- Email: [email protected]
Crossreactivity of mimotopes and peptide homologues of a sequential epitope with a monoclonal antibody does not predict crossreactive immunogenicity
Vaccine 18 (2000) 284±290
www.elsevier.com/locate/vaccine
Crossreactivity of mimotopes and peptide homologues of a sequential epitope with a monoclonal antibody does not predict crossreactive immunogenicity Karim C. El Kasmi a, b, Sabrina Deroo a, Dietmar M. Theisen a, Nicolaas H.C. Brons a, Claude P. Muller a, b,* a
Laboratoire National de SanteÂ, P. O. Box 1102, L-1011 Luxembourg, Luxembourg b Medizinische FakultaÈt, UniversitaÈt TuÈbingen, D-72076 TuÈbingen, FRG
Abstract The sequence H236±256 of the measles virus (MV) hemagglutinin (H) contains the sequential epitope of the neutralizing and protective monoclonal antibody (mAb) BH129 with the minimal epitope E245L±QL249. Using this mAb, we have recently developed 7mer mimotopes binding up to 135 better than the corresponding 7mer epitope H244±250. In this study, we combined T cell epitopes (TCE) with either highly crossreactive 7mer mimotopes, 13mer mimotopes or less crossreactive MVderived B cell epitopes (BCE). Antigenicity of these TBB, TTB and TTBB peptides was determined with BH129 in a competition ELISA against MV. We found that chimeric peptides including mimotopes were up to 80 better binders to the mAb than peptides containing the original BCEs. All peptides irrespective of their antigenicity were used for immunization to compare their virus- crossreactive immunogenicity. Unexpectedly, none of the highly antigenic mimotope-based peptides induced MVcrossreactive antibodies. In contrast, a number of peptides with the viral BCE sequence that did not bind to the mAb, induced MV-crossreactive and even neutralizing antibodies. This report describes a striking example of disparity between antigenicity and crossreactive immunogenicity and casts considerable doubt on the predictive value of antigenicity in immunogenicity studies, considerably complicating the selection of potential vaccine candidates. # 1999 Elsevier Science Ltd. All rights reserved. Keywords: Measles virus; Mimotope; Synthetic peptides; Antigenicity
1. Introduction Current live attenuated vaccines have successfully reduced mortality due to measles virus (MV) infection in the developed world. In developing countries the disease still causes the death of one million children every year [1,2]. Major drawbacks of current vaccines include the lack of resistance to neutralization by maternal antibodies and their thermal sensitivity. Subunit-vaccines based on synthetic peptides that induce neutralizing and protective antibodies could po-
* Corresponding author. Tel.: +352-4-90604; fax: +352-4-90686. E-mail address: [email protected] (C.P. Muller)
tentially circumvent these problems [3,4]. Passively transferred antibodies directed against the hemagglutinin protein (H) fully protect against MV infection, emphasizing the role of antibodies even in the absence of a T cell response [5±7]. Since conformational epitopes cannot easily be mimicked by synthetic peptides the design of peptide vaccines is mostly con®ned to sequential B cell epitopes (BCE). However, random phage display libraries or solid phase combinatorial peptide libraries have been used to ®nd peptides that mimic more complex epitopes. Steward et al. reported mimotopes that generated neutralizing and/or protective antibodies against MV and respiratory syncytial virus [8,9]. Although, in the case of the MV mimotope this result is surprising,
0264-410X/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 4 - 4 1 0 X ( 9 9 ) 0 0 1 9 9 - 1
K.C. El Kasmi et al. / Vaccine 18 (2000) 284±290
285
Table 1 T and B cell epitopes used Abbreviation
Sequence
Description
T1 T2 T3 T4 B7 B8 B15 Ma7 Mb7 PMa7 Ma8 Mb8 Mo13 Ma13 Mb13
PVVEVNGVTIQVGSR LLGILESRGIKARIT DIEKKIAKMEKASSVFNVVNS QYIKANSKFIGITEL SELSQLS RSELSQLS KPNLSSKRSELSQLS LYMPQLA/S SEMPQLP PFYSHSADGALYMPQLGAAGA RLYMPQLA/S RSEMPQLP AVFVTSFLPQLSY AVFVTLYMPQLSY AVFVTSEMPQLPY
TCE (MV F421-435)[14] TCE (MV F256-270)[14] TCE (CST3 378-398; P. falcip.)[15] TCE (tt830-844; Tetanus toxoid)[16] MV sequence H244-250[10,11] MV sequence H243-250[10] MV sequence H236-250[10] mimotope; corresponding to H244-250[11] idem mimotope between phage ¯anking residues mimotope; corresponding to H243±250 idem mimotope idem idem
since the template monoclonal antibody F2.41 does not neutralize or protect, these observations emphasize the potential of mimotopes as candidate vaccines. A sequential epitope of MV±H has recently been mapped to the region H236±256 with a series of neutralizing and protective monoclonal antibodies (mAbs) including BH129 [10]. In a previous study, we have generated mimotopes against BH129 using a 6mer phage displayed random library. Most of the selected clones had little sequence homology with the original epitope but the corresponding peptides bound BH129 up to 135 better than the peptides homologous with the sequence of this sequential BCE, suggesting that the mimotopes mimicked the conformation of the viral epitope more eciently than the MV-derived peptides [11]. In the present study, we compared the ability of the mimotopes and of MV-derived peptides to induce MVcrossreactive antibodies when combined with T cell epitopes (TCE) in iteratively optimized synthetic TB peptides. Although, in these TB peptides the mimotopes were considerably better binders to the mAb than were the MV sequences, only the latter generated MV-crossreactive antibodies.
2. Materials and methods 2.1. Peptide synthesis and immunizations All peptides were synthesized by automated solid phase Fmoc chemistry on a SYRO peptide synthesizer (Multisyntech, Bochum) [12,13]. Purity was >60%, as determined by reversed phase high-performance liquid chromatography (HPLC). Peptides used in this study are shown in Table 1. Peptides containing the sequence of the viral epitope will be referred to as wild-type peptides.
Male and female speci®c pathogen-free BALB/c mice (H2d) (8±15 weeks old) were immunized twice (day 0, 14) intraperitoneally with a CFA (priming)/ IFA (boost) (Sigma, USA) emulsion of 150 mg peptide (mixed 1:1). Blood was collected seven to ten days after the boost to prepare serum. 2.2. ELISA and ¯ow cytometry All ELISAs and ¯ow cytometry experiments were performed as previously described [10,17]. In brief, the antigenicity of a peptide was determined as its ability to inhibit binding of mAb BH129 to MV and is expressed as 50%-inhibiting concentration (IC50% in mM). It was determined in competition ELISA with MV-coated microtiter plates, by preincubating the lowest saturating concentration of BH129 with increasing concentrations of test peptide [10,11]. Serum titres were determined by serial dilutions in microtiter plates coated overnight at 48C with homologous and heterologous peptides (1 mg/well in 50 ml water). Antibody titers were de®ned as the highest serum dilution that gave a signal corresponding to fourfold the background that was obtained with a naive control serum. The reactivity of immune sera (1:100) with virus was tested by ¯ow cytometry using a MV-superinfected EBV-transformed human B cell line (WMPT; gift from Dr. B. Chain, London, UK) as described [17]. 2.3. Generation of mimotopes The 6mer and 13mer sequences were selected by several rounds of panning with mAb BH129 of a ®lamentous phage library with a random 6mer or 15mer insert at the N-terminus of coat protein III as described previously [11,18]. The selected sequences were operationally de®ned as mimotopes. Binding of
286
K.C. El Kasmi et al. / Vaccine 18 (2000) 284±290
Fig. 1. Antigenicity of TBB and TTBB peptides containing the TCEs T1 or T2 and the wild-type BCEs or mimotopes. Antigenicity is plotted as IC50% values in mM of the peptides tested in competition ELISA with BH129 against immobilized MV. For peptides inducing MV-crossreactivity the corresponding mean arbitrary ¯uorescence value of the sera as measured by ¯ow cytometry is given in brackets. Values are truncated at 60 mM.() denotes peptides not used for immunization since they do not contain a TCE.
the peptides to the mAb will be referred to as antigenicity. 3. Results 3.1. TBB and TTBB peptides containing 7mer mimotopes or the 7mer wild-type BCE In a previous study, two 6mer mimotopes (EMPQLP, LYMPQL) of the wild-type epitope E245LSQLS250 were identi®ed using mAb BH129. Elongated with an additional N-terminal serine of the MV sequence (SEMPQLP (Mb7)) or a C-terminal alanine (LYMPQLA (Ma7)) these mimotopes were up to
95±135 more ecient competitors than the corresponding wild-type peptide of the same length (S244ELSQLS250, B7) [11]. Therefore, these mimotopes appeared to be attractive candidates to induce functional antibodies against MV. When the 7mer BCE H244±250 (B7; IC50%=94 mM) was colinearly synthesized as a TBB peptide with the TCE T1 (F421±435) or T2 (F256±270) the IC50% of these peptides improved 3 and 12 respectively in comparison to the 7mer BCE alone (Fig. 1). When in these TBB peptides the BCE was replaced by the 7mer mimotope Ma7 or Mb7, the IC50% further improved up to 80 (Fig. 1). In all cases TBB peptides containing T1 appeared to be more antigenic than peptides containing T2 (by a factor 4, 60 and 2 respectively;
K.C. El Kasmi et al. / Vaccine 18 (2000) 284±290
287
Fig. 2. Titration of anti-peptide sera against immobilized homologous and heterologous constructs (shown within the bar).
Fig. 1). At least two other TCEs that were tested (T3 and T4 of Table 1) did not further improve the IC50% value. A TB peptide containing T1 and Ma7 embedded in the phage ¯anking residues (PMa7) also showed no improved IC50% value (data not shown). Despite their good binding properties with respect to the mAb none of these peptides induced MV-crossreactive antibodies. Since the above TBB peptides failed to induce MVcrossreactive serum, the potential of tandem TCEs was investigated. Doubling of the T1 (the best of the above TCEs) did not increase binding to the mAb but these peptides induced higher titres of peptide-crossreactive antibodies (data not shown), albeit without MV crossreactivity (Fig. 1). The same was true for a TTBB peptide with T2 and Mb7 (IC50%=6 mM) (data not shown).
3.2. Peptides with the 8mer wild-type BCE or elongated mimotopes The absence of MV-crossreactivity may be due to missing ¯anking residues of the BCE. Therefore, the BCE was tested as an 8mer by adding the Arg243 (H243±250; B8) in TTB and TTBB peptides containing the TCEs T1 or T2. Interestingly, these peptides no longer crossreacted with BH129 but they induced signi®cant levels of MV-crossreactive sera (Fig. 1). The titre against MV was 5 higher with peptides containing T2 than T1 (data not shown). Therefore TTBB peptides with TCE T2 and based on mimotopes (RSEMPQLP=Mb8; elongated with Arg243 RLYMPQLA=Ma8; Table 1) were tested. However, these constructs were not recognized by the mAb and they did not induce MV-crossreactive serum (Fig. 1).
3.3. Mimotopes of a 15mer phage library and the 15mer wild-type sequence The improved crossreactive immunogenicity of the elongated BCE (Fig. 1) prompted us to screen a random 15mer phage library. A single clone expressing AVFVTSFLPQLSY (Mo13) was selected. This sequence shares only (Q248L) with the MV sequence and (PQ) with the 6mer mimotopes. In analogy to the 6mer mimotopes, we also designed 13mer mimotopes (Ma13 and Mb13) by replacing SFLPQLS with Ma7 or Mb7. These mimotopes were compared as free peptides or TTB derivatives with 15mer BCE H236±250 (B15; Fig. 1). Whereas some of the peptides containing the mimotopes were highly antigenic with respect to the mAb, the B15 peptides were essentially negative for BH129 (Fig. 1). However, after immunization of these peptides only the TTB peptide with the MV sequence induced crossreactive antibodies.
3.4. Crossreactivity of anti-wild-type and anti-mimotope sera with homologous and heterologous peptides Since we did not detect antibodies against MV after immunization with mimotope-based or B7-based constructs, we evaluated peptide immunogenicity by testing anti-peptide sera against homologous and heterologous peptides. Anti-mimotope sera were titrated against wild-type peptides and against mimotope-based peptides and vice versa. Fig. 2 shows that anti-T1T1B7B7 serum reacted with the homologous and with the mimotope-based peptide T2T2Mb7Mb7. On the other hand, peptide T2T2Mb7Mb7 induced antibodies against peptides based on B7 and Ma7. Similar results were obtained with TBB peptides (data not shown).
288
K.C. El Kasmi et al. / Vaccine 18 (2000) 284±290
Table 2 Antigenicity (IC50%, mM) and crossreactive immunogenicity of TB peptides based on one N-terminal TCE and one wild-type BCE[19]. The MV-reactivity of anti-peptide sera was determined by ¯ow cytometry on MV-superinfected EBV-transformed cells TCE LAC (b-lactoglobulin)[20] INF (In¯uenza virus H)[21] F288 (MV F288-304)[22] F421[14] F256[14] tt830[16] LAC[20] CST3[15] INF[21] F421[14] Idem idem idem idem idem idem
B7 B7 B15 B15 B15 B15 B15 B15 B15 B12 B15 B15 B15 B16 B18 B20
BCE
IC50% (mM)
MV reactivity
(H244±250) (H244±250) (H236±250) (H236±250) (H236±250) (H236±250) (H236±250) (H236±250) (H236±250) (H244±255) (H238±252) (H240±254) (H242±256) (H236±251) (H233±250) (H236±255)
(500) (500) (500) (500) (500) (500) (500) (500) (500) (500) (43) (25) (500) (17) (500) (500)
negative negative positive positive positive positive negative negative negative negative negative negative negative negative negative neutralizing
4. Discussion Phage display and peptide libraries have been extensively used to de®ne lead structures that could mimic conformations of complex ligands. In the case of antigens, lead structures are selected on the basis of their enhanced interaction with antibodies. We have used a sequence-speci®c peptide library [10,17] and a random phage display library [11] to generate dierent sets of lead-peptides mimicking the helical [11] H244±250 domain. The phage library generated up to 135 better binders than the sequence-speci®c peptides of the corresponding length [11]. To induce antibodies against peptide-BCEs these were combined with TCEs. This can have a profound in¯uence on the structural propensity of the BCE, since ¯anking sequences codetermine its conformation [19,23,24] and its ability to induce virus-crossreactive antibodies. In order to identify peptides that would perform well during immunization it seems convenient to select TB peptides with the desired conformation on the basis of their antigenicity with appropriate mAbs [25±28]. In this study, we started from low-antigenic sequence homologues of the wild-type BCE and highly antigenic mimotope sequences. Combining these sequences as tandem repeats of BCEs with a TCE improved the antigenicity of the wild-type BCE (B7) but not of the mimotope. The four TCEs tested showed considerable dierences in their ability to improve the binding of the BCE to the mAb. When analogous synthetic TB constructs, containing either the phage mimotope or the wild-type BCE were tested in parallel, the better binders were invariably those containing the phage-derived mimotopes. Longer peptides have a more stable secondary struc-
ture than shorter peptides; in addition they can supply important ¯anking residues [11,19,29±31]. Therefore, we screened 15mer libraries. Although Mb13 was 90 more antigenic than B15 both as free peptides or as TTB peptides, only the nonantigenic TTB peptide with the wild-type BCE induced anti-MV antibodies (Fig. 1). Although the mimotopes strongly crossreacted with the mAb they failed to induce virus-crossreactive antibodies. In contrast a large series of TB peptides with the wild-type BCE did not bind the mAb, yet most of them induced MV-crossreactive serum some of which were neutralizing (see Table 2). The observation that mAb-crossreactive peptides failed to induce virus-crossreactive antibodies can be attributed to a number of reasons. (i) The crossreactive peptide may contain only part of the epitope. This is a possibility, since the minimal wild-type epitope that induces optimal crossreactive antibodies is the 8mer with the R243. However, the addition of this amino acid or longer ¯anking sequences to the minimal mimotope (T2-T2±Ma8±Ma8) did not improve viruscrossreactivity. (ii) The structural presentation of the mimotope is dierent in the peptide than in the phage environment. If this was the case, it was not possible to correct the conformation by simply adding the phage sequences that ¯ank the mimotope-insert, to the peptide. This may indicate that solid phase peptide libraries are more appropriate for the generation of peptide mimotopes than phage libraries. (iii) Since sera raised either against mimotopes or against the wildtype peptide crossreacted equally well with all peptides it seems that the mimotopes mimic the peptide BCE but do not adequately mimic the epitope present in the MV protein. (iv) A peptide may bind to residues of the
K.C. El Kasmi et al. / Vaccine 18 (2000) 284±290
mAb that are not part of the paratope de®ned by the original antigen; in the case of overlapping paratopes, however such peptides may still be able to inhibit binding of the mAb to the original epitope. These peptides are unlikely to induce antibodies crossreacting with the original epitope. Therefore such peptides should not be considered true mimotopes of the parent antigen. Antigenicity measures both spontaneous and induced-®t binding to preformed antibodies, whereas during immunization peptides with multiple con®gurations can bind to surface Ig molecules, many of which do not necessarily recognize the virus. It is therefore dicult to understand why peptides that were ignored by the mAb were able to induce viruscrossreactive and even neutralizing antibodies. We speculate that the lipophilic adjuvant (CFA/ IFA) used for immunization critically in¯uences peptide conformation, and that the peptide folds dierently in the aqueous solution of the competition assays. Alternatively, our observation may support the concept of multiple conformations that can be adopted by a given epitope of a viral protein. Accordingly, a mAb speci®c for a particular epitope conformation would not recognize all peptide conformations that could induce crossreactive antibodies. Our report describes a striking example of disparity between crossreactive antigenicity and crossreactive immunogenicity and casts considerable doubt on the predictive value of antigenicity in immunogenicity studies. This is disturbing because it considerably complicates the selection of potential vaccine candidates on the basis of their binding to a mAb.
[4] [5]
[6]
[7] [8] [9]
[10]
[11]
[12] [13]
[14]
Acknowledgements The ®nancial support of the Centre de Recherche Public-SanteÂ, of the MinisteÁre de l'Education Nationale, the MinisteÁre de la SanteÂ, Luxembourg and the EU Leonardo da Vinci fellowship to K.C.E. are gratefully acknowledged. Parts of this work were done by K.C.E. in partial ful®llment of his M.D. thesis and by S.D. and D.T. in partial ful®llment of their Ph.D. theses.
[15]
[16] [17]
References [1] Sabin AB. My last will and testament on rapid elimination and ultimate global eradication of poliomyelitis and measles. Pediatrics 1992;90:162±9. [2] Tulchinsky TH, Ginsberg GM, Abed Y, Angeles MT, Akukwe C, Bonn J. Measles control in developing and developed countries: the case for a two-dose policy. Bull World Health Organ 1993;71:93±103. [3] Garenne M, Leroy O, Beau JP, Sene I. Child mortality after
[18] [19]
[20]
289
high-titer measles vaccines: prospective study in Senegal. Lancet 1991;338:903±7. Weiss R. Measles battle loses potent weapon. Science 1992;258:546±7. Giraudon P, Wild TF. Correlation between epitopes on hemagglutinin of measles virus and biological activities: passive protection by monoclonal antibodies is related to their hemagglutination inhibiting activity. Virology 1985;144:46±58. Malvoisin E, Wild TF. Contribution of measles virus fusion protein in protective immunity: anti-F monoclonal antibodies neutralize virus infectivity and protect mice against challenge. J Virol 1990;64:5160±2. Varsanyi TM, Utter G, Norrby E. Puri®cation, morphology and antigenic characterization of measles virus envelope components. J Gen Virol 1984;65:355±66. Steward MW, Stanley CM, Obeid OE. A mimotope from a solid±phase peptide library induces a measles virus-neutralizing and protective antibody response. J Virol 1995;69:7668±73. Chargelegue D, Obeid OE, Hsu SC, Shaw MD, Denbury AN, Taylor G, Steward MW. A peptide mimic of a protective epitope of respiratory syncytial virus selected from a combinatorial library induces virus-neutralizing antibodies and reduces viral load in vivo. J Virol 1998;72:2040±6. Fournier P, Brons NH, Berbers G, Wiesmuller KH, Fleckenstein BT, Schneider F, Jung G, Muller CP. Antibodies to a new linear site at the topographical or functional interface between the haemagglutinin and fusion proteins protect against measles encephalitis. J Gen Virol 1997;78:1295±302. Deroo S, El Kasmi KC, Fournier P, Theisen D, Brons NHC, Herrmann M, Desmet J, Muller CP. Enhanced antigenicity of a four-contact-residue epitope of the measles virus hemagglutinin protein by phage display libraries: evidence of a helical structure in the putative active site. Mol Immunol 1998;35:435. Merri®eld RB. Solid-phase peptide synthesis. Adv Enzymol Relat Areas Mol Biol 1969;32:221±96. Wiesmuller KH, Spahn G, Handtmann D, Schneider F, Jung G, Muller CP. Heterogeneity of linear B-cell epitopes of the measles virus fusion protein reacting with late convalescent sera. J Gen Virol 1992;73:2211±6. Muller CP, BuÈnder R, Mayser H, Ammon S, Weinmann M, Brons NHC, Schneider F, Jung G, WiesmuÈller KH. Intramolecular immunodominance and intermolecular selection of H2d-restricted peptides de®ne the same immunodominant region of the measles virus fusion protein. Mol Immunol 1995;32:37±47. Sinigaglia F, Guttinger M, Kilgus J, Doran DM, Matile H, Etlinger H, Trzeciak A, Gillessen D, Pink JR. A malaria T-cell epitope recognized in association with most mouse and human MHC class II molecules. Nature 1988;336:778±80. Demotz S, Lanzavecchia A, Eisel U, Niemann H, Widmann C, Corradin G. Delineation of several DR-restricted tetanus toxin T cell epitopes. J Immunol 1989;142:394±402. Ziegler D, Fournier P, Berbers GA, Steuer H, Wiesmuller KH, Fleckenstein B, Schneider F, Jung G, King CC, Muller CP. Protection against measles virus encephalitis by monoclonal antibodies binding to a cystine loop domain of the H protein mimicked by peptides which are not recognized by maternal antibodies. J Gen Virol 1996;77:2479±89. Parmley SF, Smith GP. Antibody-selectable ®lamentous fd phage vectors: Anity puri®cation of target genes. Gene 1988;73:305±18. El Kasmi KC, Theisen D, Brons NHC, Muller CP. The molecular basis of virus crossreactivity and neutralisation after immunisation with optimised chimeric peptides mimicking a putative helical epitope of the measles virus hemagglutinin protein. Mol Immunol (in press). Sakurai T, Ametani A, Nakamura Y, Shimizu M, Idota T,
290
[21]
[22] [23] [24]
[25]
K.C. El Kasmi et al. / Vaccine 18 (2000) 284±290 Kaminogawa S. Cryptic B cell determinant in a short peptide: T cells do not induce antibody response of B cells when their determinants entirely overlap each other. Int Immunol 1993;5:793±800. Hackett CJ, Hurwitz JL, Dietzschold B, Gerhard W. A synthetic decapeptide of in¯uenza virus hemagglutinin elicits helper T cells with the same ®ne recognition speci®cities as occur in response to whole virus. J Immunol 1985;135:1391±4. Partidos CD, Steward MW. Prediction and identi®cation of a T cell epitope in the fusion protein of measles virus immunodominant in mice and humans. J Gen Virol 1990;71:2099±105. Partidos CD, Steward MW. The eects of a ¯anking sequence on the immune response to a B and a T cell epitope from the fusion protein of measles virus. J Gen Virol 1992;73:1987±94. Martineau P, Leclerc C, Hofnung M. Modulating the immunological properties of a linear B-cell epitope by insertion into permissive sites of the MalE protein. Mol Immunol 1996;33:1345± 58. Denton G, Hudecz F, Kajtar J, Murray A, Tendler SJ. Price MR Sequential order of T and B cell epitopes aects immunogenicity but not antibody recognition of the B cell epitope. Pept Res 1994;7:258±64.
[26] Godard I, Estaquier J, Zenner L, Bossus M, Auriault C, Darcy F, Gras-Masse H, Capron A. Antigenicity and immunogenicity of P30-derived peptides in experimental models of toxoplasmosis. Mol Immunol 1994;31:1353±63. [27] Lee MK, Kim KL, Hahm KS, Yang KH. Structure-antigenicity relationship of peptides from the pre-S2 region of the hepatitis B virus surface antigen. Biochem Mol Biol Int 1994;34:159±68. [28] Marguerite M, Bossus M, Mazingue C, Wolowczuk I, GrasMasse H, Tartar A, Capron A, Auriault C. Analysis of antigenicity and immunogenicity of ®ve dierent chemically de®ned constructs of a peptide. Mol Immunol 1992;29:793±800. [29] Vijayakrishnan L, Sarkar S, Roy RP, Rao KV. B cell responses to a peptide epitope: IV. Subtle sequence changes in ¯anking residues modulate immunogenicity. J Immunol 1997;159:1809± 19. [30] Su JY, Hodges RS, Kay CM. Eect of chain length on the formation and stability of synthetic alpha-helical coiled coils. Biochemistry 1994;33:15501±10. [31] Gras-Masse H, Jolivet M, Drobecq H, Aubert JP, Beachey EH, Audibert F, Chedid L, Tartar A. In¯uence of helical organization on immunogenicity and antigenicity of synthetic peptides. Mol Immunol 1988;25:673±8.
www.elsevier.com/locate/vaccine
Crossreactivity of mimotopes and peptide homologues of a sequential epitope with a monoclonal antibody does not predict crossreactive immunogenicity Karim C. El Kasmi a, b, Sabrina Deroo a, Dietmar M. Theisen a, Nicolaas H.C. Brons a, Claude P. Muller a, b,* a
Laboratoire National de SanteÂ, P. O. Box 1102, L-1011 Luxembourg, Luxembourg b Medizinische FakultaÈt, UniversitaÈt TuÈbingen, D-72076 TuÈbingen, FRG
Abstract The sequence H236±256 of the measles virus (MV) hemagglutinin (H) contains the sequential epitope of the neutralizing and protective monoclonal antibody (mAb) BH129 with the minimal epitope E245L±QL249. Using this mAb, we have recently developed 7mer mimotopes binding up to 135 better than the corresponding 7mer epitope H244±250. In this study, we combined T cell epitopes (TCE) with either highly crossreactive 7mer mimotopes, 13mer mimotopes or less crossreactive MVderived B cell epitopes (BCE). Antigenicity of these TBB, TTB and TTBB peptides was determined with BH129 in a competition ELISA against MV. We found that chimeric peptides including mimotopes were up to 80 better binders to the mAb than peptides containing the original BCEs. All peptides irrespective of their antigenicity were used for immunization to compare their virus- crossreactive immunogenicity. Unexpectedly, none of the highly antigenic mimotope-based peptides induced MVcrossreactive antibodies. In contrast, a number of peptides with the viral BCE sequence that did not bind to the mAb, induced MV-crossreactive and even neutralizing antibodies. This report describes a striking example of disparity between antigenicity and crossreactive immunogenicity and casts considerable doubt on the predictive value of antigenicity in immunogenicity studies, considerably complicating the selection of potential vaccine candidates. # 1999 Elsevier Science Ltd. All rights reserved. Keywords: Measles virus; Mimotope; Synthetic peptides; Antigenicity
1. Introduction Current live attenuated vaccines have successfully reduced mortality due to measles virus (MV) infection in the developed world. In developing countries the disease still causes the death of one million children every year [1,2]. Major drawbacks of current vaccines include the lack of resistance to neutralization by maternal antibodies and their thermal sensitivity. Subunit-vaccines based on synthetic peptides that induce neutralizing and protective antibodies could po-
* Corresponding author. Tel.: +352-4-90604; fax: +352-4-90686. E-mail address: [email protected] (C.P. Muller)
tentially circumvent these problems [3,4]. Passively transferred antibodies directed against the hemagglutinin protein (H) fully protect against MV infection, emphasizing the role of antibodies even in the absence of a T cell response [5±7]. Since conformational epitopes cannot easily be mimicked by synthetic peptides the design of peptide vaccines is mostly con®ned to sequential B cell epitopes (BCE). However, random phage display libraries or solid phase combinatorial peptide libraries have been used to ®nd peptides that mimic more complex epitopes. Steward et al. reported mimotopes that generated neutralizing and/or protective antibodies against MV and respiratory syncytial virus [8,9]. Although, in the case of the MV mimotope this result is surprising,
0264-410X/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 4 - 4 1 0 X ( 9 9 ) 0 0 1 9 9 - 1
K.C. El Kasmi et al. / Vaccine 18 (2000) 284±290
285
Table 1 T and B cell epitopes used Abbreviation
Sequence
Description
T1 T2 T3 T4 B7 B8 B15 Ma7 Mb7 PMa7 Ma8 Mb8 Mo13 Ma13 Mb13
PVVEVNGVTIQVGSR LLGILESRGIKARIT DIEKKIAKMEKASSVFNVVNS QYIKANSKFIGITEL SELSQLS RSELSQLS KPNLSSKRSELSQLS LYMPQLA/S SEMPQLP PFYSHSADGALYMPQLGAAGA RLYMPQLA/S RSEMPQLP AVFVTSFLPQLSY AVFVTLYMPQLSY AVFVTSEMPQLPY
TCE (MV F421-435)[14] TCE (MV F256-270)[14] TCE (CST3 378-398; P. falcip.)[15] TCE (tt830-844; Tetanus toxoid)[16] MV sequence H244-250[10,11] MV sequence H243-250[10] MV sequence H236-250[10] mimotope; corresponding to H244-250[11] idem mimotope between phage ¯anking residues mimotope; corresponding to H243±250 idem mimotope idem idem
since the template monoclonal antibody F2.41 does not neutralize or protect, these observations emphasize the potential of mimotopes as candidate vaccines. A sequential epitope of MV±H has recently been mapped to the region H236±256 with a series of neutralizing and protective monoclonal antibodies (mAbs) including BH129 [10]. In a previous study, we have generated mimotopes against BH129 using a 6mer phage displayed random library. Most of the selected clones had little sequence homology with the original epitope but the corresponding peptides bound BH129 up to 135 better than the peptides homologous with the sequence of this sequential BCE, suggesting that the mimotopes mimicked the conformation of the viral epitope more eciently than the MV-derived peptides [11]. In the present study, we compared the ability of the mimotopes and of MV-derived peptides to induce MVcrossreactive antibodies when combined with T cell epitopes (TCE) in iteratively optimized synthetic TB peptides. Although, in these TB peptides the mimotopes were considerably better binders to the mAb than were the MV sequences, only the latter generated MV-crossreactive antibodies.
2. Materials and methods 2.1. Peptide synthesis and immunizations All peptides were synthesized by automated solid phase Fmoc chemistry on a SYRO peptide synthesizer (Multisyntech, Bochum) [12,13]. Purity was >60%, as determined by reversed phase high-performance liquid chromatography (HPLC). Peptides used in this study are shown in Table 1. Peptides containing the sequence of the viral epitope will be referred to as wild-type peptides.
Male and female speci®c pathogen-free BALB/c mice (H2d) (8±15 weeks old) were immunized twice (day 0, 14) intraperitoneally with a CFA (priming)/ IFA (boost) (Sigma, USA) emulsion of 150 mg peptide (mixed 1:1). Blood was collected seven to ten days after the boost to prepare serum. 2.2. ELISA and ¯ow cytometry All ELISAs and ¯ow cytometry experiments were performed as previously described [10,17]. In brief, the antigenicity of a peptide was determined as its ability to inhibit binding of mAb BH129 to MV and is expressed as 50%-inhibiting concentration (IC50% in mM). It was determined in competition ELISA with MV-coated microtiter plates, by preincubating the lowest saturating concentration of BH129 with increasing concentrations of test peptide [10,11]. Serum titres were determined by serial dilutions in microtiter plates coated overnight at 48C with homologous and heterologous peptides (1 mg/well in 50 ml water). Antibody titers were de®ned as the highest serum dilution that gave a signal corresponding to fourfold the background that was obtained with a naive control serum. The reactivity of immune sera (1:100) with virus was tested by ¯ow cytometry using a MV-superinfected EBV-transformed human B cell line (WMPT; gift from Dr. B. Chain, London, UK) as described [17]. 2.3. Generation of mimotopes The 6mer and 13mer sequences were selected by several rounds of panning with mAb BH129 of a ®lamentous phage library with a random 6mer or 15mer insert at the N-terminus of coat protein III as described previously [11,18]. The selected sequences were operationally de®ned as mimotopes. Binding of
286
K.C. El Kasmi et al. / Vaccine 18 (2000) 284±290
Fig. 1. Antigenicity of TBB and TTBB peptides containing the TCEs T1 or T2 and the wild-type BCEs or mimotopes. Antigenicity is plotted as IC50% values in mM of the peptides tested in competition ELISA with BH129 against immobilized MV. For peptides inducing MV-crossreactivity the corresponding mean arbitrary ¯uorescence value of the sera as measured by ¯ow cytometry is given in brackets. Values are truncated at 60 mM.() denotes peptides not used for immunization since they do not contain a TCE.
the peptides to the mAb will be referred to as antigenicity. 3. Results 3.1. TBB and TTBB peptides containing 7mer mimotopes or the 7mer wild-type BCE In a previous study, two 6mer mimotopes (EMPQLP, LYMPQL) of the wild-type epitope E245LSQLS250 were identi®ed using mAb BH129. Elongated with an additional N-terminal serine of the MV sequence (SEMPQLP (Mb7)) or a C-terminal alanine (LYMPQLA (Ma7)) these mimotopes were up to
95±135 more ecient competitors than the corresponding wild-type peptide of the same length (S244ELSQLS250, B7) [11]. Therefore, these mimotopes appeared to be attractive candidates to induce functional antibodies against MV. When the 7mer BCE H244±250 (B7; IC50%=94 mM) was colinearly synthesized as a TBB peptide with the TCE T1 (F421±435) or T2 (F256±270) the IC50% of these peptides improved 3 and 12 respectively in comparison to the 7mer BCE alone (Fig. 1). When in these TBB peptides the BCE was replaced by the 7mer mimotope Ma7 or Mb7, the IC50% further improved up to 80 (Fig. 1). In all cases TBB peptides containing T1 appeared to be more antigenic than peptides containing T2 (by a factor 4, 60 and 2 respectively;
K.C. El Kasmi et al. / Vaccine 18 (2000) 284±290
287
Fig. 2. Titration of anti-peptide sera against immobilized homologous and heterologous constructs (shown within the bar).
Fig. 1). At least two other TCEs that were tested (T3 and T4 of Table 1) did not further improve the IC50% value. A TB peptide containing T1 and Ma7 embedded in the phage ¯anking residues (PMa7) also showed no improved IC50% value (data not shown). Despite their good binding properties with respect to the mAb none of these peptides induced MV-crossreactive antibodies. Since the above TBB peptides failed to induce MVcrossreactive serum, the potential of tandem TCEs was investigated. Doubling of the T1 (the best of the above TCEs) did not increase binding to the mAb but these peptides induced higher titres of peptide-crossreactive antibodies (data not shown), albeit without MV crossreactivity (Fig. 1). The same was true for a TTBB peptide with T2 and Mb7 (IC50%=6 mM) (data not shown).
3.2. Peptides with the 8mer wild-type BCE or elongated mimotopes The absence of MV-crossreactivity may be due to missing ¯anking residues of the BCE. Therefore, the BCE was tested as an 8mer by adding the Arg243 (H243±250; B8) in TTB and TTBB peptides containing the TCEs T1 or T2. Interestingly, these peptides no longer crossreacted with BH129 but they induced signi®cant levels of MV-crossreactive sera (Fig. 1). The titre against MV was 5 higher with peptides containing T2 than T1 (data not shown). Therefore TTBB peptides with TCE T2 and based on mimotopes (RSEMPQLP=Mb8; elongated with Arg243 RLYMPQLA=Ma8; Table 1) were tested. However, these constructs were not recognized by the mAb and they did not induce MV-crossreactive serum (Fig. 1).
3.3. Mimotopes of a 15mer phage library and the 15mer wild-type sequence The improved crossreactive immunogenicity of the elongated BCE (Fig. 1) prompted us to screen a random 15mer phage library. A single clone expressing AVFVTSFLPQLSY (Mo13) was selected. This sequence shares only (Q248L) with the MV sequence and (PQ) with the 6mer mimotopes. In analogy to the 6mer mimotopes, we also designed 13mer mimotopes (Ma13 and Mb13) by replacing SFLPQLS with Ma7 or Mb7. These mimotopes were compared as free peptides or TTB derivatives with 15mer BCE H236±250 (B15; Fig. 1). Whereas some of the peptides containing the mimotopes were highly antigenic with respect to the mAb, the B15 peptides were essentially negative for BH129 (Fig. 1). However, after immunization of these peptides only the TTB peptide with the MV sequence induced crossreactive antibodies.
3.4. Crossreactivity of anti-wild-type and anti-mimotope sera with homologous and heterologous peptides Since we did not detect antibodies against MV after immunization with mimotope-based or B7-based constructs, we evaluated peptide immunogenicity by testing anti-peptide sera against homologous and heterologous peptides. Anti-mimotope sera were titrated against wild-type peptides and against mimotope-based peptides and vice versa. Fig. 2 shows that anti-T1T1B7B7 serum reacted with the homologous and with the mimotope-based peptide T2T2Mb7Mb7. On the other hand, peptide T2T2Mb7Mb7 induced antibodies against peptides based on B7 and Ma7. Similar results were obtained with TBB peptides (data not shown).
288
K.C. El Kasmi et al. / Vaccine 18 (2000) 284±290
Table 2 Antigenicity (IC50%, mM) and crossreactive immunogenicity of TB peptides based on one N-terminal TCE and one wild-type BCE[19]. The MV-reactivity of anti-peptide sera was determined by ¯ow cytometry on MV-superinfected EBV-transformed cells TCE LAC (b-lactoglobulin)[20] INF (In¯uenza virus H)[21] F288 (MV F288-304)[22] F421[14] F256[14] tt830[16] LAC[20] CST3[15] INF[21] F421[14] Idem idem idem idem idem idem
B7 B7 B15 B15 B15 B15 B15 B15 B15 B12 B15 B15 B15 B16 B18 B20
BCE
IC50% (mM)
MV reactivity
(H244±250) (H244±250) (H236±250) (H236±250) (H236±250) (H236±250) (H236±250) (H236±250) (H236±250) (H244±255) (H238±252) (H240±254) (H242±256) (H236±251) (H233±250) (H236±255)
(500) (500) (500) (500) (500) (500) (500) (500) (500) (500) (43) (25) (500) (17) (500) (500)
negative negative positive positive positive positive negative negative negative negative negative negative negative negative negative neutralizing
4. Discussion Phage display and peptide libraries have been extensively used to de®ne lead structures that could mimic conformations of complex ligands. In the case of antigens, lead structures are selected on the basis of their enhanced interaction with antibodies. We have used a sequence-speci®c peptide library [10,17] and a random phage display library [11] to generate dierent sets of lead-peptides mimicking the helical [11] H244±250 domain. The phage library generated up to 135 better binders than the sequence-speci®c peptides of the corresponding length [11]. To induce antibodies against peptide-BCEs these were combined with TCEs. This can have a profound in¯uence on the structural propensity of the BCE, since ¯anking sequences codetermine its conformation [19,23,24] and its ability to induce virus-crossreactive antibodies. In order to identify peptides that would perform well during immunization it seems convenient to select TB peptides with the desired conformation on the basis of their antigenicity with appropriate mAbs [25±28]. In this study, we started from low-antigenic sequence homologues of the wild-type BCE and highly antigenic mimotope sequences. Combining these sequences as tandem repeats of BCEs with a TCE improved the antigenicity of the wild-type BCE (B7) but not of the mimotope. The four TCEs tested showed considerable dierences in their ability to improve the binding of the BCE to the mAb. When analogous synthetic TB constructs, containing either the phage mimotope or the wild-type BCE were tested in parallel, the better binders were invariably those containing the phage-derived mimotopes. Longer peptides have a more stable secondary struc-
ture than shorter peptides; in addition they can supply important ¯anking residues [11,19,29±31]. Therefore, we screened 15mer libraries. Although Mb13 was 90 more antigenic than B15 both as free peptides or as TTB peptides, only the nonantigenic TTB peptide with the wild-type BCE induced anti-MV antibodies (Fig. 1). Although the mimotopes strongly crossreacted with the mAb they failed to induce virus-crossreactive antibodies. In contrast a large series of TB peptides with the wild-type BCE did not bind the mAb, yet most of them induced MV-crossreactive serum some of which were neutralizing (see Table 2). The observation that mAb-crossreactive peptides failed to induce virus-crossreactive antibodies can be attributed to a number of reasons. (i) The crossreactive peptide may contain only part of the epitope. This is a possibility, since the minimal wild-type epitope that induces optimal crossreactive antibodies is the 8mer with the R243. However, the addition of this amino acid or longer ¯anking sequences to the minimal mimotope (T2-T2±Ma8±Ma8) did not improve viruscrossreactivity. (ii) The structural presentation of the mimotope is dierent in the peptide than in the phage environment. If this was the case, it was not possible to correct the conformation by simply adding the phage sequences that ¯ank the mimotope-insert, to the peptide. This may indicate that solid phase peptide libraries are more appropriate for the generation of peptide mimotopes than phage libraries. (iii) Since sera raised either against mimotopes or against the wildtype peptide crossreacted equally well with all peptides it seems that the mimotopes mimic the peptide BCE but do not adequately mimic the epitope present in the MV protein. (iv) A peptide may bind to residues of the
K.C. El Kasmi et al. / Vaccine 18 (2000) 284±290
mAb that are not part of the paratope de®ned by the original antigen; in the case of overlapping paratopes, however such peptides may still be able to inhibit binding of the mAb to the original epitope. These peptides are unlikely to induce antibodies crossreacting with the original epitope. Therefore such peptides should not be considered true mimotopes of the parent antigen. Antigenicity measures both spontaneous and induced-®t binding to preformed antibodies, whereas during immunization peptides with multiple con®gurations can bind to surface Ig molecules, many of which do not necessarily recognize the virus. It is therefore dicult to understand why peptides that were ignored by the mAb were able to induce viruscrossreactive and even neutralizing antibodies. We speculate that the lipophilic adjuvant (CFA/ IFA) used for immunization critically in¯uences peptide conformation, and that the peptide folds dierently in the aqueous solution of the competition assays. Alternatively, our observation may support the concept of multiple conformations that can be adopted by a given epitope of a viral protein. Accordingly, a mAb speci®c for a particular epitope conformation would not recognize all peptide conformations that could induce crossreactive antibodies. Our report describes a striking example of disparity between crossreactive antigenicity and crossreactive immunogenicity and casts considerable doubt on the predictive value of antigenicity in immunogenicity studies. This is disturbing because it considerably complicates the selection of potential vaccine candidates on the basis of their binding to a mAb.
[4] [5]
[6]
[7] [8] [9]
[10]
[11]
[12] [13]
[14]
Acknowledgements The ®nancial support of the Centre de Recherche Public-SanteÂ, of the MinisteÁre de l'Education Nationale, the MinisteÁre de la SanteÂ, Luxembourg and the EU Leonardo da Vinci fellowship to K.C.E. are gratefully acknowledged. Parts of this work were done by K.C.E. in partial ful®llment of his M.D. thesis and by S.D. and D.T. in partial ful®llment of their Ph.D. theses.
[15]
[16] [17]
References [1] Sabin AB. My last will and testament on rapid elimination and ultimate global eradication of poliomyelitis and measles. Pediatrics 1992;90:162±9. [2] Tulchinsky TH, Ginsberg GM, Abed Y, Angeles MT, Akukwe C, Bonn J. Measles control in developing and developed countries: the case for a two-dose policy. Bull World Health Organ 1993;71:93±103. [3] Garenne M, Leroy O, Beau JP, Sene I. Child mortality after
[18] [19]
[20]
289
high-titer measles vaccines: prospective study in Senegal. Lancet 1991;338:903±7. Weiss R. Measles battle loses potent weapon. Science 1992;258:546±7. Giraudon P, Wild TF. Correlation between epitopes on hemagglutinin of measles virus and biological activities: passive protection by monoclonal antibodies is related to their hemagglutination inhibiting activity. Virology 1985;144:46±58. Malvoisin E, Wild TF. Contribution of measles virus fusion protein in protective immunity: anti-F monoclonal antibodies neutralize virus infectivity and protect mice against challenge. J Virol 1990;64:5160±2. Varsanyi TM, Utter G, Norrby E. Puri®cation, morphology and antigenic characterization of measles virus envelope components. J Gen Virol 1984;65:355±66. Steward MW, Stanley CM, Obeid OE. A mimotope from a solid±phase peptide library induces a measles virus-neutralizing and protective antibody response. J Virol 1995;69:7668±73. Chargelegue D, Obeid OE, Hsu SC, Shaw MD, Denbury AN, Taylor G, Steward MW. A peptide mimic of a protective epitope of respiratory syncytial virus selected from a combinatorial library induces virus-neutralizing antibodies and reduces viral load in vivo. J Virol 1998;72:2040±6. Fournier P, Brons NH, Berbers G, Wiesmuller KH, Fleckenstein BT, Schneider F, Jung G, Muller CP. Antibodies to a new linear site at the topographical or functional interface between the haemagglutinin and fusion proteins protect against measles encephalitis. J Gen Virol 1997;78:1295±302. Deroo S, El Kasmi KC, Fournier P, Theisen D, Brons NHC, Herrmann M, Desmet J, Muller CP. Enhanced antigenicity of a four-contact-residue epitope of the measles virus hemagglutinin protein by phage display libraries: evidence of a helical structure in the putative active site. Mol Immunol 1998;35:435. Merri®eld RB. Solid-phase peptide synthesis. Adv Enzymol Relat Areas Mol Biol 1969;32:221±96. Wiesmuller KH, Spahn G, Handtmann D, Schneider F, Jung G, Muller CP. Heterogeneity of linear B-cell epitopes of the measles virus fusion protein reacting with late convalescent sera. J Gen Virol 1992;73:2211±6. Muller CP, BuÈnder R, Mayser H, Ammon S, Weinmann M, Brons NHC, Schneider F, Jung G, WiesmuÈller KH. Intramolecular immunodominance and intermolecular selection of H2d-restricted peptides de®ne the same immunodominant region of the measles virus fusion protein. Mol Immunol 1995;32:37±47. Sinigaglia F, Guttinger M, Kilgus J, Doran DM, Matile H, Etlinger H, Trzeciak A, Gillessen D, Pink JR. A malaria T-cell epitope recognized in association with most mouse and human MHC class II molecules. Nature 1988;336:778±80. Demotz S, Lanzavecchia A, Eisel U, Niemann H, Widmann C, Corradin G. Delineation of several DR-restricted tetanus toxin T cell epitopes. J Immunol 1989;142:394±402. Ziegler D, Fournier P, Berbers GA, Steuer H, Wiesmuller KH, Fleckenstein B, Schneider F, Jung G, King CC, Muller CP. Protection against measles virus encephalitis by monoclonal antibodies binding to a cystine loop domain of the H protein mimicked by peptides which are not recognized by maternal antibodies. J Gen Virol 1996;77:2479±89. Parmley SF, Smith GP. Antibody-selectable ®lamentous fd phage vectors: Anity puri®cation of target genes. Gene 1988;73:305±18. El Kasmi KC, Theisen D, Brons NHC, Muller CP. The molecular basis of virus crossreactivity and neutralisation after immunisation with optimised chimeric peptides mimicking a putative helical epitope of the measles virus hemagglutinin protein. Mol Immunol (in press). Sakurai T, Ametani A, Nakamura Y, Shimizu M, Idota T,
290
[21]
[22] [23] [24]
[25]
K.C. El Kasmi et al. / Vaccine 18 (2000) 284±290 Kaminogawa S. Cryptic B cell determinant in a short peptide: T cells do not induce antibody response of B cells when their determinants entirely overlap each other. Int Immunol 1993;5:793±800. Hackett CJ, Hurwitz JL, Dietzschold B, Gerhard W. A synthetic decapeptide of in¯uenza virus hemagglutinin elicits helper T cells with the same ®ne recognition speci®cities as occur in response to whole virus. J Immunol 1985;135:1391±4. Partidos CD, Steward MW. Prediction and identi®cation of a T cell epitope in the fusion protein of measles virus immunodominant in mice and humans. J Gen Virol 1990;71:2099±105. Partidos CD, Steward MW. The eects of a ¯anking sequence on the immune response to a B and a T cell epitope from the fusion protein of measles virus. J Gen Virol 1992;73:1987±94. Martineau P, Leclerc C, Hofnung M. Modulating the immunological properties of a linear B-cell epitope by insertion into permissive sites of the MalE protein. Mol Immunol 1996;33:1345± 58. Denton G, Hudecz F, Kajtar J, Murray A, Tendler SJ. Price MR Sequential order of T and B cell epitopes aects immunogenicity but not antibody recognition of the B cell epitope. Pept Res 1994;7:258±64.
[26] Godard I, Estaquier J, Zenner L, Bossus M, Auriault C, Darcy F, Gras-Masse H, Capron A. Antigenicity and immunogenicity of P30-derived peptides in experimental models of toxoplasmosis. Mol Immunol 1994;31:1353±63. [27] Lee MK, Kim KL, Hahm KS, Yang KH. Structure-antigenicity relationship of peptides from the pre-S2 region of the hepatitis B virus surface antigen. Biochem Mol Biol Int 1994;34:159±68. [28] Marguerite M, Bossus M, Mazingue C, Wolowczuk I, GrasMasse H, Tartar A, Capron A, Auriault C. Analysis of antigenicity and immunogenicity of ®ve dierent chemically de®ned constructs of a peptide. Mol Immunol 1992;29:793±800. [29] Vijayakrishnan L, Sarkar S, Roy RP, Rao KV. B cell responses to a peptide epitope: IV. Subtle sequence changes in ¯anking residues modulate immunogenicity. J Immunol 1997;159:1809± 19. [30] Su JY, Hodges RS, Kay CM. Eect of chain length on the formation and stability of synthetic alpha-helical coiled coils. Biochemistry 1994;33:15501±10. [31] Gras-Masse H, Jolivet M, Drobecq H, Aubert JP, Beachey EH, Audibert F, Chedid L, Tartar A. In¯uence of helical organization on immunogenicity and antigenicity of synthetic peptides. Mol Immunol 1988;25:673±8.