of the molecules do not appear antigenic. We propose that the primary reason why certain polypeptide-chain segments are antigenic is their exceptional surface exposure, making them readily available for contacts with antigen-combining sites. Exposure of these segments frequently results in high mobility and, in consequence, to the reported correlation between antigenicity and segmental flexibility.

Influence of protein flexibility and peptide conformation on reactivity of monoclonal anti-peptide antibodies with a protein alpha-helix.

Monoclonal antibodies against an alpha-helical region of the iron-containing, oxygen-binding protein myohemerythrin were isolated following immunization of mice with either the whole protein or a peptide homolog of the helix. Three distinct epitopes within the myohemerythrin helix were identified. The individual residues within two of these epitopes that were essential for antibody binding were determined by measuring antibody binding to a set of peptides in which each amino acid of the epitope was replaced in turn by each of the other 19 amino acids. Hydrophilic residues that are exposed in the native conformation and buried, hydrophobic residues were both shown to be irreplaceable, suggesting their direct involvement in antibody binding. The influence of antigen conformation on antibody binding to these amphipathic epitopes was assessed by measuring the relative affinities of the antibodies for peptides, intact protein, and apoprotein. All of the antibodies bound to apoprotein better than to native protein, indicating that relaxation of the native structure by removal of the iron center increases antibody affinity for myohemerythrin. However, not all of the antibodies tested bound to peptides better than to protein, suggesting that increased antigen flexibility is not always sufficient to maximize antibody binding. Antibody binding to peptides appeared to also be influenced by the ability of the peptides to attain secondary structure at the epitopes, either alone or due to carrier influences.

The reactivity of anti-peptide antibodies is a function of the atomic mobility of sites in a protein.

To study the nature of antigenic recognition, antibodies have been prepared against a set of peptide sequences representing both highly mobile and well-ordered regions of myohaemerythrin, based on X-ray crystallographic temperature factors. Anti-peptide antibodies against highly mobile regions react strongly with the native protein; anti-peptide antibodies from well-ordered regions do not. Mobility is a major factor in the recognition of the native protein by anti-peptide antibodies; this may be of general significance in protein-protein interactions.

Location of 'continuous' antigenic determinants in the protruding regions of proteins.

A simple method is described to locate 'antigenic' peptides from the alpha-carbon co-ordinates of a protein, based on protrusion from the protein's globular surface. A good correlation is found between those parts of a protein which protrude and the experimentally determined antigenic peptides in myoglobin, lysozyme and myohemerythrin. A comparison is made between the use of protrusion index, mobility, solvent accessibility and hydrophilicity for predicting the most likely antigenic peptides.

Binding to native proteins by antipeptide monoclonal antibodies.

mAb raised against synthetic peptides derived from cholera toxin, myohemerythrin, and sickle hemoglobin were analyzed by both solid-phase and solution-phase methods. Antipeptide mAb against cholera toxin (mAb TE32 and TE33), against myohemerythrin (mAb B13I2, B13C2, and B13F2), and against sickle hemoglobin (mAb HuS-1 and HuS-2), had been previously described and used for vaccine development, structural characterization, or identification of a specific antigenic determinant, and each was apparently capable of binding both peptide and native Ag. In this study, all were found to bind whole protein when tested against immobilized Ag in a standard solid-phase assay (ELISA), yet none of the antibodies recognized the Ag in its true native form, failing to bind when tested in several solution-phase assay systems, including size exclusion HPLC. This discrepancy may be the result of modifications of the epitope created by interaction and possible denaturation of the protein on the solid-phase matrix. As a consequence, binding of these antibodies to peptides, either immobilized or in solution, or to immobilized protein, cannot be used to infer that the peptide has assumed a conformation that corresponds to that of the cognate sequence in the native protein. A re-evaluation of binding data that relates antipeptide mAb to native structural characteristics may be necessary.

Nigerythrin and rubrerythrin from Desulfovibrio vulgaris each contain two mononuclear iron centers and two dinuclear iron clusters.

The trivial name 'rubr-erythrin' is a contraction of two other trivial names: rubredoxin (ruber, red) and hemerythrin. It names a protein of undetermined biological function which putatively carries rubredoxin-like mononuclear iron and hemerythrin-like dinuclear iron. The name 'nigerythrin' (niger, black) is an analogy of rubrerythrin. It identifies a second protein of undetermined function which has prosthetic groups similar to rubrerythrin. Rubrerythrin was initially described [LeGall, J., Prickril, B. C., Moura, I., Xavier, A. V., Moura, J. J. G. & Huynh, B.-H. (1988) Biochemistry 27, 1636-1642] as a homodimer with four iron ions arranged into two rubredoxin sites and one inter-subunit dinuclear cluster. Nigerythrin is a novel protein. Here, we report that both proteins are homodimers, each dimer carrying not four but six iron ions in two mononuclear centers and two dinuclear clusters. Rubrerythrin and nigerythrin are probably both located in the cytoplasm; they are differentially characterized with respect to molecular mass, pI, N-terminal sequence, antibody cross-reactivity, optical absorption, EPR spectroscopy, and reduction potentials. All three reduction potentials in both proteins are > +200 mV. These appear too high to be of practical relevance in the cytoplasm of the sulfate reducer Desulfovibrio vulgaris (Hildenborough). We suggest the possibility of a non-redox role for both proteins with all six iron ions in the ferrous state.

Preliminary crystallographic data and primary sequence for anti-peptide Fab' B13I2 and its complex with the C-helix peptide from myohemerythrin.

Crystals of the Fab' fragment from the monoclonal anti-peptide antibody B1312 and of the Fab'-peptide antigen complex have been characterized. The monoclonal antibodies were raised against a synthetic homologue of the C-helix of myohemerythrin (residues 69-87 in myohemerythrin). The Fab'-peptide complex crystallizes in space group P6322 with unit cell dimensions a = b = 142.5 A, c = 101.5 A, alpha = beta = 90 degrees, gamma = 120 degrees, and Z = 1. The native Fab' crystallizes in space group P212121 with unit cell dimensions a = 98.0 A, b = 151.7 A, c = 80.8 A, alpha = beta = gamma = 90 degrees, and Z = 2. Both crystal forms diffract to beyond 2.6 A resolution. We also report the cDNA and predicted amino acid sequences for the variable regions of both the light and heavy chains of this anti-peptide antibody.

Altering the antigenicity of proteins.

To better understand the binding interaction between antigen and antibody we need to distinguish protein residues critical to the binding energy and mechanism from residues merely localized in the interface. By analyzing the binding of monoclonal antibodies to recombinant wild-type and mutant myohemerythrin (MHr) proteins, we were able to test the role of individual critical residues at the highly antigenic site MHr-(79-84), within the context of the folded protein. The results directly show the existence of antigenically critical residues, whose mutations significantly reduce antibody binding to the folded protein, thus verifying peptide-based assignments of these critical residues and demonstrating the ability of buried side chains to influence antigenicity. Taken together, these results (i) distinguish the antigenic surface from the solvent-exposed protein surface before binding, (ii) support a two-stage interaction mechanism allowing inducible changes in protein antigens by antibody binding, and (iii) show that protein antigenicity can be significantly reduced by alteration of single critical residues without destroying biological activity.