WO1991006562A1 - Immunological domains of hepatitis delta virus antigen - Google Patents

Abstract

In an effort to map the antigenic domains of HDAg, 209 overlapping hexapeptides, spanning the entire 214 amino acid residues of the protein, were synthesized on polyethylene pins and probed by ELISA with sera containing high titers of anti-HD antibodies. Domains recognized by antibodies present in serum from human chronic carriers of this virus included residues 2-7, 63-74, 86-91, 94-100, 159-172, 174-195 and 197-207. Oligopeptides 15 to 29 residues in length and representing epitopes of HDAg found to be dominant in man (residues 2-17, 156-184 and 197-211) were synthesized in bulk and found to possess significant antigenic activity by microtiter ELISA. The reacativity of the 197-211 peptide with human sera confirms that the entire 214 amino acids of HDAg are expressed during infection in vivo. The aforementioned peptides are useful as diagnostic reagents and as vaccines.

Description

____-IONO__OGIC-_L DOMAINS OF HEPATITIS DELTA VIRUS ANTIGEN

BACKGROUND OF THE INVENTION

Field of the Invention

This invention relates to synthetic peptides useful as diagnostic reagents in the detection of Hepatitis delta virus infection, as well as, in the form of an immunogenic conjugate, as vaccines.

Information Disclosure Statement

Hepatitis delta virus (HDV) is a unique human pathogen which has been implicated, in association with hepatitis B virus (HBV) , as a cause of severe acute hepatitis and progressive chronic liver disease (7) . The HDV virion consists of a circular, single-stranded RNA genome approximately 1.7 ilobases in length (17,19,26,27) and a highly basic phosphoprotein (hepatitis delta antigen, or HDAg) (4,5) packaged within an envelope comprised of hepatitis B surface antigen (HBsAg) (3,24). HBsAg present in the HDV virion is encoded by HBV DNA, and current data indicate that HBV - confection is an absolute prerequisite for significant infection and disease associated with HDV in man.

Complete genomic sequences have been independently derived from cDNA clones established from two different HDV strains, one after chimpanzee passage (26,27) and the other directly from human serum (19) . Analysis of these sequences has shown that virion RNA contains 2 open reading frames capable of encoding proteins of longer than 100 amino acids, while the antigenomic sequence contains 3 additional open reading frames of equivalent length (19,26,27). Thus, the HDV genome potentially encodes up to 5 different proteins of significant length. HDAg has been shown to be encoded by one of the open reading frames in antigenomic sense RNA (5,28): ORF 5 according to the nomenclature of Wang et al. (26,27), or ORF-2 according to Makino et al. (19) . Thus, HDV is a negative- stranded RNA virus. HDAg is present in HDV particles in two distinct forms having molecular weights of 27,000 (p27d) and 24,000 daltons (p24d) (3,28). Present knowledge of the differences between these two HDAg species is incomplete. ORF 5 potentially encodes a protein 214 amino acids in length, thought to represent p27d (19,26,27). It has been suggested that the p24d form of HDAg may represent expression of a C- terminal truncated protein of 195 amino acids, due to heterogeneity in the HDV genome reflected by the presence of an amber stop codon in some cloned cDNAs (28) . Alternatively, the two molecular species may reflect posttranslational processing of the primary ORF 5 expression product. HDAg is phosphorylated at multiple serine residues, and has been shown to have RNA-binding activity of uncertain specificity (5) . Thus far, it is the only protein product clearly recognized to be expressed by the HDV genome during in vivo infection. Further understanding of the structure and function of this viral protein is central to unraveling the molecular events involved in HDV replication and the pathogenesis of delta hepatitis.

Houghton, EP Appl. 251,575 sequenced 0RF5 and set forth a predicted amino acid sequence for the encoded protein. They suggested that this protein, or antigenic fragments thereof, might be useful for HDV diagnosis and vaccination. However, no epitope mapping was in fact performed and Houghton provides no guidance as to which subsequences of ORF5 might be antigenic determinants of HDV.

One traditional method of mapping the antigenic determinants (epitopes) of protein antigens was to laboriously prepare a large number of well-characterized chemical derivatives and peptide fragments from the original protein and then to test the derivatives for immunological activity. Developments in the art of solid state peptide synthesis have provided an alternative approach to obtaining potentially antigenic subsequences of a known protein antigen. By relaxing the usual criteria for quantity and purity, a large set of overlapping peptides corresponding to a larger protein may be prepared (11-12,29).

An important limitation of the pin-based oligopeptide approach to epitope mapping is related to the purity of the peptides synthesized on pins. As these peptides are not cleaved from the pins following their synthesis and undergo no physical purification steps, impurities due to unplanned side chain reactions or other irregularities in the synthesis accumulate with the addition of residues to the immobilized peptide. Thus, there are inherent limitations to the length of peptides that can be usefully synthesized in this fashion. Indeed, its developers have remarked that "because of the uncertainty of the final purity of the peptide on any given rod, the methods suffers from the disadvantage that a negative result cannot be taken as proof of the absence of antibody able to bind that nominal sequence." They also emphasize the importance of highly selective antibodies, preferably monoclonal antibodies, for screening purposes. Monoclonal antibodies against HDV delta antigen are not yet commercially available.

It should also be considered that if hexapeptides are prepared, so as to minimize the purity problem, epitopes may be overlooked since 6-mers are the smallest peptides that would be expected to have a reasonable probability of preserving antigenic activity (12) . Even so, a hexapeptide in isolation will not necessarily assume the exact same conformation that it does when it is a subsequence of a protein, so its affinity for antibodies may be different than that of the corresponding site on the protein. Finally, while the assay identifies potential epitopes, it tells nothing about their uniqueness.

A number of mathematical methods have been developed for using amino acid sequence data to locate protein determinants. These look, for example, at local average hydrophilicity (14) . These methods have their limitations, however. Hopp (14) , for example, was only confident of the antigenicity of his peak hydrophilic peptide; the second and third highest peaks resulted in a mixture of correct and incorrect assignments. Gamier (10) , reviewing Chou and Fasman's (6) method of predicting the secondary structure, found that it was only about 60% accurate in assigning residues to one of four conformational states. These deficiencies are attributable partially ^'to the relatively small size of the database used to parametrize the models, and partially to the fact that the assumption that these characteristics are purely locally determined is simplistic.

SUMMARY OF THE INVENTION

Surprisingly, we have found that the major antigenic domains of the human hepatitis delta virus antigen which are recognized by human antibodies are found in the less hydrophilic carboxyl region (residues 145-214) of the protein. In particular, the domains in the region 159-207 appear dominant.

These domains may be used in the form of synthetic oligopeptides as diagnostic reagents. Many different assay formats exist, and the oligopeptides may be labeled or immobilized as required by the format.

The oligopeptides may also be conjugated to an i_araunoge_-ic carrier and used to elicit an HDV-specific immune response to the oligopeptide. The monoclonal or polyclonal antibodies produced after such immunization may likewise be used as diagnostic reagents. Additionally, the immunogenic conjugates may be used as vaccines.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1: Analysis of the predicted HDAg (19) secondary structure by the PEPTIDESTRUCTURE program (15) . The graphical output of the PLOTSTRUCTURE program, detailing hydrophilicity (Kyte-Doolittle) , surface probability (Emini) , chain flexibility (Karplus-Schulz) , antigenic index (Jameson- Wolf) , and secondary structure by methods of Chou-Fas an (CF) and Garnier-Osguthorpe-Robson (GOR) is shown.

Figure 2: Amino acid sequences of several screened HDV peptides. Note,that the leading cysteines are provided for coupling purposes and are not a part of the cognate HDV sequences. Their presence is believed to be optional, as indicated in the claims by parenthesizing them.

Figure 3: shows the results of screening six oligopeptides for binding by 23 anti-Hd positive sera and 17 antgi-HD negative sera. All sera wsere HBsAg positive. The oligopeptides corresponded to HDAg 2-17, 58-78, 82-102, 123- 143, 156-184 and 197-211.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

The approach we utilized for mapping the epitopes of HDAg involved the synthesis of 209 hexapeptides, overlapping each other by five residues and spanning the entire 214 residues that most likely comprise the p27d form of HDAg. These peptides were synthesized on polyethylene pins configured in a microtiter format, a method first described by Geysen et al. (11) that allows their direct and repeated use in ELISA tests to determine antibody binding activity. The peptides of the present invention may be prepared in any convenient manner. Preferably, they are prepared by purely chemical synthesis. For direct chemical synthesis of peptides, see Merrifield (21) . Secondly, they may be prepared by expression of the peptides or of a larger peptide including the desired peptide from a corresponding gene (whether synthetic or natural in origin) in a suitable host. These techniques may be combined. This fusion peptide may contain a cleavage site whereby the peptide of interest may be released by cleavage of the fused molecule.

Preferably, the peptide comprises a sequence substantially homologous with one of the domains of Table 2; more preferably, it is one of the peptides defined in Figure 2. However, analogous peptides with similar immunological activity are encompassed by the present invention. These peptidies may differ from those recited by one or more substitutions, insertions or deletions of amino acids, either within the peptides or at the termini. Preferred substitutions are indicated by the table below of groups of often equivalent amino acids (with single letter codes indicated in parenthesis) :

(a) Ala(A) Ser(S) Thr(T) Pro(P) Gly(G) ;

(b) Asn(N) Asp(D) Glu(E) Gln(Q) ;

(c) His(H) Arg(R) Lys(K);

(d) Met(M) Leu(L) Ile(I) Val(V);

(e) Phe(F) Tyr(Y) Trp(W); and

(f) Cys(C).

Candidate analogues may be chosen on the basis of the effect of the amino acid changes on local hydrophilicity, surface probability, flexibility, antigenic index and secondary structure, as well as on the immunoreactivity of the most homologous hexapeptides. The peptides used as reagents or vaccines preferably include at least ten amino acids of an antigenic -doaain of HDAg, and more preferably at least twenty amino acides. Smaller peptides are less likely to accurately reproduce the desired antigenic . domain in its native conformation. They preferably should not exceed 50 amino acids, and more preferably 30 amino acids, in length. This upper limit is partially a function of synthetic considerations. Additionally, it may be desirable that the peptide mimic only selected epitopes of HFAg, since some epitopes will be less cross-reactive with epitopes of non-HFAg protein than others.

The peptide may also comprise a repeated sequence corresponding to one or more of the antigenic domains identified herein, with or without a spacer region separating the repeats.

In one embodiment, a panel of peptides according to the present invention are used as diagnostic agents. As will be seen hereafter, no one peptide determinant was recognized by all of the delta positive sera considered. However, a panel comprising a small number of oligopeptides, each corresponding to a significant antigenic domain of delta antigen, is more readily prepared and characterized than is a diagnostic reagent comprising the entire 200+ amino acid delta antigen.

A peptide of particular interest, though not listed in Figure 1, is 156 (or 159) to 195 (or 212) .

The peptides may be conjugated to an immunogenic carrier, which may be, e.g., a polypeptide or polysaccharide. If the carrier is a polypeptide, the desired conjugate may be expressed as a fusion protein. Alternatively, the HDV-specific peptide and the carrier may be obtained separately and then conjugated.

Numerous enzyme immunoassay formats, labels, conjugation and immobilization techniques, etc., are disclosed in the following publications, hereby incorporated by reference herein: O'Sullivan, Annals Clin. Biochem. , 16:221-240 (1976); McLaren, Med. Lab. Sci., 38:245-51 (1981); Ollerich, J. Clin. Chem. Clin. Biochem. , 22:895-904 (1984); Ngo and Lenhoff, Mol. Cell. Biochem., 44:3-12 (1982). The disclosed peptides may be labeled or immobilized, directly or indirectly, as required by the format selected. The present invention is directed generally to the use of the disclosed haptens and antigens in assays for HDV antigens and antibodies in body fluids of infected subjects and is not limited to any particular form of immunoassay.

The label may be, without limitation, an enzyme, enzyme substrate, radioisotope, or fluorescent label and may be attached directly or via an antibody-antigen, carbohydrate- lectin or biotin-avidin bridge. If immobilized, covalent or noncovalent means may be used to associate the peptide with the desired support, which may be for example a plate, tube, dipstick, bead or particle.

The invention in certain of its preferred embodiments is further described in the Examples which follow.

Materials and Methods

Human sera

Sera were collected from hemophiliac patients who were chronic carriers of HBV (HBsAg-positive) and enrolled in a longitudinal study exploring the contribution of HDV infection to chronic liver disease in hemophilia. Sera with uniquely high anti-HD antibody titers were chosen for these studies to enhance the signal to noise ratio in epitope screening by hexapeptide ELISA. These sera represented a select group (less than 10%) of all anti-HD positive human sera. Preferably, the anti-HD titer (measured as for Table 1) is at least about 1:12,500 (cp. AF043) . Sera having lesser titers of anti-HD antibody provided no useful information in hexapeptide ELISA assays. Control sera were obtained from two healthy individuals of comparable age and sex, one of whom was an anti- HD negative, chronic HBsAg carrier. Additional sera were collected from a woodchuck (Marmota monax) , WC862, which was positive for woodchuck hepatitis virus (WHV) and experimentally superinfected with human HDV. WHV is an hepadnavirus that is closely related to HBV and capable of supplying helper functions necessary for the replication of HDV (22) . The anti- HD activity of these sera was determined by a commercially available competitive ELISA (Delta EIA, Abbott Laboratories, N. Chicago, IL) . HBsAg was also detected in sera by a commercial ELISA (Auszyme, Abbott Laboratories) .

Anti-HD radioimmunoassay

Additional testing for anti-HD was carried out by a microtiter solid-phase competitive radioimmunoassay modified from that described by Rizzetto et al. (25) . HDAg employed in this assay was extracted from the liver of an acutely superinfected, WHV-positive woodchuck (WC643) by preparing tissue homogenates in 6 M guanidine HC1 (pH 6.0), followed by dialysis against phosphate buffered saline (PBS) , as described by Bergmann and Gerin (1) . Flexible polyvinyl chloride microtiter plate wells were coated for 2 hr at 35°C with 100 μl of a human anti-HD positive serum (AI035) diluted 1:4000 in 50 mM carbonate buffer, pH 9.6, washed with PBS containing 0.05% Tween 20 (PBS-T) , and loaded with 40 μl of the HDAg preparation. After 2 h incubation at 37°C, the plates were washed with PBS-T, and 50 μl of a 1:100 dilution of test serum in PBS was added to each well. Following an overnight incubation at 4°C, wells were washed with PBS-T and incubated for 4 hr at 4°C with 50 μl of [1251]-labelled IgG (50,000 cpm) , isolated from the serum of an anti-HD positive hemophiliac patient (ET87) by chromatography through DEAE-Sephacryl (Pharmacia, Piscataway NJ) , iodinated by the chloramine-T method, and diluted in 10% fetal calf serum. Immunoblot detection of anti-HD

Immunoblots were carried out with HDAg prepared from woodchuck serum, as this source provides antigen of less complexity than that present in liver (1). Serum (1.0 ml) taken from an acutely superinfected woodchuck (WC643) was layered over an 11 ml cushion containing 20% sucrose in 0.02 M HEPES (pH 7.4), 0.01 M CaCl₂, 0.01 M MgCl₂ and 0.1% bovine serum albumin, and centrifuged for 5 hr at 150,000 x G in an SW40 rotor (Beckman, Palo Alto CA) . Pelleted HDAg was resuspended in distilled water and stored at -70°C until use. Antigen diluted in sample buffer (0.0625 M Tris, pH 6.8, 1% sodium dodecyl sulfate, 1% 2-mercaptoethanol, 10% glycerol, 0.002% bromophenol blue) was separated by SDS-PAGE with stacking and separating gels containing 4% and 12.5% polyacrylamide respectively. Separated polypeptides were electrophoretically transferred to nitrocellulose paper at 85 A for 3 hr at 4°C. Nitrocellulose papers were blocked with blocking buffer (3% milk, 50mM Tris, 150mM NaCl, 5mM EDTA, 0.25% gelatin, 0.05% NaN₃, pH 7.4) for 30 min, and incubated with test serum diluted 1:1000 in blocking buffer for 1 hr at room temperature. After washing in PBS-T, nitrocellulose papers were incubated for 30 min with horseradish peroxidase (HRP)-conjugated goat anti-human IgG (Bethesda Research Laboratories, Gaithersberg, MD) diluted in PBS. Following an additional washing step, the papers were placed in freshly prepared substrate solution (25 mg 3,3-diaminobenzidine in 50 ml 0.05 M Tris, 0.02% hydrogen peroxide, pH 7.4) for color development.

Pin-based oligopeptide synthesis

Overlapping hexapeptides spanning the entire HDAg molecule were synthesized on polyethylene pins with materials provided as components of the "Epitope Mapping Kit" manufactured by ambridge ^"Research ^"Biochemicals, Inc. (Valley Stream, NY) . This method of peptide synthesis employs preformed active ester coupling with 9-fluoroenylmethyl- oxycarbonyl (FMOC) and t-butyloxycarbonyl (TBOC) protection. Its application to epitope mapping has been extensively described by Geysen et al. (11-13) .

Pin-based oligopeptide ELISA

Oligopeptide-bearing pins, after sonication as described below, were blocked by incubation for 1 hr in microtiter plates containing 200 ul per well blocking buffer at room temperature. Pins were then transferred to a microtiter plate containing 175 ul per well of test serum diluted 1:6000 or more in blocking buffer. Following incubation at 4°C for 24 to 48 hr, pins were subjected to 3 cycles of washing in PBS-T, 30 min per cycle with agitation, and transferred to wells containing HRP-conjugated goat anti-human IgG diluted in blocking buffer without sodium azide. Woodchuck antibody was detected by sequential incubations with rabbit antiserum raised to woodchuck immunoglobulin, and HRP-conjugated goat anti- rabbit IgG. After washing as above, pins were placed in microtiter wells containing 150 μl of freshly prepared substrate solution (25 mg 2,2'-azino-di-[3-ethylbenzthiazoline sulfonate (6)] (ABTS, Boehringer-Mannhei Biochemicals, Indianapolis, IN) in 50 ml of 0.1 M disodium hydrogen phosphate, 0.08 M citric acid, 0.015% hydrogen peroxide, pH 4.0) and held at room temperature in the dark for 30 min. Color development was stopped by removing the pins and the absorbance of the substrate solutions determined immediately by reading in an automated ELISA plate reader at 405 n . Data were downloaded to an IBM-XT computer and analyzed using Lotus 1-2-3 software. Reactions with human sera were considered positive if absorbance was > 0.2, while woodchuck serum reactions were considered positive if absorbance was > 0.5. Pins were stripped of immunoglobulin by sonication for 30 min in 1% sodium dodecyl sulfate, 0.1% 2-mercaptoethanol and 0.1 M sodium dihydrogen orthophosphate at 60°C, washed with hot (60°C) distilled water followed by boiling methanol for 2 min, and air dried.

Bulk synthesis of oligopeptides

Peptide synthesis was carried out with a Biosearch Model 9500 Peptide Synthesizer (Novato CA) . TBOC chemistry was employed for peptide synthesis (21) .

Oligopeptide ELISA for anti-HD

Oligopeptides (750 ng in 75 μl carbonate buffer) were applied to the wells of flat-bottom ELISA plates (Falcon) by incubation for 2 hours at 35°C in a humidified chamber. Plates were washed with PBS-T and blocked by the addition of 75 μl blocking buffer. Serum specimens diluted in blocking buffer (50 μl per well) were added to the plates which were then incubated overnight at 4°C. Following washing with PBS-T, 50 μl of HRP-conjugated goat anti-human IgG diluted in blocking buffer without sodium azide was added for 50 min at 37°C. Plates were washed with PBS, followed by the addition of o- phenylenediamine substrate solution (55 μl per well) at room temperature for 15 min in the dark. Color development was stopped by the addition of 160 μl of 1 N H2SO4, and the absorbance at 490 nm determined by reading in an automated ELISA plate reader.

HDAg secondary structure predictions

Hydrophilic domains of HDAg were predicted by the method of Hopp and Woods (14) , while hydrophobic regions were independently predicted by a program which determined the mean hydrophobicity score for 11 residue windows using hydrophobic indices derived- for individual side chains by Fauchere and Pliska (9) . More detailed predictions of the secondary structure of HDAg were done by means of the PEPTIDESTRUCTURE and PLOTSTRUCTURE programs of the University of Wisconsin Genetics Computer Group Sequence Analysis Software Package (15) . These programs include predictions of hydrophilicity according to the method of Kyte and Doolittle (18) , peptide secondary structure according to Chou and Fasman (6) and Gamier et al. (10) , and surface probability according to Emini et al. (8) . From these predictive measures, an antigenic index (Al) is derived by the algorithm developed by Jameson and Wolf (15) . The Al is a measure of the probability that a domain is antigenic, and is calculated by summing weighted values for surface accessibility, regional backbone flexibility, and certain features of predicted secondary structure. In addition, regions predicted to be amphipathic alpha helices (putative T cell determinants) were identified by a program developed by Margalit et al. (20) .

Results

HDAg epitope mapping with overlapping hexapeptides

We synthesized 209 overlapping hexapeptides spanning the 214 amino acids of human HDAg (ORF-2 expression product) predicted from the nucleotide sequence of cloned HDV cDNA repαrted by Makino et al. (19) . Hexapeptides were synthesized on polyethylene pins using a synthesis system supplied by Cambridge Research Biochemicals. Control peptides (4-mers) included in the synthesis reacted specifically with murine monoclonal antibody supplied by the manufacturer in subsequent pin-based ELISA tests (mean signal to noise (S/N) ratio was 4) , confirming the validity of the synthesis. However, preliminary pin-based ELISA tests demonstrated substantial non-specific binding of human IgG to pins supporting HDAg hexapeptides at test serum dilutions below 1:2000. To overcome this problem, we determined the endpoint titer of each of a panel of 27 anti-HD positive sera in a competitive radioimmunoassay which employed as antigen HDAg extracted from woodchuck liver. These sera had been collected from hemophiliac patients who were chronic carriers of HBsAg. We selected for further study 5 sera with particularly high titers of anti-HD, ranging from 1:12,500 to > 1:312,500 (Table 1) . These titers of anti-HD are highly suggestive of chronic infection with HDV. Each of these 5 sera were strongly positive in immunoblot assays against denatured HDAg concentrated from woodchuck serum, reacting with both p24d and p27d (Figure 1) . We thus considered it likely that these sera had high titer antibody against sequential epitopes of HDAg that would be detectable in hexapeptide ELISA assays. Each serum was tested at a dilution of 1:6000 to 1:8000 against the array of HDAg hexapeptides.

Antibodies present in this serum (AC036) demonstrated binding activity against hexapeptides representing at least six discrete regions in the linear sequence of the HDAg protein. Epitopes recognized by binding of antibody to hexapeptides spanned residues 2-7, 63-69, 159-165, 167-172, 174-181, and 201-207. Additional hexapeptides demonstrating low level binding (above background, but less than the arbitrary cut-off of absorbance of > 0.2) represented residues 85-91, and 152-157 (Figure 2) . These hexapeptides may represent minor determinants. Hexapeptide ELISA results with this positive serum were highly reproducible, while two anti-HD negative sera (A008 and A073, Table 1) yielded consistently negative results (mean OD490 0.113 and 0.121, and maximum OD490 0.176 and 0.157, respectively) when tested against the hexapeptide-bearing pins at similar dilutions (results not shown) . Results of assays with alternating positive and negative sera demonstrated that the stripping procedure was effective in completely removing IgG bound to the pins. In the course of these studies, the hexapeptide-bearing pins were reused through 26 repetitive ELISA cycles without noticeable loss of activity. Subsequent screening of HDAg hexapeptides with four other high-titer anti-HD positive human sera identified identical or closely positioned epitopes within the protein, as well as two additional antigenic domains. Antigenic domains of HDAg were defined as regions containing overlapping or contiguous hexapeptides found to be antigenic in screening with any of the 5 human anti-HD positive sera. Altogether, these domains included regions spanned by residues 2-7, 63-74, 86-91, 94-100, 159-172, 174-195, and 197-207 (Table 2). Although the carboxyl terminal 50 residues of the HDAg protein (residues 159-207) appeared immunodominant with each of the human sera tested in these assays, significant variation was evident between individual sera with respect to the degree to which certain hexapeptides were bound by antibody. Among the 5 anti- HD sera tested, AD099 demonstrated relatively low level binding of antibodies to HDAg hexapeptides (maximum OD490 of 0.398, compared with 0.858-2.272 obtained with the other four sera). These results suggest that the sequential epitopes recognized by AD099 antibodies in HDAg immunoblots may be poorly mimicked by synthetic peptides as short as six residues in length.

Mapping of HDAg epitopes recognized by the woodchuck

Hexapeptide ELISA tests were carried out with serum from a WHV-carrier woodchuck (WC862) which was acutely superinfected with HDV. Serum was collected following HDAg clearance from serum and the appearance of anti-HD. The second antibody employed in these hexapeptide ELISAs, rabbit anti- woodchuck immunoglobulin, generated higher background activity and required a different standard for positivity (OD490 > 0.5). The results of these tests indicated that the epitopes bound by woodchuck antibodies (residues 1-7, 63-71, 121-128, and 197- 204) overlap with at least some of the antigenic domains recognized by humans (residues 2-7, 63-74, and 197-207). However, in contrast to the apparent immunodominance of the carboxyl terminal region of HDAg in humans, HDAg domains defined by hexapeptides spanning residues 1-7 and 63-71 were dominant in this one woodchuck. In addition, woodchuck antibodies bound also to hexapeptides spanning residues 121- 128, a region not recognized by any of the human sera tested (Table 2) . These results suggest significant inter-species differences in the recognition of HDAg epitopes.

Comparison of hexapeptide ELISA with computer algorithm predictions

We analyzed the predicted amino acid sequence of HDAg for regions of relative hydrophilicity and hydrophobicity using the method of Hopp and Woods (14) and values of relative hydrophobicity developed for individual amino acid side chains by Fauchere and Pliska (9) . Although these two methods yield somewhat different predictions of the hydrophilic and hydrophobic domains of the protein, both suggest the carboxyl terminal region (residues 145-214) of HDAg is less hydrophilic than the amino terminal 144 residues (not shown) . Neither analysis would have predicted the antigenic domains occurring within residues 159-207 that appear dominant in hexapeptide ELISAs with human sera.

Predictions of the secondary structure of HDAg made by the PEPTIDESTRUCTURE program (15) are shown in Figure 6.

To determine the extent of correlation between residues predicted to be antigenic by this program and those observed to part of antigenic domains in hexapeptide ELISA assays, a statistical analysis was carried out. For each residue within the protein, the observed antigenic activity (i.e., inclusion within an antigenic domain) was cross- classified against predicted antigenic activity derived by each of the five separate algorithms contained within PEPTIDE STRUCTURE (15) . The prediction criteria chosen for this analysis included (a) nydrophilicity score (IS) greater than or equal to 1.3, (b) surface probability score (8) greater than or equal to 5, (c) flexibility score greater than or equal to 1.04, (d) antigenic index (15) greater than or equal to 1.2, and the presence of a turn ("T" or "t") predicted by Chou- Fasman (6) or method of Gamier (10) . Each prediction method was found to have relatively poor sensitivity or specificity (Table 7) . Regardless of the prediction rule, the kappa statistic, which measures chance-adjusted agreement, never exceed 0.21, indicating poor agreement between the methods.

Finally, although regions predicted to be amphipathic alpha helices are generally considered predictive of T cell and not B cell determinants (20) , it was of interest that 5 of 9 predicted amphipathic segments (residues 117-120, 131-133, 150- 159, 170-177 and 179-183) were adjacent to or overlapped with antigenic sites determined in hexapeptide ELISAs.

Microtiter oligopeptide ELISA for anti-HD

Based on the results of hexapeptide ELISA tests with human sera, four oligopeptides representing HDAg residues 2-17, 156-184, 167-184, and 197-211 were synthesized in bulk and tested for antigenic activity by a microtiter ELISA. Maximal antigenic activity was associated with the largest peptide, representing residues 156-184, which demonstrated substantial antigenic activity with as little as 75 pg of peptide applied in solution to ELISA plate wells. The 156-184 peptide contains an internal tri-peptide sequence of 3 glycine residues (164- 166) suggesting that it may comprise two distinct functional domains; thus peptide 167-184 was synthesized to represent the carboxyl domain of this peptide. The results suggested that domains on either side of the three glycine residues contribute to the antigenic activity of the 156-184 peptide. Peptide 167- 184 was strongly antigenic, although less so than the 156-184 peptide. It should be noted, however, that the hexapeptide APGGGF (162-167) had significant antigenic activity in hexapeptide screening assays. Less antigenic activity was evident with a peptide representing residues 197-211, and only limited activity was found with the 2-17 peptide. The degree of reactivity of each of the human sera with individual peptides was proportionate to the degree of binding of antibodies in pin-based hexapeptide ELISAs (Table 3) . Serum AD099 did not react with any of the four abovementioned bulk- synthesized peptides, consistent with the fact that its reactivity with hexapeptides suggested dominant (but relatively weak) reactivity in the region 180-193. Figure 3 presents bar charts illustrating the reactivity of oligopeptides comprising amino acids 2-17, 58-78, 82-102, 123-143, 156-184 or 197-211 of HDAg with anti-HD positive and negative sera. Study of these charts reveals that many anti-HD positive sera react with more than one antigenic domain of HDAg. For example, positive serum 6 reacted strongly with oligopeptides 58-78, 82-102, 123-143 and 156-184.

Oligopeptide 2-17 was the least reactive of the six peptides treated in Figure 3. Additionally, oligopeptide 197- 211 reacted strongly only with positive serum 2. Oligopeptide 156-184 wa the most reliable antigen, but several sera reacted much more strongly with 82-102 than with 156-184. Oligopeptide 123-143 showed a pattern of reactivity similar to that of 156- 184, but with fewer strong reactions.

None of the six peptides evinced significant reactivity with the seventeen anti-HD negative sera.

The data presented in Table 3 and in Figure 3 indicate that synthetic peptides might be useful diagnostic reagents that could be employed in new assays for anti-HD antibodies. Studies with the 156-184 peptide suggest that it is recognized by IgG antibody present in approximately 90% of anti-HD positive sera, and that it may provide the basis for an antibody test that is significantly more sensitive than existing competitive im unoassays. Further studies of the immune response to this and other HDAg peptides will help in determining whether there are differences in the fine structure (e.g., disease stage) of the B cell response to this protein in acute and chronic HDV infections, and in assessing the isotype specific responses to this antigen.

Although current understanding of the HDV-εpecific immunity is limited, previous infection of hepatitis B-positive chimpanzees with HDV has been shown to result in protection against severe hepatitis following rechallenge to HDV (23) . Although it is possible that antibodies to HDAg may play a role in mediating such protection (as antibodies to hepatitis B core protein may offer some protection against hepatitis B infection) , it is likely that T-cell-mediated immunity is also important (particularly cytotoxic T-cells) . The antigenic peptides described herein may thus be capable of providing protection against HDV disease if administered as immunogenic conjugates with carrier proteins such as hepatitis B core protein. Because peptide 156-184 contains predicted amphipathic helical structures (20) at residues 150-159, 170- 177 and 179-183, we determined its ability to induce antibody as a free peptide and as a peptide coupled to keyhole limpet hemocyanin (KLH) (Table 8) . The peptide and peptide conjugate were administered to New Zealand white rabbits using Freund's adjuvant (complete and incomplete) . The demonstrated immunogenicity of peptide 156-184 in the absence of coupling to a carrier protein suggests that this peptide contains important T-cell determinants as well as B-cell epitopes, and thus might be a particularly useful component in an HDV vaccine.

Table 1. Anti-HD Positive and Control Human Sera.

* Highest dilution yielding 50% inhibition in competitive irihibition radioimmunoassay with woodchuck liver-derived HDAg (ccmparative endpoint titer in ccmmercial ELISA) .

Table 2. Antigenic Domains of HDAg Defined by Overlapping

Hexapeptides with Antigenic Activity

Antigenic Domains of HDAg

Serum 1-7 63-74 86-92 94-100 121-128* 159-172 174-195 197-207

3 2

3 3 2 2 5 3

5

Note: Antigenic domains were defined by overlapping or con¬ tiguous hexapeptides with antigenic activity determined by screening against the panel of human sera (see Figure 3) . Shewn in the table are the number of hexapeptides within each domain that were determined to have antigenic activity with each individual serum specimen (absorbance > 0.2 for human sera, or > 0.5 for woodchuck serum) .

* Recognized only by woodchuck serum.

Table 2A. Comparison of Antigenic Domains and Tested Peptides

Antigenic Substantially Homologous

Domain Peptides Tested

HD(l-7) (Cys)-HD(2-17)

HD(63-74) (Cys)-HD(58-78)

HD(86-92) (Cys)-HD(82-102)

HD(94-100) ditto

HD(121-128) (Cys)-HD(123-143)

HD(159-172) (Cys) -HD(156-184)

HD(167-184)

HD(174-195) none tested

HD(197-207) HD(198-212)

Table 3. Detection of Anti-HD by Synthetic Peptide E ESA

Peptide Antigen

Serum 2-17 156-184 167-184 197-211

AD062 .07 (<10²)* >2.95 (>10⁶) >2.95 ( 10⁵) 0.34 ( 10²)

AF043 .06 (<10²) >2.95 ( 10⁵) 1.29 ( 10³) 0.28 ( 10²)

AC036 .33 ( 10²) >2.95 (>10⁶) >2.95 (>10⁵) 2.53 ( 10²)

AC039 .09 (<10²) >2.95 (>10⁶) >2.95 ( 10⁵) >2.95 ( 10⁴)

*OD490 at 1:100 serum dilution (endpoint titer)

Table 4: Predicted Antigenic Sites of HDAg

PEPTIDESTRUCTURE of: humanhdv.pep check: 6004 from: 1 to: 214

TRANSLATE of: human.frg /rev check: 1646 from: 1 to: 645

Hydrophilicity (Kyte-Doolittle) averaged over a window of: 7

Surface Probability according to Emini

Chain Flexibility according to Karplus-Schulz

Secondary Structure according to Chou-Fasman

Secondary Structure according to Garnier-Osguthorpe-Robson

Antigenicity Index according to Jameson-Wolf

Position numbers corresponding to preferred peptides of Figure 1 are boldfaced. Asterisks mark the antigenic domains of Table 2.

Pos AA HyPhil SurfPr FlexPr CF-Pred GORPred Al-Ind

Table 5: HDAg Hexapeptide Screening Results

Serum

Table 6. Examples of Two Antigenic Domains of HDAg Determined fcy Pin-Based Hexapeptide ELISA

A.

Reactive

S RGAP GGG FVP SMQ OD490 Sera

S R G A P G R G A P G G

A P G G G F

G F V P S M

FVP SMQ

B. VP E S P FART G E G L D I RG S Q G F P

V P E S P

P E S P

E S P

S P

Antigenic domains were defined by overlapping or contiguous peptide sequence found to be reactive as hexapeptides when tested against any anti-HD positive serum. Only reactive hexapeptides are shewn. Where the hexapeptide was reactive with more than one of the tested sera, the OD490 represents the mean absorbance: A) antigenic domain at residues 159-172, B) antigenic domain at residues 174-195. See also Table 2. Table 7. Cc.rpari.son of Predicted and Observed Antigenic Residues of HDAg

Sens = Sensitivity; Spec = Specificity

Table 8. Dάπmunogenicity of HDV Peptides and Peptide Conjugates

Anti-peptide

Rabbit Immunogen (cognate peptide) Western Blot

LP-1 HD2-17/KLH >=1:156,000 neg LP-2 HD156-184/KLH >=1:156,000 + LP-3 HD197-211/KEH >=1:156,000 neg IP-4 HD156-184 >=1:1250 + LP-5 HAV-VP4 n.d neg

Imrnunogens were given to rabbits in Freund's complete adjuants, followed by subsequent booster immunizations in Freund's incomplete adjuvant. Rabbit LP-5 received a control peptide representing the putative UP4 capsid protein of hepatitis A virus. Sera collected from immunized rabbits were tested in ELISA assays against the cognate peptide (uncoupled) and by Western blot (1:100 serum dilution) against HDAg purified from an infected woodchuck.

References

1. Bergmann, K.F., and J.L. Gerin. 1986. Antigens of hepatitis delta virus in the liver and serum of humans and animals. J. Infect. Dis. 154:702-706.

2. Berzofsky, J.A. 1985. Intrinsic and extrinsic factors in protein antigenic structure. Science 229:932-940.

3. Bonino, F., K.H. Heermann, M. Rizzetto, and W.H. Gerlich. 1986. Hepatitis delta virus: protein composition of delta antigen and its hepatitis B virus-derived envelope. J. Virol. 58:945-950.

4. Bonino, F. , B. Hoyer, J.W.-K. Shih, M. Rizzetto, R.H. Purcell, and J.L. Gerin. 1984. Delta hepatitis agent: structural and antigenic properties of the delta-associated particle. Infect. Immun. 43:1000-1005.

5. Chang, M.-F., S.C. Baker, L.H. Soe, T. Kamahora, J.G. Keck, S. Makino, S. Govindarajan, and M.M.C. Lai. 1988. Human hepatitis delta antigen is a nuclear phosphoprotein with RNA-binding activity. J. Virol. 62:2403-2410.

6. Chou, P.Y., and G.D. Fasman. 1978. Prediction of the secondary structure of proteins from their amino acid sequence, p.45-148. In A. Meister (ed.), Advances in Enzymology and Related Areas of Molecular Biology, Volume 47. John Wiley & Sons, New York City.

7. De Cock, K.M. , S. Govindarajan, and A.G. Redeker. 1987. HDV infection in the Los Angeles area, p.167-179. In M. Rizzetto, J.L. Gerin, and R.H. Purcell (eds.), The hepatitis delta virus and its infection. Alan R. Liss, Inc., New York City. 8. Emini, E.A. , J.V. Hughes, D.S. Perlow, and J. Boger. 1985. Induction of hepatitis A virus-neutralizing antibody by a virus-specific synthetic peptide. J. Virol. 55:836-839.

9. Fauchere, J.-L., and V. Plishka. 1983. Hydrophobic parameters P of amino-acid side chains from the partitioning of N-acetyl-amino-acid amides. Eur. J. Med. Chem. 18:369-375.

10. Gamier, J. , D.J. Osguthorpe, and B. Robson. 1978. Analysis of the accuracy and implications of simple methods for predicting the secondary structure of globular proteins. J. Mol. Biol. 120:97-120.

11. Geysen, H.M. , R.H. Meloen, and S.J. Barteling. 1984. Use of peptide synthesis to probe viral antigens for epitopes to a resolution of a single amino acid. Proc. Natl. Acad. Sci. USA 81:3998-4002.

12. Geysen, H.M. , S.J. Rodda, T.J. Mason, G. Tribbick, and P.G. Schoofs. 1987. Strategies for epitope analysis using peptide synthesis. J. Immunol. Methods 102:259- 274.

13. Geysen, H.M. , J.A. Tainer, S.J. Rodda, T.J. Mason, H. Alexander, E.D. Getzoff, and R.A. Lerner. 1988. Chemistry of antibody binding to a protein. Science 235:1184- 1190.

14. Hopp, T.P., and K.R. Woods. 1981. Prediction of protein antigenic determinants from amino acid sequences. Proc. Natl. Acad. Sci. USA 78:3824-3828.

15. Jameson, B.A., and H. Wolf. 1988. The antigenic index; a novel algorithm for predicting antigenic determinants. CABIOS 4:181-186. 16. Kuo, M.-Y.P., M. Chao, and J. Taylor. 1989. Initiation of replication of the human hepatitis delta virus genome from cloned DNA: role of delta antigen. J. Virol. 63:1945-1950.

17. Kuo, M.-Y.P., J.D. Goldberg, L. Coates, W.S. Mason, J.L. Gerin, and J. Taylor. 1988. Molecular cloning of hepatitis delta virus RNA from an infected woodchuck liver: sequence, structure, and applications. J. Virol. 62:1855- 1861.

18. Kyte, J. , and R.F. Doolittle. 1982. A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 157:105-132.

19. Makino, S., M.-F. Chang, C.-K. Shieh, T. Kamahora, D.M. Vannier, S. Govindarajan, and M.M.C. Lai. 1987. Molecular cloning and sequencing of a human hepatitis delta (d) virus RNA. Nature 329:343-346.

20. Margalit, H. , J.L. Spouge, J.L. Comette, K.B. Cease, C. DeLisi, and J.A. Berzofsky. 1987. Prediction of immunodominant helper T cell antigenic sites from the primary sequence. J. Immunol. 138:2213-2229.

21. Merrifield, R.B. 1963. Solid phase peptide synthesis. I. The synthesis of a tetrapeptide. J. Am. Chem. Soc. 85:2149-2154.

22. Ponzetto, A., P.J. Cote, H. Popper, B.H. Hoyer, W.T. London, E.C. Ford, F. Bonino, R.H. Purcell, and J.L. Gerin. 1984. Transmission of the hepatitis B virus- associated d agent to the eastern woodchuck. Proc. Natl. Acad. Sci. USA 81:2208-2212. 23. Purcell, R.H. , W.C. Satterfield, K.F. Berg ann, A. S edile, A. Ponzetto, and J.L. Gerin. 1987. Experimental hepatitis delta virus infection in the chimpanzee. Prog. Clin. Biol. Res. 234:27-36.

24. Rizzetto, M. , B. Hoyer, M.G. Canese, J.W.-K. Shih, R.H. Purcell, and J.L. Gerin. 1980. d agent: association of d antigen with hepatitis B surface antigen and RNA in serum of d-infected chimpanzees. Proc. Natl. Acad. Sci. USA 77:6124-6128.

25. Rizzetto, M. , J.W.-K. Shih, and J.L. Gerin. 1980. The hepatitis B virus-associated d antigen: isolation from liver, development of solid-phase radioimmunoassays for d antigen and anti-d and partial characterization of d antigen. J. Immunol. 125:318-324.

26. Wang, K.-S., Q.-L. Choo, A.J. Weiner, J.-H. Ou, R.C. Najarian, R.M. Thayer, G.T. Mullenbach, K.J. Denniston, J.L. Gerin, and M. Houghton. 1986. Structure, sequence and expression of the hepatitis delta (d) viral genome. Nature 323:508-514.

27. Wang, K.-S., Q.-L. Choo, A.J. Weiner, C.-Y. Ou, R.C. Najarian, R.M. Thayer, G.T. Mullenbach, K.J. Denniston, J.L. Gerin, and M. Houghton. 1987. Corrigendum: structure, sequence and expression of the hepatitis delta (d) viral genome. Nature 328:456-456.

28. Weiner, A.J., Q.-L. Choo, K.-S. Wang, S. Govindarajan, A.G. Redeker, J.L. Gerin, and M. Houghton. 1988. A single antigenomic open reading frame of the hepatitis delta virus encodes the epitope(s) of both hepatitis delta antigen polypeptides p24d and p27d. J. Virol. 62:594-599.

29. Geysen, U.S. 4,708,871, Antigenically Active Amino Acid Sequences (1987) .

Claims

1. A first peptide of ten to fifty amino acids having the sequence

(Cys-)Ser-Arg-Ser-Glu-Arg-Arg-Lys-Asp-Arg-Gly-Gly-Arg-Glu-Asp- Ile-Leu or a peptide of ten to fifty amino acids which substantially inhibits the binding of said first peptide by human anti-hepatitis delta virus antibodies.

2. A first peptide of ten to fifty amino acids having the sequence

(Cys-)Ile-Gly-Lys-Lys-Asp-Lys-Asp-Gly-Glu-Gly-Ala-Pro-Pro-Ala- Lys-Lys-Leu-Arg-Met-Asp-Gin or a peptide of ten to fifty amino acids which substantially inhibits the binding of said first peptide by human anti-hepatitis delta virus antibodies.

3. A first peptide of ten to fifty amino acids having the sequence

(Cys-)Asp-Ala-Gly-Pro-Arg-Lys-Arg-Pro-Leu-Arg-Gly-Gly-Phe-Thr- Asp-Lys-Glu-Arg-Gln-Asp-His or a peptide of ten to fifty amino acids which substantially inhibits the binding of said first peptide by human anti-hepatitis delta virus antibodies.

4. A first peptide of ten to fifty amino acids having the sequence

(Cys-)Ser-Arg-Glu-Glu-Glu-Glu-Glu-Leu-Lys-Arg-Leu-Thr-Glu-Glu- Asp-Glu-Lys-Arg-Glu-Arg-Arg or a peptide of ten to fifty amino acids which substantially inhibits the binding of said first peptide by human anti-hepatitis delta virus antibodies.

5. A first peptide of ten to fifty amino acids having the sequence

(Cys-)Glu-Gly-Gly-Ser-Arg-Gly-Ala-Pro-Gly-Gly-Gly-Phe-Val-Pro- Ser-Met-Gln-Gly-Val-Pro-Glu-Ser-Pro-Phe-Ala-Arg-Thr-Gly-Glu or a peptide of ten to fifty amino acids which substantially inhibits the binding of said first peptide by human anti- hepatitis delta virus antibodies.

6. A first peptide of ten to fifty amino acids having the sequence

Phe-Val-Pro-Ser-Met-Gln-Gly-Val-Pro-Glu-Ser-Pro-Phe-Ala-Arg- Thr-Gly-Glu or a peptide of ten to fifty amino acids which substantially inhibits the binding of said first peptide by human anti-hepatitis delta virus antibodies.

7. A first peptide of ten to fifty amino acids having the sequence

Asp-Ile-Leu-Phe-Pro-Ala-Asp-Pro-Pro-Phe-Ser-Pro-Gln-Ser-Cysor a peptide of ten to fifty amino acids which substantially inhibits the binding of said first peptide by human anti- hepatitis delta virus antibodies.

8. A peptide of ten to fifty amino acids comprising one of the following sequences,

(a) (Cys-)Ser-Arg-Ser-Glu-Arg-Arg-Lys-Asp-Arg-Gly- Gly- Arg-Glu-Asp-Ile-Leu,

(b) (Cys-)Ile-Gly-Lys-Lys-Asp-Lys-Asp-Gly-Glu-Gly- Ala- Pro-Pro-Ala-Lys-Lys-Leu-Arg-Met-Asp-Gin,

(c) (Cys-)Asp-Ala-Gly-Pro-Arg-Lys-Arg-Pro-Leu-Arg- Gly- Gly-Phe-Thr-Asp-Lys-Glu-Arg-Gln-Asp-His,

(d) (Cys-)Ser-Arg-Glu-Glu-Glu-Glu-Glu-Leu-Lys-Arg- Leu- Thr-Glu-Glu-Asp-Glu-Lys-Arg-Glu-Arg-Arg, (e) (Cys-)Glu- Gly-Gly-Ser-Arg-Gly-Ala-Pro-Gly-Gly-Phe- Val-Pro-Ser-Met-Gln- Gly-Val-Pro-Glu-Ser-Pro-Phe-Ala-Arg-Thr-Gly-Glu,

(f) Phe-Val-Pro-Ser-Met-Gln-Gly-Val-Pro-Glu-Ser-Pro- Phe-Ala-Arg-Thr-Gly-Glu, or

(g) Asp-Ile-Leu-Phe-Pro-Ala-Asp-Pro-Pro-Phe-Ser-Pro- Gln-Ser-Cys, or a peptide of ten to fifty amino acids which substantially inhibits the binding of one and only one of said peptides (a)- (g) by human anti-hepatitis delta virus antibodies.

9. In an immunoassay to detect the presence or measure the level of anti-hepatitis delta virus antibodies in a sample, the improvement which comprises use of one or more peptides according to claim 8 as an antigenic diagnostic reagent.

10. A vaccine comprising a peptide of claim 8 conjugated with an immunogenic carrier.

11. A peptide comprising a sequence substantially homologous with one and only one of the following antigenic domains of hepatitis delta antigen: 1-7, 63-74, 86-92, 94- 100, 121-128, 159-172, 174-195 and 197-207.

12. A peptide comprising at least one of the 159- 172, 174-195 and 197-207 antigenic domains of HDAg and consisting of no more than about 100 amino acids.

13. The method of claim 9 in which a plurality of such peptides are used as a diagnostic panel.

14. The vaccine of claim 10, wherein the peptide comprises a sequence substantially homologous with the hepatitis delta antigen sequence 156-184.