Grafting of a hepatitis B S-preS(2) T-cell epitope on lysozyme enhances the immunogenicity of lysozyme in responder mice primed with the T-cell epitope

Immunology Letters, 41 (1994) 25-32 0165 - 2478 / 94 / $ 7.00 © 1994 Elsevier Science B.V. All rights reserved IMLET 02115

Grafting of a hepatitis B S-preS(2) T-cell epitope on lysozyme enhances the immunogenicity of lysozyme in responder mice primed with the T-cell epitope Jean-Pierre Y. Scheerlinck a'*, Alain Michel b and Patrick De Baetselier a aUnit of Cellular Immunology, Institute of Molecular Biology, Vrije UniversiteitBrussel, St-Genesius-Rode, Belgium; and bLaboratoire de Chimie Biologique, Universitd de Mons-Hainaut, Mons, Belgium (Received 26 December 1993; accepted 21 February 1994)

1.

Summary

Subunit immunogens composed of well-defined Tand B-cell epitopes might represent a valuable approach to design vaccines. The reduction of the size of the T-cell epitope is clearly in the line of this strategy. In this study we evaluated the capacity of a hepatitis B S-preS(2) surface antigen-derived T-cell epitope (i.e., S2b) to enhance the humoral immune response towards lysozyme when covalently linked to this antigen. We hereby anticipated that new problems, related to processing of a subunit immunogen, may emerge when grafting minimalized T-cell epitopes on protein antigens. Indeed, insertion of a T-cell epitope containing peptide (i.e., S2b) in a new protein context does not warrant a correct processing of the T-cell epitope. To avoid such potential processing problems an acid labile linker between T-cell and B-cell epitopes was devised in order to provide a processing-independent cleavage site. Using a T-cell hybridoma specific for the S2b T-cell epitope the S2bC-lysozyme conjugate was found to be presented by functional antigen-presenting cells. However, fixed APC did not present the conjugate in vitro indicating that processing is required for the release and Key words: T-cell epitope; Vaccine *Corresponding author: Jean-Pierre Y. Scheerlinck, Department of Cellular Immunology, Institute of Molecular Biology, Vrije Universiteit Brussel, Paardenstraat 65, B-1640 Sint-Genesius-Rode, Belgium. Abbreviations: APC, antigen-presenting cell; Ag, antigen; FCS, fetal calf serum; IL-2, interleukin-2, 2-ME, 2-mercaptoethanol; PBS, phosphate-buffered saline. SSDI 0 1 6 5 - 2 4 7 8 ( 9 4 ) 0 0 0 3 1 - L

presentation of S2b. The ability of the conjugate to generate an enhanced immune response was investigated in vivo. In S2b-primed mice the S2bC-lysozyme conjugate was found to elicit a faster and higher anti-lysozyme humoral response, as compared to uncoupled mixtures of lysozyme and S2b. Moreover, more IgG was produced when the mice were immunised with the conjugate as compared to the mixture of lysozyme and S2b, indicating the induction of a 'secondary-type' of immune response against lysozyme. In contrast, in non-responder mice for the S2b T-cell epitope, the same immunization protocol did not induce anti-lysozyme antibodies. These results demonstrate that S2b-induced memory T cells provide help to lysozyme-specific B cells during primary immunizations with the S2bC-lysozyme conjugate, resulting in a secondary IgG-type anti-lysozyme humoral response.

2. Introduction T cells play a central role in the regulation of the immune response, and the use of carriers to enhance the humoral immune response to weak immunogens has been well documented [1,2]. In order to obtain a secondary immune response, it is essential that antigen-specific memory T cells provide help to the antibody-producing B cells. However, on the basis of current models for T-B cell collaboration [3], it is not essential that the antigens used for generation of the memory T cells and for induction of the secondary immune response should be molecularly identical. Indeed, an effective immune response against any type of antigen will occur, provided that the immu-

nising antigen shares a T-cell epitope with the priming antigen [4]. Consequently, it was proposed that one can take advantage of preceding vaccinations to improve the production of antibodies against a new antigen, provided that both antigens shared the identical T-cell epitopes [1]. However, the general application of this simple concept is hampered by several phenomena [5] such as epitope-specific regulatory mechanisms [6,7]. Several types of epitope-specific suppressor mechanisms have been hypothesized, including clonal dominance of carrier-specific B cells [8], epitope-specific T suppressor cells [9] and anergy of unprimed hapten-specific B cells [10]. In order to avoid these annoying side effects, mostly encountered when using large carriers, the possibility of using minimal carriers, i.e., small peptides encompassing T-cell epitopes, was considered. First, it could be expected that, because of their limited size, minimal carriers might avoid antigen sequestration by carrier-specific B cells. Second, if as suggested previously [I 1,12], the T-helper and T-suppressor repertoires are distinct and properly selected, small carrier peptides might contain only T-helper epitopes, thus avoiding this type of epitope-specific suppression. Third, minimal-carrier epitopes may represent human 'promiscuous T-cell epitopes', which can be presented in the context of different MHC haplotypes [13], hereby allowing their broader use in outbred populations. However, there is one important drawback when using minimal-carrier peptides conjugated to antigens, namely unpredictable processing. Indeed, since the T-cell epitope is separated from its original protein environment and grafted to a different antigen, the processing of this new complex might be altered, resulting in an ineffective or altered release of the T-cell epitope. The insertion of an acid-labile link between the carrier peptide and hapten may circumvent this problem. Such type of chemical link is pH sensitive and hence the low pH of the endosomal compartment would cleave the Tcell epitope from the antigen, resulting in a processing-independent release of the T-cell epitope. Consequently, this type of linking may allow conjugation of a large variety of T- and B-cell epitopes. Taking advantage of T-cell hybridomas specific for the used T-cell epitope (i.e., the S2b T-cell epitope derived from the hepatitis B S-preS(2) surface antigen sequence (156-170) [14]), we analyzed the antigen presentation requirements of S2b conjugated to lysozyme through an acid-labile covalent bond. Subsequently, the immunopotentiating effect of the grafted T-cell epitope on the humoral anti-lysozyme response was investigated in vivo. The T-cell priming 26

effect was compared in high- and low-responder mouse strains for the T-cell epitope.

3. Materials and Methods

3.1. Mice and antigens Female BALB/c mice (6-10 weeks old) were purchased from the Animal Breading Centre of the SCK, Mol, Belgium. Female C57B1/6 mice (6-10 weeks old) were obtained from Banten and Kingman, North Humberside, UK. The S2b peptide (156-170: NIASHISSSSARTGD) and the S2bC peptide (with a cystein residue added at the C-terminal end of S2b) were synthesised by the Merrifield [15] solid-phase method and were subjected to HPLC on a C18 reverse-phase column.

3.2. Preparation of the T-cell epitope-lysozyme conjugates Lysozyme (10 mg/ml, 2 ml) was derivatized with 1 thiol group/protein using Traut's reagents (2-iminothiolane, Pierce Chemical; Rockford, USA) at pH 8 in 0.05 M phosphate buffer for 2 h. Precipitated product was removed by centrifugation and thiolated lysozyme was freed from excess 2-iminothiolane by gel filtration on a sephadex G-25 F column. The recovered product was reacted with a 15-fold molar excess of bis-maleimide reagent over the free -SH concentration for 30 min at pH 8.0. Bis-(maleimido-ethoxy)-propane, prepared as described by Srinivasachar and Neville [16], was used for the synthesis of the acid-labile conjugate. Maleimido-derivatized lysozyme was separated from excess reagent by chromatography on a sephadex G25 F column. Finally, the -SH group of the S2bC peptide was reacted with the maleimido-derivatized lysozyme (molar ratio lysozyme/peptides, 1:4) at pH 8.0 for 1 h and subsequently passaged on a Sephadex G25 F column.

3.3. Antigen presentation assay The B-cell hybridoma TA3 [17] was maintained in complete medium (i.e., RPMI-1640 (Gibco, Grand Island, NY) supplemented with 0.30 /~g/ml h-glutamine (Flow, McLean, VA), 100 U/ml penicillin, 0.1 /~g/ml streptomycin (Gibco), 10% heat-inactivated FCS (Gibco). TA3 or syngeneic spleen cells were also used as a source of APC. 'Fixed APC' refer to APC treated with 0.5% formaldehyde for 30 min at room temperature, and subsequently washed exten-

sively with PBS-10% FCS as described previously [18]. APC, at the indicated concentration, were mixed with 2 × 104 T-cell hybridomas in micro-well plates in a final volume of 200 #1. Antigen (50 #g/ml) was added to the appropriate wells. When conjugates were used as antigen, free peptides were removed from the conjugates by extensive dialysis against PBS using 6000-8000 molecular weight cut-off membranes. After 24-h incubation, 100 #1 of the culture supernate was transferred to a new micro-well plate and frozen to lyse remaining cells. 106 CTL-L cells [19] (a kind gift of Dr. C. Uytenhoven from the Universit6 Catbolique de Louvain, ICP, Brussels, Belgium) were added to the freeze-thawed supernate and after 24 h, the micro cultures were pulsed with [methyl-3H]thymidine (Amersham, Brucks, UK) during 18 h. The cells were then harvested and the incorporated [methyl-3H]thymidine was measured by scintillation counting. The results are presented as cpm upon subtraction of background proliferation in the absence of Ag. 3.4. Identification of low-responder mouse strains to

S2b Mice were primed with 100 #g of S2b peptide emulsified in complete Freund's adjuvant (CFA: Difco, Detroit, MI), inoculated intra-footpad. One week later, the popliteal lymph node cells were recovered and 2 x 104 cells/200 /~1 were re-stimulated in vitro with 10 #g/ml S2b peptide. At different time intervals the cells were pulsed with [methyl-3H]thymidine during 18 h. The incorporated [methyl-3H]thymidine was measured and the background proliferation without antigen was subtracted.

3.5. In vivo antibody production assay Mice were injected intraperitoneally with either PBS or 100 #g of S2b peptide emulsified in CFA. Three weeks later the mice were boosted with either 100 pg of lysozyme mixed with 7/~g of S2b peptide or 100 #g of S2bC-lysozyme conjugate. Blood was harvested at regular time intervals, diluted 100 times in PBS-I% BSA and centrifuged to remove cells and frozen. This concentration was found optimal to discriminate between high- and low-responder mice using the described immunisation protocol. After 3 weeks, the different blood samples were assayed for the presence of antibodies specific for lysozyme in ELISA. For the ELISA detection, 50 #g/ml lyso-

zyme in PBS was coated to maxisorb ELISA plates (Nunc, Roskilde, Denmark) and the overcoating was done using 1 mg/ml BSA. Test sample (100 pl) was then added to the micro wells. The presence of specific antibodies was detected by adding alkaline phosphatase-conjugated anti-mouse antibodies. Subsequently, 2 mg/ml 4-nitrophenyl phosphate disodium hexahydrate was used to detect the presence of alkaline phosphatase. The subclass of antigen-specific antibodies was tested using a serolab isotyping kit.

4. Results 4.1. Conjugation of the S2b T-cell epitope to lysozyme The S2b peptide was selected as a candidate T-cell epitope. When presented in a I-A d context this peptide induces IL-2 secretion by the T-cell hybridoma HB 68 [14]. In order to perform a controlled covalent bond between the S2b T-cell epitope and lysozyme it was necessary to add a cystein residue to the original sequence of S2b. Since the removal of 3 amino acids from the C-terminal did not abrogate the Tcell epitope activity in vitro (data not shown), this part of the peptide does not contain the T-cell epitope. In order not to interfere with sites close to the actual T-cell epitope, the cystein residue was added to the C-terminal of the S2b peptide, resulting in the new peptide S2bC. To introduce a -SH group in lysozyme, a lysine residue was modified by Traut's reagents [20,21]. Using the 3-dimensional structure of lysozyme and the molecular graphics program BRUGEL [22] the accessibility of the 6 lysine -NH3 + groups was evaluated. Table 1 shows that 5 lysine -NH3 + groups were very accessible while one was much more buried, indicating that several putative attachment sides for the peptide are present. It was evaluated that approximately 0.7 to 0.8 -SH groups were added per lysozyme molecule. We assumed that during conjugation the S2bC peptide would be randomly coupled to different lysine residues of lysozyme. In these conditions it can be expected that following conjugation with the T-cell epitope, no particular B-cell epitope of lysozyme would be destroyed preferentially. A bis-maleimide derivative was used in which the 2 maleimide groups were linked with an acid-labile spacer. The chosen acid-labile spacer was described to be stable at neutral pH for at least several hours but is cleavable at pH 4.5-5.0 with a half-life of the same order of magnitude as the time needed for antigen processing in the endosomes (i.e., about 30 min 27

TABLE 1 ACCESSIBLE S U R F A C E A R E A O F T H E LYSINE RESID U E S IN L Y S O Z Y M E

TABLE 2 P R E S E N T A T I O N O F T H E S2BC/LYSOZYME C O N J U G A T E A N D R E L A T E D PEPTIDES TO T H E S2B-SPECIFIC T-CELL H Y B R I D O M A S HB 68

Lysine residue a Accessible surface area b Contact residue c

APC Lys 1 Lys 13 Lys 33 Lys96 Lys 97 Lys 116

43.3 44.3 0.5 27.9 48.1 39.1

2 2 2 2 2 2

F3, E7, $86, D87 D18 R5, V29, F38, W123 H15, G16, V92, N93 L75, C76, N77 Y23, N106, W i l l

A a TA3 Spleen B b Non-fixed Fixed

aLysine residue in lysozyme, bAccessible surface area of the N zatom of the lysine residues. Radius of a ball rolling over the surface is 1 ,~. CResidues making direct contact with the lysine residue. Note that the contact residues are calculated as residues making at least 1 contact with 1 atom of the lysine residue, and not necessarily with the N atom of this lysine.

Stimulating antigens S2b peptide

S2bC-HEM

S2bClysozyme

50+10 50_ 5 64+4 44 ___4

73+4 47 + 7 N.D. N.D.

52+5 71 ± 6 52+5 12 _ 6

aSupernates of a co-culture of 2.104 TA3 cells or 2.105 spleen cells, 2.104 HB 68 T-cell hybridomas and antigen were added to C T L - L cells. The results are expressed in cpm x 103 of [3H]thymidine incorporation as a result of proliferation of C T L - L cells. The background, co-cultures without antigen, less than 2 × 103 cpm, was subtracted. The m e a n _ SD of triplicate cultures are shown. b2.104 TA3 cells were used as a source of A P C either not treated or fixed for 30 min using 0.5% formaldehyde. The presentation of S2b peptide and the S2bC-lysozyme conjugate were compared. The results are expressed as indicated above.

[23]). At pH 4.0-5.0 a N-(2-hydroxyethyl)-maleimide group remains attached to the -SH groups of released S2bC peptide and of the modified lysine of lysozyme (Fig. 1).

NIASHISSSSARTGDC 8

0

I

Hl'~'--~'O

CB3

.

S

//

%

AcidicpHinenaosomes/~

/

\

5-4 s

NIASHISSSSARTGDC

H--H

I,,I--C~-- Cll2- O-- if

H

Fig. 1. Schematic representation of the acid-cleavable linker between S2b (T-cell epitope) and lysozyme (B-cell epitope). The expected cleaved product upon acid treatment is shown.

28

4.2. The T-cell epitope of S2b is functionally active following conjugation

1.5 m c "o m

.m

0

0.5

:

:

:

:

:

:

:

3

6

9

12

15

18

21

'.

Days after injection. Fig. 2. Anti-lysozymeantibody responsein S2b-primedmice challenged with S2b-lysozymeconjugates or S2b-lysozymemixture. Three Balb/c (H-2a) mice per group were primed with 100 mg of S2b emulsified in complete Freund's adjuvants and boosted 3 weeks later with either S2b-lysozymeconjugate (ll) or a mixture of S2b and lysozyme(A). The results are expressedas the optical density of the anti-lysozymeELISA detection system.

Since the cleavage of the acid-labile spacer is expected to result in a release of a modified S2b peptide (i.e., S2bC-HEM: S2bC conjugated to a N-(2hydroxyethyl)-maleimide), this expected cleavage product was synthesised and tested for its T-cell stimulating properties in vitro. From Table 2A it is clear that the addition of a C-terminal N-(2-hydroxyethyl)-maleimide modified cystein residue does not impair the functional presentation of this T-cell epitope. Hence, hydrolysis of the acid-labile spacer, by treatment at acidic pH, is expected to release a functional T-cell epitope. In order to show that this T-cell epitope is effectively linked to a lysozyme molecule the presentation capacity of S2bC-lysozyme conjugate and S2b peptides by fixed and normal APC was compared. Table 2B shows that the S2bC-lysozyme conjugate is effectively presented by normal APC but only marginally by fixed APC. The negative control using lysozyme alone as antigen did not stimulate the T-cell hybridomas as compared to co-cultures without antigen (data not shown). These results indicate that the S2bC-lysozyme conjugate needs a form of processing, through functional APC, in order to present the S2b epitope to the T-cell hybridomas. Hence, the stimulation of the T-cell hybridomas is not due to contaminating free S2b peptides, either present at the beginning of the experiment or formed by hydrolysis of the acid-labile linker in vitro. B

A

0.8

0.8

0.7

0.7

0.6

0.6

'~ 0.5

'~ 0.5

m

C

: 0.4

0.4

~0.3 0

~:.l 0.3 O

0.2 0.1

0.2

I,I

....

IgM IgG1 IgG2a IgG2b IgG3

o: Ig

, IgM IgG1 IgG2a IgG2b IgG3

Fig. 3. Subclass of the anti-lysozyme antibodies in S2b-primed mice elicited with S2b-lysozyme conjugates or S2b-lysozyme mixtures. Three Balb/c (H-2d) mice per group were primed with 100 mg of S2b emulsified in complete Freund's adjuvants and boosted 3 weeks later with either a mixture of S2b and lysozyme (A) or S2b-lysozyme conjugate (B). The sera collected at days 3, 6 and 9 were pooled and subjected to a lysozyme-specific ELISA. The subclass of the anti-lysozyme antibodies was determined using subclass-specific secondary antibodies. The results are shown as the mean of 3 independent replicates.

29

A

B

100

1.2

"

75

'~

CO O T-

x =S

0.8 -

50

'0 0.6-

26

~ 0.4 0 0.2

O.

O

m

l

l

30

60

90

Time of pulsing

0

I

I

I

I

I

I

6

0

12

15

18

21

Days after Booster Injection

Fig. 4. Anti-lysozyme antibody response in S2b-primed low and high responder mice. A: in vitro T-cell proliferative response of Balb/c (0, H-2 a) and C57B1/6 (m H.2 b) mice to the S2b T-cell epitope. Mice were primed intra-footpad with 100 mg of S2b peptide emulsified in complete Freund's adjuvants. B: immunogenicity of the S2b-lysozyme conjugate in high ( 0 , Balb/c H-2 d) and low (ll, C57B1/6 H-2 b) responder mice strains. Mice were immunized with 100 mg of S2b peptide emulsified in complete Freund's adjuvants. Three weeks later the S2b-lysozyme conjugate was injected. The presence of anti-lysozyme antibodies was detected at different time intervals using a lysozymespecific ELISA.

4.3. Anti-lysozyme response using S2bC-lysozyme conjugate as booster antigen The results, shown in Fig. 2, indicate that mice primed with S2b peptides and subsequently challenged with the S2bC-lysozyme conjugate, develop a faster and higher anti-lysozyme response as compared to control animals that were challenged with a mixture of lysozyme and S2b peptides. The kinetics of the anti-lysozyme immune response suggest that a secondary-type of immune response is induced in mice challenged with S2bC-lysozyme conjugates. This possibility was corroborated by analyzing the isotypes of the generated anti-lysozyme antibodies. The results presented in Fig. 3 indicate that mice injected with the S2bC-lysozyme conjugate produced more lysozyme-specific IgG antibodies as compared to mice challenged with a mixture of S2b peptides and lysozyme. This result confirms that a single immunisation, of S2b-primed mice, with the S2bC-lysozyme conjugate resulted in a secondary-type anti-lysozyme response. 4.4. Anti-lysozyme response against the S2bClysozyme conjugate in S2b non-responder mice In order to prove that the observed immunopotentiating effect is truly T-cell dependent, the immunogenicity of the S2bC-lysozyme conjugate was compared in S2b responder and non-responder mouse strains. As shown in Fig. 4A, S2b-primed lymph 30

node cells from C57bl/6 mice (H-2 b) do not proliferate in vitro after antigen-specific re-stimulation indicating that this mouse strain, in contrast to BALB/c mice (H-2d), is non-responsive to S2b. The immunogenicity of S2bC-lysozyme conjugates in both mouse strains, primed with the S2b peptide, was compared. From Fig. 4B, it is clear that only Balb/C (H-2 d) mice, in which the S2b peptide is functional, produce anti-lysozyme antibodies, whereas C57B1/6 mice (H2 b) fail to generate an anti-lysozyme response. Thus, these results clearly demonstrate that the immunopotentiating effect of S2b on the anti-lysozyme humoral immune response is MHC restricted.

5. Discussion

Lysozyme was shown to possess only one natural T-cell epitope, capable of being presented in a H-2 d context [24], and this antigen has a low immunogenicity in BALB/c mice. Aiming at enhancing the humoral immune response to this antigen an additional T-cell epitope (i.e., S2b) was covalently linked to lysozyme. The conjugation of the S2b T-cell epitope to lysozyme did not abolish the immunological properties of S2b as tested by the capacity of this conjugate to stimulate specific T-cell hybridomas in the presence of functional APC. Since fixed APC were capable of presenting free S2b peptides but not the conjugate the observed presentation was not due to contaminating free S2b peptides.

By coupling T-cell epitope containing peptides to weak immunogens the molecular context in which this antigenic peptide is placed can be drastically altered. Hence, this might influence the correct processing of the conjugate, resulting in an impaired release of the original T-cell epitope and consequently in an abolishment of a potential T-helper effect. In order to circumvent this potential problem an acid-labile linker was generated between S2b (T-cell epitope) and lysozyme (B-cell epitope). Such link would be hydrolyzed during the passage of the antigen through the endosomal compartment where the pH drops from neutral to 4.5-5. As such this type of linker would assure a release of the peptide in the intracellular compartment where it would most probably associate with the MHC class II molecule [25]. In vivo experiments showed a fast and increased humoral immune response in mice primed with S2b and boosted with the conjugate as compared to animals challenged with a mixture of lysozyme and S2b. Moreover, both the kinetics of the antibody production as well as the isotype of the generated antibodies suggest that a secondary-type immune response has been induced. Since in these experiments, the animals encountered lysozyme for the first time, this result implies that naive B cells can be stimulated to produce IgG, provided adequate T-cell help is available. However, we observed sporadically that mice challenged with a mixture of S2b and lysozyme could develope a higher anti-lysozyme response, as compared to mice injected with lysozyme alone (data not shown). This observation is in accordance with the recent investigation reporting that high antibody titers could be induced in mice injected with a mixture of a T-cell epitope containing peptide and protein [26]. The reason why in our study, only a minor proportion of the mice reacted similarly might be related to the fact that in our experimental set-up the mixture was injected in PBS while in the case of Sarobe et al. [26] the antigens were emulsified in adjuvants. It is conceivable that emulsifying the mixture of antigens in adjuvants limits the diffusion of the antigens in vivo and as such promotes a close proximity of both the T-cell epitope and the B-cell epitope. A close proximity of both epitopes in vivo was proposed by Sarobe et al. [26] to account for the observed enhanced immunogenicity. In S2b non-responder C57B1/6 (H-2 b) mice which had been primed with S2b, no anti-lysozyme response could be obtained with the S2bC-lysozyme conjugates when priming these mice with S2b. Since in non-responder mice no S2b-specific T-helper cells

are expected to be induced, this result corroborates with the notion that S2b-specific T-helper cells provide help to the lysozyme-specific B cells during challenge with the S2bC-lysozyme conjugates. It should be noted that our results do not imply that the acid-labile link is an absolute requirement and that stable covalent bonds between T- and B-cell epitopes would be unsuitable. However, it can be expected that the use of a cleavage site, independent of processing enzymes, between the T- and B-cell epitope, may be required for some but not all T/B-cell epitope combinations. Furthermore, recent results suggest that disulphide bonds are cleaved during antigen processing [27]. Hence a disulphide bridge between the T- and B-cell epitope might represent an attractive alternative for the acid-cleavable linker.

Acknowledgements J.P.Y. Scheerlinck is a research assistant of the Nationaal Fonds voor Wetenschappelijk Onderzoek (NFWO). P. De Baetselier is a research associate of the NFWO. The authors thank Dr. I. Lasters from Plant Genetic Systems (Gent, Belgium) for his help using the BRUGEL software. This work was supported by grants from the Fond National pour la Recherche Scientifique (to A. Michel) and the Governmental VLAB grand (to P. De Baetselier).

References [1] Etlinger, H.M., Gillessen, D., Lahm, H.-W., Mattle, H., Sch6nfeld, H.-J. and Trzeciak, A. (1990) Science 249, 423. [2] Michison, N.A. (1971) Eur. J. Immunol. 1, 10. [3] Abbas, A.K. (1988) Immunol. Today 9, 89. [4] Milich, D.R., McLachlan, A., Thornton, G.B. and Hughes, J.L. (1987) Nature 329, 547. [5] Schutze, M.-P., Leclerc, C., Vogel, F.R. and Chedid, L. (1987) Cell. Immunol. 104, 79. [6] Herzenberg, L.A. and Tokuhisa, T. (1980) Nature 285, 664. [7] Herzenberg, L.A., Tokuhisa, T. and Hayakawa, K. (1983) Ann. Rev. Immunol. 1,609. [8] Schutze, M.-P., Deriaud, E., Przewleocki, G. and Leclerc, C. (1989) J. Immunol. 142, 2635. [9] Herzenberg, L.A. and Tokuhisa, T. (1982) J. Exp. Med. 155, 1730. [10] Galelli, A. and Chariot, B. (1990) J. Immunol. 145, 2397. [11] Sercarz, E.E. and Krzych, U. (1991) Immunol. Today 12, 111. [12] Sercarz, E.E., Yowell, R.L., Turkin, D., Miller, A., Araneo, B.A. and Adorini, L. (1978) Immunol. Rev. 39, 108. [13] Paninabordignon, P., Tan, A., Termijtelen, A., Demotz, S., Corradin, G. and Lanzavecchia, A. (1989) Eur. J. Immunol. 19, 2237. [14] Scheerlinck, J.-P.Y., Burssens, G., Brys, L., Hauser, P. and De Baetselier, P. (1991) Immunology 73, 88.

31

[15] Merrifield, R.B., Vizioly, L.D. and Boman, H.G. (1982) Biochemistry 21, 5020. [16] Srinivasachar, K. and Neville, D.M.J. (1989) Biochemistry 28, 2501. [17] Glimcher, L.H., Hamano, T., Asofsky, R., Sachs, D.H., Pierres, M., Samelson, L.E., Sharrow, S.O. and Paul, W.E. (1983) J. Immunol. 130, 2287. [18] Kovac, Z. and Schwartz, T.Z. (1985) J. Immunol. 134, 3233. [19] Gillis, S., Ferm, M.M., Ou, W. and Smith, K.A. (1978) J. Immunol. 120, 2027. [20] Lambert, J.M., Jue, R. and Traut, R.R. (1978) Biochemistry 17, 5406. [21] Jue, R., Lambert, J.M., Pierce, L. and Traut, R.R. (1978)

32

Biochemistry 17, 5399. [22] Delhaise, P., Bardiaux, M. and Wodak, S.J. (1984) J. Mol. Graph. 2, 103. [23] Harding, C.V. and Unanue, E.R. (1989) J. Immunol. 142, 12. [24] Adorini, L., Sette, A., Buus, S., Grey, H.M., Darsley, M., Lehmann, P.V., Doria, G., Nagy, Z.A. and Appella, E. (1988) Proc. Natl. Acad. Sci. USA 85, 5181. [25] Jensen, P.E. (1991)J. Exp. Med. 174, 1111. [26] Sarobe, P., Lasarte, J.J., Golvano, J., Gull6n, A., Civeira, M.P., Prieto, J. and Borrhs-Cuesta, F. (1991) Eur. J. Immunol. 21, 1555. [27] Collins, D.S., Unanue, E.R. and Harding, C.V. (1991) J. lmmunol. 147(12), 4054.