Hybrid hepatitis B virus core antigen as a vaccine carrier moiety: I. Presentation of foreign epitopes

,“VYNAL

OI

Biotechnology ELSEVIER

Journal of Biotechnology 44 (1996) 91-96

Hybrid hepatitis B virus core antigen as a vaccine carrier moiety: I. Presentation of foreign epitopes F. Schiidel ‘,*, D. Peterson b, J. Hughes d, R. Wirtz ‘, D. Milich d a INSERM U 80, Pavilion P, HGpital Edourd Herriot, 69437 Lyon Cedex b Department of Biochemistry, Virginia Commonwealth University, Richmond, ’ Department of Entomology, Walter Reed Army Institute of Research, Washington, d Department of Molecular Biology, The Scripps Research Institute, La Jolla,

03, France VA, 23298, USA DC 20307-5100, USA CA, 92037, USA

Received 3 August 1995; accepted 7 August 1995

Abstract Hepatitis B virus (HBV) core antigen (HBcAg) is a highly immunogenic subviral particle. Here, we review recent progress in the use of HBcAg as a carrier moiety for heterologous epitopes. To define surface exposed and immunogenic insertion sites for foreign epitopes in HBcAg, peptidic epitopes representing binding sites for virus neutralizing antibodies on the HBV surface antigens were inserted at different positions within HBcAg using genetic engineering in an Escherichia coli expression system (Schodel et al. (1992) J. Virol. 66, 106- 114). While fusion to the N-terminus required a linker to become surface accessible, both fusion to the N-terminus and to the C-terminus was compatible with particle assembly and preserved the native antigenicity and immunogenicity of HBcAg. Fusion to an immunodominant internal site of HBcAg reduced the HBcAg immunogenicity and antigenicity and most drastically enhanced the immunogenicity of the inserted foreign epitope. This internal site of HBcAg was used to express circumsporozoite antigen (CS) repeat epitopes of two rodent malaria parasites and of Plasmodium falciparum (Schijdel et al. (1994b) J. Exp. Med. 180, 1037-1046 and Schodel et al. (1995a) 95th ASM General Meeting, Washington DC, Abstr. E61). When purified from recombinant Salmonella typhimurium, the hybrid HBcAg-CS proteins were particulate and displayed CS antigenicity as well as reduced HBc antigenicity, as compared to native HBcAg. Immunization of several mouse strains with HBcAg-CS hybrid particles resulted in high titered serum anti-CS antibodies representing all murine IgG isotypes. Immunization of mice with HBcAg or HBcAg-CS particles formulated on alum, complete Freunds or incomplete Freunds adjuvant resulted in equivalent anti-CS and anti-HBc serum antibody times. The possible influence of carrier-specific immunosuppression was examined and pre-existing immunity to HBcAg did not significantly alter the immunogenicity of hybrid HBcAg particles suggesting that they would be useful carrier moieties for repeated immunizations against multiple haptens or in immune subjects after HBV infection. Examination of T cell recognition of HBcAg-CS particles revealed that HBcAg-specific T cells were universally primed and CS-specific T cells were primed if the insert contained a CS-specific T cell recognition site. This indicates that the internal amino acid position in HBcAg is permissive for the inclusion of heterologous functional T helper as well as B cell epitopes. BAL,B/c mice immunized with HBcAg-CSl were protected against P. berghei challenge to 90% and lOO%, respectively, in two independent experiments. Keywords:

HBcAg; Hepatitis B virus; Core particle; Epitope presentation; Malaria; Circumsporozoite antigen; Malaria vaccine

* Corresponding author. 0168-1656/96/$15.00 0 1996 Elsevier Science B.V. All rights reserved SSD!O168-1656(95)00118-2

92

F. Schiidel

et al. / Journal

of Biowchnology

1. Introduction Hepatitis B virus (HBV) core antigen (HBcAg) is a 21-kDa protein that self-assembles to form the subviral 30-32 nm nucleocapsid particles packaging the viral polymerase and pregenomic RNA during HBV replication. HBcAg also assembles to form particles when synthesized in the absence of other HBV gene products in a wide range of prokaryotic and eukaryotic recombinant expression systems. Two particle sizes are observed, a slightly smaller particle with a T = 3 symmetry consisting of 180 subunits and a larger particle consisting of 240 subunits and displaying a T = 4 symmetry (Crowther et al., 1994). The latter is the predominant species. Apart from the ease of recombinant expression and self assembly, HBcAg has several features that make it an attractive carrier moiety for foreign haptens (reviewed in: Milich, 1988; Schijdel et al., 1994a). HBV-infected individuals mount a high titered anti-HBc antibody response indicative of an enhanced immunogenicity of HBcAg in man, at the same time its expression is apparently non-toxic. HBcAg can directly activate B cells (it is a T cell dependent as well as T cell independent antigen (Milich and McLachlan, 1986). HBcAg as a protein antigen elicits strong T cell responses (Milich et al., 1987a) and HBcAg-specific Th cells can cooperate with envelope (HBsA~)specific B cells to produce anti-HBs (Milich et al., 1987b). When H-2 congenic mouse strains are immunized with HBcAg, an MHC-dependent hierarchy of the HBcAg T cell responses and in consequence of the anti-HBc antibody titres is observed. However, every mouse strain tested reacts with high titered anti-HBc antibodies and no non-responder strain has been identified. The particulate nature of HBcAg is necessary but not sufficient for this enhanced immunogenicity. Immunization of H-2 congenie mice with HBsAg for example, another particulate antigen within a similar size range, reveals a much lower immunogenicity and a number of nonor low responder strains (Milich et al., 1987a, b). To further char-acterize HBcAg as a vaccine carrier moiety, we have at first established prokaryotic expression systems for HBcAg and then defined insertion sites for heterologous epitopes (Refs. in SchMel et al., 1992). In a more recent development, we have used an internal permissive site of HBcAg

44 (I 996)

91-96

thus defined for the expression of circumsporozoite antigen (CS) repeat epitopes on malaria parasites (Schiidel et al., 1994b, 1995a). With these hybrid HBcAg-CS particles we have compared immunogenicity on different adjuvants, studied the impact of carrier-specific immunosuppression, the inclusion of a heterologous Th cell epitope and protection against plasmodial challenge in two rodent malaria systems. Finally, we have developed Salmonella spp. based expression systems and studied the immunogenicity of oral and other mucosal route immunizations with recombinant Salmonella expressing HBcAg and HBcAg hybrids (to be reviewed separately in:Schijdel et al., 1995b).

2. Experimental

and discussion

2.1. Influence of position on immunogenicity of guest epitopes within HBcAg To study the influence of epitope location on immunogenicity of hybrid particles, we first inserted virus neutralizing epitopes from the pre-S region of HBV at the N-terminus of HBcAg, at the N-terminus through a precore linker as previously employed for other epitopes (Clarke et al., 19871, at a truncated C-terminal position, or at an internal position after amino acid 75 of HBcAg. These studies revealed that in order to render inserted epitopes surface accessible and immunogenic at the HBcAg core N-terminus, it is necessary to insert a linker sequence which can be derived from the precore region between the core N-terminus and the epitope of choice (Schijdel et al., 1992). The very N-terminus of HBcAg appears not to be surface exposed (Schijdel et al., 1992). Whereas surface accessibility of the inserted heterologous epitope is required, it is not sufficient for immunogenicity. Various amino acid sequences inserted at a truncated C-terminus became at least partly surface accessible as analyzed by direct and capture ELISA techniques but did not acquire the same level of immunogenicity as sequences fused to the N-terminus through the precore linker (Schijdel et al., 1992). Finally, an internal site of core antigen predicted by immunologic analysis to represent an immunodominant site was used for insertion of heterologous epitopes (Schijdel et al., 1992).

F. Schijdel

et al./

Journal

of Biotechnology

Epitopes from a variety of pathogens have been inserted into HBcAg (see Schijdel et al., 1994a). N-terminal, C-terminal and internal fusions of heterologous sequences resulted in the synthesis of HBcAg-like particles (SchBdel et al., 1992). Insertion of sequences at the N-terminus and at the Cterminus leaves the native antigenicity and immunogenicity of HBcAg largely intact. Such chimaeric particles also retain the T cell independence of antiHBc if not of anti-heterologous epitope antibody production. In contrast, insertions between amino acids 75 and 81 delete a major HBcAg antibody recognition site, reducing the reactivity of chimaeric particles with anti-native HBc polyclonal sera and abolishing their recognition by 5/6 anti-HBc monoclonals (Schiidel et al., 1992). These deletions defined a second epitope on HBcAg recognized by mutually crossinhibiting antibodies, which were heretofore believed to recognize the same epitope. Immunization with these internal fusion particles also drastically reduced the native HBcAg immunogenicity (while enhancing the immunogenicity of the carried epitopes). The internal fusions are thus not only more immunogenic, they also reduce the level of anti-HBc produced and would potentially escape anti-carrier antibody mediated suppression of immune responses against the carried epitope if this should present a problem. Comparing the immunogenicity of inserted epitopes we found that insertion at the internal site of HBcAg resulted in by far the highest degree of immunogenicity conferred to the guest epitopes (Schijdel et al., 1992). Fusions to the amine terminus through a precore linker were also highly immunogenic, whereas fusions to the Cterminus became surface accessible but did not achieve a similar level of enhanced immunogenicity. The internal site between 75 and 81 seems to accomodateheterologoussequencesof up to at least 45 amino acids (Schiidel, unpublisheddata). Whether or not insertion of a heterologous sequenceis compatible with core particle assembly is not only dependent on the length of the sequencebut on the specific primary sequencecontext. We have encountered two relatively short sequencesthat resulted in insoluble core material and one that apparently destabilized HBcAg resulting in proteolytic degradation. Fusions at the C- or N-terminus may allow insertion of longer heterologous amino acid se-

44 f 1996)

91-96

93

quences (Refs. in SchBdel et al., 1994a). However, these fusions do not reach the same level of enhanced immunogenicity as the internal fusions.

3. Plasmodial

CS HBcAg

hybrid

particles

Having defined the internal site of HBcAg between amino acids 75 and 81 as a useful position for inserting foreign epitopes we turned our attention to an unresolved problem in malaria vaccination. When a mosquito transmits sporozoites to a vertebrate host, the sporozoites are covered with a surface protein, the circumsporozoite antigen (CS) containing a high number of short repetitive amino acid sequences that constitute immunodominant antibody recognition sites (Refs. in Schijdel et al., 1994b). Antibodies against these sequences are elicited during infection and can protect mice against challenge with rodent parasites. It is still unclear whether high titered antibody responses against CS could protect humans from P. fulciparum infection due to the fact that it has proven difficult in spite of considerableefforts to elicit and maintain high titered antibody responses against P. fulciparum CS in man. Similarly, it has been shown that passive transfer of a monoclonal antibody against P. yoelii CS could protect mice from challenge (Charoenvit et al., 1991). However, active immunization regimens with e.g. synthetic peptides containing the epitopes recognized by the samemonoclonal antibody failed to elicit protective antibody levels. We therefore decided to insert the well-defined CS repeat epitopes of P. berghei, P. yoelii or P. fulcipurum into hybrid HBcAg-CS particles and analyze their antigenicity, immunogenicity and in their protective efficacy in rodent models. At the sametime we used thesehybrid HBcAg-CS particles to analyze their effect on IgG isotype distribution, the use of alum hydroxide as an adjuvant acceptablefor human use, the functional inclusion of Th epitopes and the effect, if any, of pre-existing carrier-specific immunity. For this purpose a P. berghei CS repeat (DP,NPN), [ = HBcAg-CSl I and a P. jiilcipurum CS repeat (NANP), [ = HBcAg-CS2] (SchMel et al., 1994b), as well as a P. yoelii CS repeat epitope (SchGdel et al., 1995a) were inserted into HBcAg by genetic engineering of an expression vector. The expression products obtained after trans-

94

F. Schijdel

et al.//ournal

of Biotechnology

formation of the expression vectors into avirulent S. fyphimurium displayed CS, pre-S and HBcAg reactivity by Western blot analysis. The resulting hybrid particles were purified from recombinant Salmonella typhimurium. Electron micrography revealed hybrid HBcAg-like particles. To analyze immunogenicity, several congenic MHC disparate mouse strains were immunized with hybrid HBcAg-CS particles. HBcAg-CS 1 and HBcAg-CS2 particles elicited high titered anti-CS and anti-peptide antibodies. To study the CS epitope-specific IgG isotype distribution, BALB/c mice were immunized with 10 lg HBcAg-CSl in CFA followed by 10 pg in IFA. Similar to the distrubution of isotypes after immunization with native HBcAg, immunization of BALB/c mice with HBcAg-CSl particles elicited a primary anti[DP,NANP], antibody response consisting of all IgG isotypes, dominated by IgG2a (Table 1). After a boost very high CS epitope-specific antibodies of all isotypes could be found. This demonstrated that HBcAg as a carrier moiety elicits a broad IgG isotype response specific for the inserted epitope and will be useful regardless of which IgG isotype is required. For vaccine purposes an antigen must be immunogenic when formulated with an adjuvant suitable for human use. The only adjuvant currently licensed as a component of vaccines for human use is alum. We therefore investigated whether the enhanced immunogenicity of HBcAg and of the carried epitope in a HBcAg-CS hybrid would be retained when alum is used as an adjuvant. To address this question BALB/c mice were immunized with HBcAg or

Table 1 Effect of HBcAg-CS

immunization

on anti-[DP,NPN],

IgG isotype Antibody

44 (1996)

Table 2 Effect of adjuvant

91-96

on immunogenicity

of HBcAg Antibody

and HBcAg-CSl titre (l/I

lmmunogen

Adjuvant

Time

HBcAg

[DP,NPNI,

HBcAg

CFA

2 4 2 4 2 4

163840 655360 163840 655360 163840 655360

0 0 0 0 0 0

2 4 2 4 2 4

2560 10240 10240 10240 10240 40960

10240 163840 2560 163840 10240 655360

IFA alum

HBcAg-CS

1

CFA IFA alum

Groups of three BALB/c mice were imunized with either 10 pg of HBcAg or hybrid HBcAg-CSl particles on alum, CFA, or IFA as indicated. Sera were collected before immunization and 2 and 4 weeks after primary immunization. Sera were pooled and analyzed by solid-phase ELISA with HBcAg (50 ng per well) and the P. berghei CS repeat synthetic peptide [DP,NPN], (1 .O pg per well) as the solid-phase ligands. (Modified from Schiidel et al., 1994a.)

HBcAg-CSl particles formulated with CFA, IFA or alum and the primary serum anti-HBc and anti[DP,NPN], IgG responses were analyzed (Table 2). The primary anti-HBc and anti-[DP,NPN], IgG serum responses were nearly identical regardless of which adjuvant was employed. The data demonstrate that HBcAg can be used as a vaccine carrier moiety with alum, an acceptable adjuvant for human use. Note that with all three adjuvants used, the anti-HBc response elicited by HBcAG-CS 1 is reduced as compared to that obtained with HBcAg. A reduced

distribution

titre (1,‘)

strain

Time

Total IgG

IgG 1

IgG2a

IgG2b

IgG3

BALBc

10d 24d 2”

10240 40 960 > 655 360

640 40960 > 655 360

640 655 360 > 655 360

2560 163840 > 655 360

2.560 163940 > 655 360

Groups of three BALB/c mice were imunized with 10 pg of hybrid HBcAg-CSl particles in CFA and boosted with 10 fig in IFA. Sera were collected before immunization, IO and 24 days after primary and 2 weeks after secondary (2’) immunization. Sera were pooled and analyzed by solid-phase ELISA: A P. berghei CS repeat synthetic peptide [DP,NPN], (1.0 wg per well) was the solid-phase ligand, goat anti-mouse IgG isotype specific antibody was the second antibody and was developed with a peroxidase-labeled swine anti-goat IgG. Data are expressed as antibody titres representing the highest dilution of serum required to yield 3-times the OD492 of preimmune sent.

F. Schiidel

et al./

Journnl

of Biotechnology

HBcAg antigenicity at the B cell level of hybrid HBcAg particles may be useful to prevent carrierspecific antibody production and recognition by preexisting carrier-specific antibodies. Pre-existing carrier protein-specific immunity has been reported to negatively influence the immunogenicity of carried haptens (Herzenberg and Tokuhisa, 1983). We have therefore investigated the influence of pre-existing anti-HBc antibodies on subsequent immunization with HBcAg-CSl. For this purpose, BALB/c mice were primed with HBcAg or controls followed by two immunizations with HBcAg-CSl particles 1 and 2 months later. At 24 days after the first injection of HBcAg-CSl, the HBcAg primed group possesed the lowest titre of anti-[DP,NPN], suggesting a degree of carrierspecific suppression. However, at week 2 and subsequently after the second HBcAg-CSl injection, no evidence of carrier-specific suppression was found and the HBcAg primed group produced extremely high titered anti-[DP,NPN], antibody (Schodel et al., 1994b). Therefore, hybrid HBcAG particles are highly effective as a vaccine carrier moiety even in the presence of pre-existing immunity to HBcAg. This suggests that they could be used in subjects with a pre-existing HBcAg immunity due to a prior or concomitant HBV infection and that they could be used repeatedly as carriers for multiple haptens. If hybrid HBcAg particles are to be used as carriers for non-HBV epitopes it could be of interest to include functional T help sites from the pathogen of interest which would possibly be available for recall memory. Previous data suggested that an Nterminally fused Th epitope did not become functional (Schodel et al., 1992) and a COOH-terminal fused Th epitope did become functional (Borisova et al., 1993). We therefore investigated the T cell recognition of HBcAg-CS hybrid particles. Immunization of three H-2 congenic murine strains with HBcAg-CS 1 and HBcAgCS2 hybrids primed HBcAg-specific T cells. Immunization of B 10 (H-2 b, mice with HBcAg-CS2 particles aditionally primed [NANPI,-specific T cells and vice versa, [NANP], synthetic peptide primed BlO T cells were also recalled with HBcAg-CS2 particles. This demonstrates that Th sites can be functionally included in hybrid HBcAg particles in at least the internal site. To study the protective efficacy of HBcAg-CS

Table 3 Protection

44 (1996)

against

91-96

95

Plamzodium

herghei

challenge

Infected/noninfected

(% protection)

lmmunogen

Expt.

Expt. 2

PBS HBcAg HBcAg-CS

8/9(11%) 9/10(10%0) 0/10(100%)

1

1

9/10 (10%) lO/lO (0%) l/10(90%)

Groups of ten female BALB/c mice were imunized with PBS, 20 pg of HBcAg or 20 @g of hybrid HBcAg-CSl particles all emulsified in CFA and boosted with identical antigen doses in IFA 2 weeks later. 3 months after immunization, the mice were challenged with P. herghei by feeding P. berghei Anka-infected A. stephensi mosquitoes on their tails. Parasitemia was monitored by microscopic examination of Giemsa-stained blood films daily for up to 25 days after challenge. The experiment was repeated a second time under identical conditions (modified from SchMel et al., 1994a).

hybrid particles we immunized BALB/c mice with PBS (in CFA), HBcAg or HBcAg-CSl particles and challenged them by feeding P. berghei-infected mosquitoes on their tail veins. This route of challenge closely mimics the natural route of infection. Between 90- 100% of PBS or HBcAg (both in CFA) immunized control mice developed parasitemia 4-6 days after challenge (Table 3). In contrast, 1O/ 10 or 9/10 mice were protected against P. berghei challenge 3 months after the last immunization with HBcAg-CSl particles in two independent experiments. Protection is mediated by polyclonal CSspecific antibodies as immunization with hybrid HBcAg-CSl particles did not elicit CS-specific T cells (Schiidel et al., 1994b). More recently we have immunized BALB/c mice with hybrid HBcAg-CS particles carrying P. yoelii repeat epitopes and could demonstrate similar levels of protection against P. yoelii challenge (Schiidel et al., 1995a). In conclusion, hybrid HBcAg-CS particles might allow us to test whether high titered anti-CS antibodies are protective in humans and, if so, might become a useful part of future malaria vaccines.

Acknowledgements Parts of this research were supported by National Institute of Health grants AI20720 and AI33562 (DRM, DP, FS) and by the World Health Organisation Transdisease Vaccinology Program (FS).

96

F. Schiidel

et al./Journnl

References Borisova, G., Biragyn, A., Dislers, A., Borschukova, O., Tsibinogin, V., Skrastina, D., Eldarov, M.A., Pumpens, P., Skryabin, K.G. and Grens, E. (1993) Hybrid hepatitis-B virus nucleocapsid bearing an immunodominant region from hepatitis-B virus surface-antigen. J. Viral. 67, 3696-3701. Clarke, B.E.. Newton, S.E., Carroll, A.R., Francis, M.J., Appleyard, G., Syred, D., Highfield, P.E., Rowlands, D.J. and Brown, F. (1987) Improved immunogenicity of a peptide epitope after fusion to hepatitis B core protein. Nature 330, 381-384. Charoenvit, Y., Mellouk, S., Cole, C., Bechara, R., Leef, M.F., Sedegah, M., Yuan, L.F., Robey. F.A., Beaudoin, R.L. and Hoffman, S. (1991) Monoclonal, but not polyclonal, antibodies protect against Pkwnodium yoelii sporozoites. Infect. lmmun. 146, 1020-1025. Crowther, R.A., Kiselev, N.A., Biittcher, B., Berriman, J.A., Borisova, G.P., Ose, V. and Pumpcns, P. (I 994) Three-dimensional structure of hepatitis B virus core particles determined by electron microscopy. Cell 77, 943-950. Herzenberg, L.A. and Tokuhisa, T. (1983) Epitope-specific regulation. 1. carrier-specific induction of suppression for IgG anti-hapten antibody response. J. Exp. Med. 155, 1730. Milich, D.R. and McLachlan, A. (1986) The nucleocapsid of the hepatitis B virus is both a T-cell in dependent and a T-cell dependent antigen. Science 234, 1398. Milich, D.R., McLachlan. A., Moriarty, A. and Thornton, G.B. (1987al Immune responses to hepatitis B virus core antigen (HBcAg): localization of T cell recognition sites within HBcAg/HBeAg. J. Immunol. 139, 1223.

of Biorechwlogy

44 (1996)

91-96

Milich, D.R., McLachlan, A., Thornton, G.B. and Hughes, J.L. (198?b) Antibody production to the nucleocapsid and envelope of the hepatitis B virus primed by a single synthetic T cell site. Nature 329, 5477549. Milich, D.R. (1988) T and B cell recognition of hepatitis B viral antigens. lmmunol. Today 9, 380-386. SchMel, F., Moriarty, A.M., Peterson, D.L., Zheng, J., Hughes, J.L., Will, H., Leturcq, D.J., McGee, J.S. and Milich, D.R. (1992) The position of heterologous epitopes inserted in hepatitis B virus core particles determines their immunogenicity. J. Viral. 66. 106- 114. Schijdel, F., Peterson, D.. Hughes, J. and Milich, D.R. (1994a) Hepatitis B virus core particles as a vaccine carrier moiety. Int. Rev. Immunol. 1 I, 153-164. Schiidel, F., Wirtz, R., Peterson, D., Hughes, J.. Warren, R., Sadoff, J. and Milich, D. (1994b) Immunity to malaria elicited by hybrid hepatitis B virus core particles carrying circumsporozoite protein epitopes. J. Exp. Med. 180, 1037- 1046. Schiiclel, F., Peterson, D., Charoenvit, Y., Sadoff, J. and Wirtz, R. (1995a) Immunity to Plasmodium yoelii infection elicited by hybrid hepatitis B virus core particles (HBcAg) carrying circumsporozoite antigen epitopes. 95th ASM General Meeting, Washington DC, Abstr. E61. Schiidel, F., Kelly, S.. Tinge, S., Hopkins, S., Peterson, D. and Curtiss, R. (I 995b) Hybrid hepatitis B virus core antigens as a vaccine carrier moiety: II. Expression in avirulent Salmonrllu spp. for mucosal immunization. In: Shaffemran, A. and Cohen, S. (Eds.), Vaccines: Novel Strategies in Design and Production. Proceedings of the 39th OHOLO Conference 1995. Plenum Press, New York, in press.