Composition for treating HBV infection

Abstract

The present invention provides a composition comprising hepatitis B virus (HBV) component(s), and which may be either nucleic acid- or polypeptide-based as well as nucleic acid molecules and vectors encoding such HBV component(s). It also relates to infectious viral particles and host cells comprising such nucleic acid molecules or vectors. It also provides composition and kits of parts comprising such nucleic acid molecules, vectors, infectious viral particles or host cells and the therapeutic use thereof for preventing or treating HBV infections.

Claims

1. An infectious adenoviral particle comprising a nucleic acid molecule encoding (i), (ii), and (iii) polypeptides wherein (i) is a polymerase moiety comprising at least 450 amino acid residues of a polymerase protein originating from a genotype D HBV virus; (ii) is a core moiety comprising at least 100 amino acid residues of a core protein originating from a genotype D HBV virus; and (iii) is an env moiety comprising a first amino acid sequence set forth in SEQ ID NO: 12 and a second amino acid sequence set forth in SEQ ID NO: 13, and wherein said env moiety does not include any immunogenic domain(s) originating from preS1 and preS2 regions; wherein either: (a) at least 45 and at most 50 amino acid residue truncation of (i) at the N-terminus of a native HBV polymerase protein or (b) at least 34 and at most 37 amino acid residue truncation of (ii) at the C-terminus of a native HBV core protein; wherein said infectious adenoviral particle is capable of inducing a T cell response against at least one of said HBV moieties and said infectious adenoviral particle is E1-defective.

2. The infectious adenoviral particle of claim 1, wherein said infectious adenoviral particle is for use in a subject infected or suspected to be infected with an HBV from a genotype D.

3. A host cell comprising the infectious adenoviral particle according to claim 1.

4. A composition comprising the infectious adenoviral particle according to claim 1 and a pharmaceutically acceptable vehicle.

5. The composition according to claim 4 which further comprises one or more adjuvant(s) suitable for systemic or mucosal application in humans.

6. The composition according to claim 4 which is formulated for intramuscular or subcutaneous administration.

7. The composition according to claim 4, which comprises from about 10.sup.5 to about 10.sup.13 infection units of an adenoviral vector or of an infectious adenoviral particle.

8. The infectious adenoviral particle according to claim 2, a host cell comprising the infectious adenoviral particle according to claim 2, or the composition comprising the infectious adenoviral particle according to claim 2, wherein said use is used in combination with standard of care.

9. The infectious adenoviral particle according to claim 1, a host cell comprising the infectious adenoviral particle according to claim 1, or a composition comprising the infectious adenoviral particle according to claim 1, wherein said genotype D HBV viruses are from HBV isolate Y07587.

10. An adenoviral vector comprising nucleic acid molecules encoding (i), (ii), and (iii) polypeptides, wherein (i) is a polymerase moiety comprising at least 450 amino acid residues of a polymerase protein originating from a genotype D HBV virus; (ii) is a core moiety comprising at least 100 amino acid residues of a core protein originating from a genotype D HBV virus; and (iii) is an env moiety comprising a first amino acid sequence set forth in SEQ ID NO: 12 and a second amino acid sequence set forth in SEQ ID NO: 13 and wherein said env moiety does not include any immunogenic domain(s) originating from preS1 and preS2 regions; wherein either: (a) at least 45 and at most 50 amino acid residue truncation of (i) at the N-terminus of a native HBV polymerase protein or (b) at least 34 and at most 37 amino acid residue truncation of (ii) at the C-terminus of a native HBV core protein; wherein said adenoviral vector is capable of inducing a T cell response against at least one of said HBV moieties and said adenoviral vector is E1-defective.

11. The adenoviral vector of claim 10, wherein said wherein said adenoviral vector is for use in a subject infected or suspected to be infected with an HBV from a genotype D.

12. A host cell comprising the adenoviral vector according to claim 10.

13. A composition comprising the adenoviral vector according to claim 10 and a pharmaceutically acceptable vehicle.

14. The composition according to claim 13 which further comprises one or more adjuvant(s) suitable for systemic or mucosal application in humans.

15. The composition according to claim 13 which is formulated for intramuscular or subcutaneous administration.

16. The composition according to claim 13, which comprises from about 10.sup.5 to about 10.sup.13 infection units of an adenoviral vector or of an infectious adenoviral particle.

17. The adenoviral vector according to claim 11, a host cell comprising the adenoviral vector according to claim 11, or the composition comprising the adenoviral vector according to claim 11, wherein said use is used in combination with standard of care.

18. The adenoviral vector according to claim 10, a host cell comprising the adenoviral vector according to claim 10, or a composition comprising the adenoviral vector according to claim 10, wherein said genotype D HBV viruses are from HBV isolate Y07587.

19. A method of inducing an immune response in a subject infected with or suspected to be infected with an HBV, comprising administering the composition of claim 4 to said subject.

20. The method of claim 19, wherein the HBV infection is a chronic HBV infection.

21. A method of inducing an immune response in a subject infected with or suspected to be infected with an HBV, comprising administering the composition of claim 13 to said subject.

22. The method of claim 21, wherein the HBV infection is a chronic HBV infection.

Description

LEGENDS OF FIGURES

(1) FIGS. 1A and 1B illustrates HBV polypeptide expression from adenovirus and MVA infected cells. A549 cells or chicken embryo fibroblasts were infected at MOI 10 or 50 for adenovirus or MOI 0.2 or 1 for MVA and cells were lysed 48h after infection. Western blot were then performed with cell lysates obtained from cells infected with the various Ad (FIG. 1A) and MVA (FIG. 1B) constructs to detect specific HBV proteins. Core-containing polypeptides were detected using an anti-Core antibody (C1-5 or 13A9, dilution 1/200) and polymerase-containing polypeptides with an anti-Pol antibody (8D5, dilution 1/200) as primary antibodies and the secondary antibody was coupled to HRP. Expected sizes for the proteins expressed by Ad TG17909 and Ad TG17910 are respectively 31.6 kDa and 88.5 kDa. Expected sizes for proteins expressed by MVA TG17971, MVA TG17972, MVA TG17993 and MVA TG17994 are respectively 20.2 kDa, 15.8 kDa, 20 kDa and 23.5 kDa. Expected sizes for proteins expressed by MVA TG17842 and MVA TG17843 are 88.5 kDa and 98.2 kDa respectively.

(2) FIGS. 2A-2C illustrates the immunogenicity of HBV polypeptides encoded by adenovirus in Elispots IFN assays. Five individual mice (HLA-A2 transgenic mice) were immunised once with either Ad TG17909 alone (black bars), Ad TG17910 alone (white bars) or in combination (Ad TG17909+Ad TG17910) (grey bars). FIG. 2A illustrates specific T cell responses targeting Polymerase protein using the HLA-A2 restricted peptide SLY (SEQ ID NO: 55) or an irrelevant one (not shown). FIG. 2B illustrates specific T cell responses targeting Core protein using the HLA-A2 restricted peptides FLP (SEQ ID NO: 56) or ILC (SEQ ID NO: 57). FIG. 2C illustrates specific T cell responses targeting Env domains using the HLA-A2 restricted peptides VLQ (SEQ ID NO: 58), FLG (SEQ ID NO: 59) or GLS (SEQ ID NO: 60). Each bar represents an individual vaccinated mouse and the hatched bars represent the median of each group. Results are shown as the mean value of the number of spots observed for 10.sup.6 spleen cells, obtained from triplicate wells. A response was considered positive if the number of spots was higher than 50 spots per 10.sup.6 cells (this cut-off is represented by a thick black line).

(3) FIGS. 3A-3C illustrates the immunogenicity of HBV polypeptides encoded by adenovirus vector in intracellular cytokine staining assays. Five individual mice (HLA-A2 transgenic mice) were immunised once with either AdTG17909 (FIG. 3B), AdTG17910 (FIG. 3A) or a combination of AdTG17909 and AdTG17910 (FIG. 3C). Splenocytes were cultured for 5h with Golgi-Plug and in presence of each HLA-A2 restricted peptide (SLY for Pol, FLP, ILC for Core, VLQ, FLG and GLS for Env) or an irrelevant one. The percentage of CD8+ cells producing cytokines (IFNg and/or TNFa) specific of each HLA-A2 restricted epitopes, was assessed by ICS assays. Each bar represents an individual vaccinated mouse, with IFNg producing cells represented by a black bar, TNFa producing cells by a white bar and IFNg+TNFa producing cells by a hatched bar and all these cell populations are piled for each mouse.

(4) FIGS. 4A-4C illustrates the ability of HBV polypeptides encoded by adenovirus vector to induce CD8 and CD4 T cell responses, detected by intracellular cytokine staining assays.

(5) Five individual mice (HLA-A2 transgenic mice) were immunised once with a mixture of AdTG17909 and AdTG17910. Splenocytes were cultured for 5h with Golgi-Plug and in presence of each HLA-A2 restricted peptide (SLY for Pol, FLP, ILC for Core, VLQ, FLG and GLS for Env) or pools of overlapping peptides (15aa overlapping by 1 lamino acids, 2 pools of peptides for Core and 2 pools of peptides for Env) covering the whole antigenic domains or an irrelevant peptide. Induced specific CD8 T cells producing IFNgamma and/or TNFalpha (FIG. 4A) and induced specific CD4 T cells producing IFNgamma and/or TNFalpha (FIG. 4B) or producing IFNgamma and/or IL2 (FIG. 4C) were monitored by ICS assays. Each bar represents an individual vaccinated mouse, with IFNg producing cells represented by a grey bar, TNFa or IL2 producing cells by a white bar and IFNg+TNFa or IFNg+IL2 producing cells by a hatched bar and all these cell populations are piled for each mouse. The median of each group is also showed.

(6) FIGS. 5A and 5B illustrates the ability of adenovirus vector encoding HBV polypeptides to induce in vivo functional cytolysis against target cells loaded with HBV HLA-A2 restricted epitopes. Three individual mice (HLA-A2 transgenic mice) were immunised once with a combination of AdTG17909 and AdTG17910 (M1 to M3) and one mouse was immunized once with an empty adenovirus vector as negative control (M0). CFSE stained splenocytes from syngenic mice, loaded with HBV HLA-A2 epitopes or not (negative control) were injected intraveinously to vaccinated mice. The in vivo lysis of stained cells was assessed for each mouse 24h later by flow cytometry and calculated as indicated in Material and methods. The mean of specific lysis observed for each peptide for the 3 mice vaccinated with AdHBV was calculated and showed (Mean M1-M3).

(7) FIGS. 6A and 6B illustrates the immunogenicity of HBV polypeptides encoded by MVA vector as determined by Elispots IFNgamma assays. Individual mice (HLA-A2 transgenic mice) were immunised three times at one week interval with either MVATG17842 or MVATG17843 (FIG. 6A) or MVATG17971 (FIG. 6B) or MVA TG17972 or the negative control MVA TGN33.1 (data not shown). FIG. 6A illustrates specific T cell responses targeting Polymerase protein following immunization with MVA TG17842 (dark grey bars) or MVATG17843 (light grey bars) using the HLA-A2 restricted peptide SLY (SEQ ID NO: 55), pool 8 of peptides covering the C-terminal part of the polymerase protein (25 peptides of 15 amino acids overlapping by 11 amino acids/pool), an irrelevant peptide or medium (negative controls). FIG. 6B illustrates specific T cell responses targeting Core protein following immunization with MVATG17971 using the HLA-A2 restricted peptides FLP (SEQ ID NO: 56), ILC (SEQ ID NO: 57), peptides pools core 1 and core 2 (21 to 22 peptides of 15 amino acids overlapping by 11 amino acids/pool), an irrelevant peptide or medium (negative controls). Each bar represents an individual vaccinated mouse and the hatched bars represent the mean of each group. Results are shown as the mean value of the number of spots observed for 10.sup.6 spleen cells, obtained from triplicate wells. A response was considered positive if the number of spots was higher than 50 spots per 10.sup.6 cells (this cut-off is represented by a dotted black line).

(8) FIGS. 7A-7C illustrates the immunogenicity of HBV polypeptides encoded by MVA vectors co-injected in mouse, as determined by Elispots IFNgamma assays. Individual mice (HLA-A2 transgenic mice) were immunised three times at one week interval with a mix of MVATG17843 and either MVATG17972 (FIG. 7A) or MVATG17993 (FIG. 7B) or MVATG17994 (FIG. 7C) or with MVA TG N33.1 alone as negative control (data not shown). Specific T cell responses targeting Polymerase protein were determined using the HLA-A2 restricted peptide SLY (SEQ ID NO: 55) and specific T cell responses targeting Env domains using the HLA-A2 restricted peptide GLS (SEQ ID NO: 60) or a pool of peptides covering the Env2 domain (pool of 15 amino acid-long peptides overlapping by 11 amino acids). An irrelevant peptide and medium were used as negative controls. Each bar represents an individual vaccinated mouse and the hatched bars represent the mean of each group. Results are shown as the mean value of the number of spots observed for 10.sup.6 spleen cells, obtained from triplicate wells. A response was considered positive if the number of spots was higher than 92 spots per 10.sup.6 cells (this cut-off is represented by a dotted black line).

(9) FIGS. 8A and 8B illustrates the cross-reactive potential of T-cells induced by HBV polypeptides encoded by adenoviruses and specific of HBV Core antigen in HLA-A2 transgenic mouse model.

(10) HLA-A2 mice were immunized by subcutaneous route with 10.sup.8 iu of AdTG17909 and 10.sup.8 iu of AdTG17910. Splenocytes were taken 2 weeks after immunisation and ELISpot IFNg (FIG. 8A) and ICS (FIG. 8B) were performed following stimulation with peptides which amino acid sequence is homologous to the HBV sequence encoded by the adenovirus (FLP and ILC) and the major and minor variants identified in Table 1. Open symbols and thin lines represent the values obtained for individual animals whereas filled symbols and bold lines correspond to group means. Horizontal lines represent the cut-off of positivity. Main variants of the different genotypes are framed by a pointed line. Responder frequency (percentage of animals above the cut-off) and normalised mean (group mean for a given epitope variant divided by the group mean for the homologous peptide) are indicated below the graphs. ILC variants are ranked in decreasing order of their Elispot normalised mean. Statistical difference in the level of observed T cell responses between the homologous peptide and its respective variants is indicated by a star. As a log-scale was used to represent the results, Elispot values equal to zero were represented as 1 on the graph and ICS values under 0.03% were represented as equal to 0.03% (but real values were used to calculate means).

(11) FIGS. 9A and 9B illustrates the cross-reactive potential of T-cells specific of HBV Env domains in HLA-A2 transgenic mouse model.

(12) HLA-A2 mice were immunized by subcutaneous route with 10.sup.8 iu of AdTG17909 and 10.sup.8 iu of AdTG17910. Splenocytes were taken 2 weeks after immunisation and ELISpot IFNg (FIG. 9A) and ICS (FIG. 9B) were performed following stimulation with peptides which amino acid sequence is homologous to the HBV sequence encoded by the adenovirus (VLQ, FLG and GLS) and the major and minor variants identified in Table 1. Open symbols and thin lines represent the values obtained for individual animals whereas filled symbols and bold lines correspond to group means. Horizontal lines represent the cut-off of positivity. Main variants of the different genotypes are framed by a pointed line. Responder frequency (percentage of animals above the cut-off) and normalised mean (group mean for a given epitope variant divided by the group mean for the homologous peptide) are indicated below the graphs. Epitope variants are ranked in decreasing order of their normalised mean. Statistical difference in the level of observed T cell responses between the homologous peptide and its respective variants is indicated by a star. As a log-scale was used to represent the results, Elispot values equal to zero were represented as 1 on the graph and ICS values under 0.03% were represented as equal to 0.03% (but real values were used to calculate means).

EXAMPLES

1. Material and Methods

(13) The constructions described below are carried out according to the general genetic engineered and molecular cloning techniques detailed in Maniatis et al. (1989, Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor N.Y.) or according to the manufacturer's recommendations when a commercial kit is used. PCR amplification techniques are known to the person skilled in the art (see for example PCR protocolsA guide to methods and applications, 1990, published by Innis, Gelfand, Sninsky andWhite, Academic Press). The recombinant plasmids carrying the ampicillin resistance gene are replicated in the E. coli C600 (Stratagene) on agar or liquid medium supplemented with 100 g/ml of antibiotic. The constructions of the recombinant vaccinia viruses are performed according to the conventional technology in the field in the documents above cited and in Mackett et al. (1982, Proc. Natl. Acad. Sci. USA 79, 7415-7419) and Mackett et al. (1984, J. Virol. 49, 857-864). The selection gene gpt (xanthine guanine phosphoribosyltransferase) of E. coli (Falkner and Moss, 1988, J. Virol. 62, 1849-1854) is used to facilitate the selection of the recombinant vaccinia viruses.

(14) 1.1. Vectors Constructions and Production

(15) 1.1.1. Selected Antigens and HBV Sequence Strain

(16) The vectors exemplified hereinafter have been engineered to express the polymerase, core polypeptides and immunogenic domains of the envelope protein. They all originate from HBV strain Y07587 which sequence is described in international databases (Genbank Y07587) and in different publications. It is a genotype D virus of serotype ayw.

(17) The Core polypeptide is either wild-type (aa 1-183) or a Core polypeptide deleted of amino acids 77 to 84 (i.e. Core containing amino acid 1 to 76 and 85 to 183 designated core*) or a C-terminally truncated polypeptide (1-148) or a C-terminally truncated core (1-148) further deleted of amino acids 77 to 84 (i.e. Core containing amino acid 1 to 76 and 85 to 148 designated core*t).

(18) The polymerase polypeptide is either wild type or a N-terminally truncated polypeptide lacking the first 47 amino acids (48-832) or a N-truncated polymerase (48-832) further mutated at posititon 540 (D in H) and 718 (E in H) (positions 450 and 718 being given with respect to the wild-type polymerase) or the truncated (48-832) and mutated polymerase (D540H and E718H) which is fused to the peptide signal and transmembrane domain of the rabies virus glycoprotein (Pol*TMR).

(19) The selected Env domains are: domain from amino acids 14 to 51 of the S protein (Env 1) and domain from amino acids 165 to 194 of the HBs protein (Env 2) and domain from amino acid 202 to 226 of the HBs protein (Env 4).

(20) 1.1.2. Construction and Production of a MVATG17842 Expressing a Truncated and Mutated HBV Polymerase (Pol*)

(21) The nucleotide sequences encoding a modified HBV polymerase polypeptide were synthesized by Geneart company using synthetic oligonucleotides and PCR products. The modified HBV polymerase corresponds to the polymerase protein of HBV Y07587 (SEQ ID NO:1) mutated at position 540 (D in H) and 718 (E in H) in order to eliminate Rnase H and RTase activities exhibited by the native HBV polymerase (resulting in amino acid sequence shown in SEQ ID NO: 8 and nucleotide sequence shown in SEQ ID NO: 22). The reassembled Pol sequence was then cloned in a plasmid vector resulting in PGA15-pol (SEQ ID NO: 27). A truncated version deleted of the first 47 amino acids present at the N-terminus of the native HBV polymerase was amplified by PCR from pGA15-Pol plasmid using the following primers OTG19037 (GAGCGATATCCACCATGAATGTTAGTATTCCTTGGAC) (SEQ ID NO: 28) and OTG19038 (GATCGCTAGCTCACGGTGGTCTCCATGCGAC) (SEQ ID NO: 29). The resulting fragment was inserted into the NheI and EcoRV restriction sites of a MVA transfer plasmid downstream the p7.5K promoter (Cochran et al, 1985, J. Virol. 54:30), resulting in pTG17842. The mutated and truncated polymerase is designated hereinafter pol*.

(22) The MVA transfer plasmid is designed to permit insertion of the nucleotide sequence to be transferred by homologous recombination in deletion III of the MVA genome. It originates from plasmid pTG1E (described in Braun et al., 2000, Gene Ther. 7:1447) into which were cloned the flanking sequences (BRG3 and BRD3) surrounding the MVA deletion III (Sutter and Moss, 1992, Proc. Natl. Acad. Sci. USA 89:10847). The transfer plasmid also contains a fusion between the Aequorea victoria enhanced Green Fluorescent protein (eGFP gene, isolated from pEGP-C1, Clontech) and the Escherichia coli xanthine-guanine phosphoribosyltransferase gene (gpt gene) under the control of the early late vaccinia virus synthetic promoter p 11K7.5 (kindly provided by R. Wittek, University of Lausanne). Synthesis of xanthine-guanine phosphoribosyltransferase enables GPT.sup.+ recombinant MVA to form plaques in a selective medium containing mycophenolic acid, xanthine, and hypoxanthine (Falkner et al, 1988, J. Virol. 62, 1849-54) and eGFP enables the visualisation of recombinant MVA plaques. The selection marker eGFP-GPTis placed between two homologous sequences in the same orientation. When the clonal selection is achieved, the selection marker is easily eliminated by several passages without selection allowing the growth of eGFP-GPT.sup. recombinant MVA.

(23) Generation of MVATG17842 virus was performed by homologous recombination in primary chicken embryos fibroblasts (CEF) infected with MVA and transfected with pTG17842 (according to the standard calcium phosphate DNA precipitation). Viral selection was performed by three round of plaque purification in the presence of a selective medium containing mycophenolic acid, xanthine and hypoxanthine. As mentioned above, the selection marker was then eliminated by passage in a non-selective medium. Absence of contamination by parental MVA was verified by PCR.

(24) Analysis of expression of HBV polymerase was performed by Western-blot. A549 cells (ATCC CCL-185) were infected at MOI of 1 with MVATG17842 (Pol*) in presence or in absence of proteasome inhibitor MG-132 (10 M) added to growth medium. After 24 hours, cells were harvested. Western-blot analysis was performed using commercial monoclonal anti-Pol antibody Hep B Pol (8D5, Santa Cruz, #sc-81591).

(25) 1.1.3. Construction and Production of MVATG17843 Expressing a Truncated and Mutated HBV Polymerase Fused to the Membrane-Anchoring Domain of Rabies Envelop Glycoprotein (Pol*TMR)

(26) The HBV Pol* sequence was then modified by fusion at its N-terminus to a peptide signal (SS) and at its C-terminus to a membrane-anchoring sequences (TMR) derived from the glycoprotein of the rabies virus (ERA isolate; described in Genbank No M38452). The SS and TMR sequences were amplified from plasmid pTG8042 (described in WO99/03885) by PCR using respectively primer pairs OTG19045 (SEQ ID NO: 30) (GAGTGATATCCACCATGGTTCCTCAGGCTCTCCTG)/OTG19047 (SEQ ID NO: 31) (GTCCAAGGAATACTAACATTAATAGGGAATTTCCCAAAACACAATG) and OTG19049 (SEQ ID NO: 32) (GTCGCATGGAGACCACCGTATGTATTACTGAGTGCAGGG/OTG19050 (SEQ ID NO: 33) (GAGTGCTAGCTCACAGTCTGGTCTCACCC). Pol* sequence was amplified from plasmid pGA15-Pol by PCR using primer pair OTG19046 (SEQ ID NO: 34) (GTTTTGGGAAATTCCCTATTAATGTTAGTATTCCTTGGACTC)/OTG19048 (SEQ ID NO: 35) (CTGCACTCAGTAATACATACGGTGGTCTCCATGCGACGTGC). Then, SS-Pol*-TMR sequence was reassembled by triple PCR using the following primers OTG19045 (SEQ ID NO: 30) and OTG19050 (SEQ ID NO: 33). The resulting fragment was inserted into the NheI and EcoRV restriction sites of a vaccinia transfer plasmid downstream the p7.5K promoter (Cochran et al, 1985, J. Virol. 54:30), resulting in pTG17843.

(27) Generation of MVATG17843 virus was performed in CEF by homologous recombination as described above.

(28) Analysis of Pol*-TMR expression was performed by Western-blot. A549 cells were infected at MOI 1 with MVATG17843 in presence or in absence of proteasome inhibitor MG-132 (10 M) added to growth medium. After 24 hours, cells were harvested. Western-blot analysis was performed using commercial monoclonal anti-Pol antibody Hep B Pol (8D5, Santa Cruz, #sc-81591).

(29) 1.1.4 Construction and Production of MVATG17971 Expressing a Deleted Core Polypeptide (Core*)

(30) Core* corresponds to the Core sequence of HBV Y07587 (SEQ ID NO: 2) deleted of amino acids 77 to 84.

(31) The Core* encoding sequences were reconstituted by double PCR from pGA4-Core plasmid. This plasmid was made by Geneart company. It contains a full length coding sequence of modified HBV Core gene which was assembled from synthetic oligonucleotides and/or PCR products. The last two codons CAA TGT of the coding sequence were modified in CAG TGC to avoid sequence homology with Pol (SEQ ID NO: 36).

(32) Core sequence from positions 1 to 76 was amplified by PCR using the following primers OTG19290 (SEQ ID NO: 37) (GACTGTTAACCACCATGGACATTGATCCTTA-TAAAGAATTTG) and OTG19292 (SEQ ID NO: 38) (GTTGACATAACTGACTA-CCAAATTACCACCCACCCAGGTAG). Core sequence from positions 85 to 183 was amplified by PCR with the following primers OTG19291 (SEQ ID NO: 39) (GTGGGTGGTAATTTGGTAGTCAGTTATGTCAACACTAATATG) and OTG19080 (SEQ ID NO: 61) (GACTCTCGAGTTAGCACTGAGATTCCCGAGATTG). A double PCR was performed using OTG19290 (SEQ ID NO: 37) and OTG19080 (SEQ ID NO: 61) and both latter generated amplicons. The resulting fragment was inserted into the XhoI and HpaI restriction sites of a vaccinia transfer plasmid downstream the pH5R promoter (Rosel et al, 1986, J Virol. 60:436), resulting in pTG17971.

(33) Generation of MVATG17971 virus was performed in CEF by homologous recombination as described above.

(34) Analysis of Core* expression was performed by Western-blot. Chicken embryo fibroblasts were infected at MOI 0.2 with MVATG17971. After 24 hours, cells were harvested. Western-blot analysis was performed using a commercial monoclonal anti-core antibody Hep B cAg (13A9) (Santa Cruz, #sc-23946).

(35) 1.1.5 Construction and Production of MVATG17972 Expressing a Deleted and Truncated Core Polypeptide (Core*t)

(36) Core*t corresponds to the Core sequence of HBV Y07587 (SEQ ID NO: 2) truncated after amino acid 148 and deleted of amino acids 77 to 84.

(37) The Core*t-encoding sequences were reconstituted by double PCR from pGA4-Core plasmid which contains the sequence encoding the full length HBV Core gene which was assembled from synthetic oligonucleotides and PCR products except that the last two codons CAA TGT of the coding sequence were modified in CAG TGC to avoid sequence homology with Pol (SEQ ID NO: 36).

(38) Core sequence from positions 1 to 76 was amplified by PCR using the following primers OTG19290 (SEQ ID NO: 37) (GACTGTTAACCACCATGGACATTGATCCTTA-TAAAGAATTTG) and OTG19292 (SEQ ID NO: 38) (GTTGACATAACTGACTA-CCAAATTACCACCCACCCAGGTAG). Core sequence from positions 85 to 148 was amplified by PCR from pGA4-core with the following primers OTG19291 (SEQ ID NO: 39) (GTGGGTGGTAATTTGGTAGTCAGTTATGTCAACACTAATATG) and OTG19299 (SEQ ID NO: 40) (GACTCTCGAGTTAAACAGTAGTCTCCGGAAGTG). The double PCR was performed using OTG19290 (SEQ ID NO: 37) and OTG19299 (SEQ ID NO: 40). The resulting fragment was inserted into the XhoI and HpaI restriction sites of a vaccinia transfer plasmid downstream the pH5R promoter (Rosel et al, 1986, J Virol. 60:436), resulting in pTG17972.

(39) Generation of MVATG17972 virus was performed in CEF by homologous recombination as described above.

(40) Analysis of Core*t expression was performed by Western-blot. Chicken embryo fibroblasts were infected at MOI 0.2 with MVATG17972. After 24 hours, cells were harvested. Western-blot analysis was performed using a commercial monoclonal anti-core antibody Hep B cAg (13A9) (Santa Cruz, #sc-23946).

(41) 1.1.6 Construction and Production of MVATG17993 Expressing a Deleted and Truncated Core Polypeptide Fused to the Env1 Immunogenic Domain (Core*t-Env1)

(42) The Core-t* moiety was fused to Env1 domain extending from amino acids 14 to 51 of the HBs protein.

(43) The Core*t-Env1 sequence was reconstituted by double PCR. Core*t sequence was amplified by PCR from pTG17972 using the following primers OTG19317 (SEQ ID NO: 41) (GACGGGATCCACCATGGACATTGATCCTTATAAAGAATTTGG) and OTG19319 (SEQ ID NO: 42) (GCCTGCTTGCAGGACAACAGTAGTCTCCGGAAGTGTTG). Env1 sequence was amplified by PCR from plasmid pMK-C/E (SEQ ID NO: 43) using the following primers OTG19318 (SEQ ID NO: 44) (CCGGAGACTACTGTTGTCC-TGCAAGCAGGCTTCTTC) and OTG19320 (SEQ ID NO: 45) (GAGTCATTCTCGAC-TTGCGGCCGCTTACTGACCCAGGCAAACCGTGG). The double PCR was performed using OTG19317 (SEQ ID NO: 41) and OTG19320 (SEQ ID NO: 45). The resulting fragment was inserted into the BamHI and NotI restriction sites of a vaccinia transfer plasmid downstream the pH5R promoter (Rosel et al, 1986, J Virol. 60:436), resulting in pTG17993.

(44) For illustrative purposes, the plasmid pMK-C/E was made by Geneart and contains a chimeric sequence consisting of an insertion of three HBV env domain sequences into core sequence (SEQ ID NO: 43). The native core and env nucleotide sequences were degenerated to avoid sequence homology with HBV Pol sequence and also sequence instability due to polyT or polyGC stretches. In addition, the core sequence was deleted of amino acids 77 to 84 and truncated at aa148. The selected Env domains are: domain from amino acids 14 to 51 of the S protein (Env 1) and domain from amino acids 165 to 194 of the S protein (Env 2) and domain from amino acid 202 to 226 of the S protein (Env 4). The three domains were inserted respectively at positions nt 127, at nt 222 and at nt 416 of core sequence. It has to be noted that insertion of this sequence in a MVA vector results in cytotoxicity in the expressing cells, emphasizing the fact that the design of the env-core fusion is not straightforward.

(45) Generation of MVATG17993 virus was performed in CEF by homologous recombination as described above.

(46) Analysis of Core*t-env1 expression was performed by Western-blot. Chicken embryo fibroblasts were infected at MOI 0.2 with MVATG17993. After 24 hours, cells were harvested. Western-blot analysis was performed using a commercial monoclonal anti-core antibody Hep B cAg (13A9)(Santa Cruz, #sc-23946).

(47) 1.1.7. Construction and Production of MVATG17994 Expressing a Deleted and Truncated Core Polypeptide Fused to Env1 and Env2 Immunogenic Domains (Core*t-Env1-Env2)

(48) The Core*t polypeptide described in 1.1.5 was then fused to two immunogenic domains extending from amino acids 14 to 51 (Env 1) and from amino acid 165 to 194 (Env2) of the HBs protein

(49) The nucleotide sequences encoding the Core*t-Env1-Env2 were reassembled by triple PCR. Core*t sequence was amplified by PCR from pTG17972 using the following primers OTG19317 (SEQ ID NO: 41) and OTG19319 (SEQ ID NO: 42). Env1 was amplified from pMK-C/E plasmid using the following primers OTG19318 (SEQ ID NO: 44) and OTG19322 (SEQ ID NO: 46 (GCGTGCGCTTGCCCACTGACCCAGGCAAACCGTGG). Env2 was amplified from pMK-C/E plasmid using the following primers OTG19321 (SEQ ID NO: 47 (CGGTTTGCCTGGGTCAGTGGGCAAGCGCACGCTTTAGC) and OTG19323 (SEQ ID NO: 48) (GAGTCATTCTCGACTTGCGGCCGCTTACACGCTCAGCCACACGGTTGG). The triple PCR was performed using OTG19317 (SEQ ID NO: 41) and OTG19323 (SEQ ID NO: 48). The resulting fragment was inserted into the BamHI and NotI restriction sites of a vaccinia transfer plasmid downstream the pH5R promoter (Rosel et al, 1986, J Virol. 60:436), resulting in pTG17994.

(50) Generation of MVATG17994 virus was performed in CEF by homologous recombination as described above.

(51) Analysis of Core*t-env1-env2 expression was performed by Western-blot. Chicken embryo fibroblasts were infected at MOI 0.2 with MVATG17994. After 24 hours, cells were harvested. Western-blot analysis was performed using a commercial monoclonal anti-core antibody Hep B cAg (13A9, Santa Cruz, #sc-23946).

(52) 1.1.8. Construction and Production of an Adenoviral Vector AdTG17909 Expressing CORE-Env1-Env2-Env4:

(53) A synthetic gene (831 nucleotides) encoding a CORE-Env1-Env2-Env4 fusion was reconstituted by double PCR. CORE was amplified by PCR from pGA4-Core (described in 1.1.4.) using the following primers OTG19152 (SEQ ID NO: 49) (GGGGGGCTAGCAAGCTTCCACCATGGACATTGATCCTTATAAAGAATTTG) and OTG19154 (SEQ ID NO: 50) (GAAAGAATCCAGCTTGCAGGACGCACT-GAGATTCCCGAGATTGAG). Env1-Env2-Env4 were amplified by PCR from pGA4-Env using the following primers OTG19153 (SEQ ID NO: 51) (CTCAATCTCGGGAATCT-CAGTGCGTCCTGCAAGCTGGATTCTTTC) and OTG19159 (SEQ ID NO: 52) (GAGTCATTCTCGACTTGCGGCCGCTTAGATATAAACCCACAAGC). The double PCR was performed using OTG19152 (SEQ ID NO: 49) and OTG19159 (SEQ ID NO: 52). The resulting fragment was inserted into the NheI and NotI restriction sites of an adenoviral shuttle plasmid containing a CMV-driven expression cassette surrounded by adenoviral sequences (adenoviral nucleotides 1-454 and nucleotides 3513-5781 respectively) to allow generation of the vector genome by homologous recombination (Chartier et al., 1996, J. Virol. 70:4805). The resulting adenoviral vector pTG17909 is E3 (nucleotides 28593-30464) and E1 (nucleotides 455-3512) deleted, with the E1 region replaced by the expression cassette containing, from 5 to 3, the CMV immediate-early enhancer/promoter, a chimeric human -globin/IgG intron (as found in pCI vector available in Promega), the sequence encoding the CORE-Env1-Env2-Env4 and the SV40 late polyadenylation signal. The recombinant adenovirus was generated by transfecting the Pac linearized viral genomes into an E1 complementation cell line. Virus propagation, purification and titration were made as described previously [Erbs, 2000] (Erbs et al., 2000, Cancer Res. 60:3813) Expression of the fusion protein was evaluated by Western-blot. 10.sup.6 A549 cells (ATCC CCL-185) were infected at MOI of 10 or 50 for 48 hours with AdTG17909 or with an empty adenovirus as negative control. The cell pellets were collected and probed with an anti CORE mouse monoclonal antibody (C1-5, sc-23945, Santa Cruz).

(54) 1.1.9. Construction and Production of Adenoviral Vector AdTG17910 Expressing Pol*:

(55) The gene encoding Pol*, a polymerase protein truncated of the 47 first amino acids except the Met initiator (48 to 832) and mutated at posititon 540 (D in H) and 718 (E in H) (with respect to the wild type polymerase) was inserted in an adenovirus vector. The Pol gene (2364 nucleotides) was amplified by PCR from pGA15-Pol (described in 1.1.2) using primers OTG19155 (SEQ ID NO: 53) (GGGGGGCTAGCAAGCTTCCACCATGAA-TGTTAGTATTCCTTGGACTCATAAG) and OTG19156 (SEQ ID NO: 54) (GAGTCATTCTCGACTTGCGGCCGCTCACGGTGGTCTCCATGCGACGTGC). The resulting fragment was inserted into the NheI and NotI restriction sites of an adenoviral shuttle plasmid containing a CMV-driven expression cassette surrounded by adenoviral sequences (adenoviral nucleotides 1-454 and nucleotides 3513-5781 respectively) to allow generation of the vector genome by homologous recombination (Chartier et al., 1996, J. Virol. 70:4805). The resulting adenoviral vector pTG17910 is E3 (nucleotides 28593-30464) and E1 (nucleotides 455-3512) deleted, with the E1 region replaced by the expression cassette containing, from 5 to 3, the CMV immediate-early enhancer/promoter, a chimeric human -globin/IgG intron (as found in pCI vector available in Promega), the sequence encoding the truncated and mutated Pol and the SV40 late polyadenylation signal. The recombinant adenovirus was generated by transfecting the PacI linearized viral genomes into an E1 complementation cell line. Virus propagation, purification and titration were made as described previously (Erbs et al., 2000, Cancer Res. 60:3813).

(56) Expression of the fusion protein was evaluated in adenovirus infected cells by Western-blot. A549 cells (10.sup.6 cells) (ATCC CCL-185) were infected at MOI of 10 or 50 for 48 hours with the adenovirus AdTG17910, as well as an empty adenovirus as negative control. The cell pellets were collected and probed with an anti Pol mouse monoclonal antibody (8D5, sc-81591, Santa Cruz).

(57) 1.2. Evaluation of Antigen Immunogenicity

(58) Antigen immunogenicity was evaluated in vivo by Elispot IFN and intracellular cytokine staining (ICS) assays following immunization of HLA transgenic mice.

(59) 1.2.1 Mouse Model

(60) The HLA-A2.1 transgenic mice used in the study were described by Pascolo et al. (1997, J. Exp. Med. 185:2043). These mice have the H-2D.sup.b and murine .sub.2-microglobulin genes knocked-out and express a transgenic monochain histocompatibility class I molecule (HHD molecule) in which the C-terminus of the human 2m is covalently linked to the N-terminus of a chimeric heavy chain (HLA-A*0201 1-2, H-2D.sup.b 3 transmembrane and intracytoplasmic domains). Seven to 10 weeks-old mice (male and female) were immunized. Average weight of the mice was around 25-30 g.

(61) 1.2.2. Immunization Protocols

(62) Mice were divided in 4 groups; group 1 immunized by AdTG17909 (encoding HBV Core fused to env1, env2 and env4 immunogenic domains), group 2 immunized AdTG17910 (encoding the truncated and mutated Pol*), group 3 immunized with both vectors and group 4 immunized with an empty adenovirus (AdTG15149) as negative control. All animals were immunized by subcutaneous injection at the base of the tail, groups 1 and 2 animals received one subcutaneous injection of 10.sup.8 IU of each Adenovirus (TG17909 or TG17910), group 3 one subcutaneous injection of a mix containing 10.sup.8 IU of each adenovirus (total of 2.Math.10.sup.8 IU: 10.sup.8 IU of AdTG17909+10.sup.8 IU of AdTG17910) and negative controls received one subcutaneous injection of 2.Math.10.sup.8 IU of AdTG15149. Cellular immune responses were assessed by IFNg Elispot and intracellular cytokine staining (ICS) assays 2 weeks after the immunization.

(63) 1.2.3 Peptides

(64) Peptides used for cells stimulation in vitro were either short peptides of 9 to 10 amino acids which are described or predicted as HLA-A2 restricted epitopes or long peptides of 15 amino acids included in peptide libraries covering all the antigens of interest.

(65) Short peptides corresponding to described or predicted HLA-A2 restricted epitopes of Polymerase protein, Core protein or Env domains were synthesized by Eurogentec (Belgium) and were dissolved in 100% DMSO (sigma, D2650) at a concentration of 10 mM.

(66) Peptides libraries covering the whole Polymerase, Core and Envelope proteins were synthesized by PROIMMUNE (Oxford, United Kingdom). The Pol, Core and Env libraries were composed of 15 mer peptides overlapping by 11 amino acids. Each crude peptide was dissolved in 100% DMSO (sigma, D2650) at a concentration of 50 mg/ml. For each library, peptides were pooled to a concentration of 2 mg/ml per peptide: HBV Pol protein is covered by 8 pools of 24 to 25 peptides from Pol library (Pool 1: 24 peptides covering residues 45 to 151; Pool 2: 24 peptides covering residues 140 to 251; Pool 3: 24 peptides covering residues 241 to 347; Pool 4: 24 peptides covering residues 337 to 447; Pool 5: 24 peptides covering residues 437 to 543; Pool 6: 24 peptides covering residues 533 to 639; Pool 7: 24 peptides covering residues 629 to 735; Pool 8: 25 peptides covering residues 725 to 832); HBV Core protein is covered by 2 pools of 21-22 peptides from Core library (Pool 1: 22 peptides covering residues 1 to 100; Pool 2: 21 peptides covering residues 89 to 183); HBV Env protein is covered by 3 pools of 6 to 10 peptides from Env library (Pool 1: 10 peptides covering HBs residues 9 to 59; Pool 2: 9 peptides covering HBs residues 157 to 194; Pool 4: 6 peptides covering HBs residues 193 to 226).
1.2.4. IFNg Elispot Assays

(67) Splenocytes from immunized mice were collected and red blood cells were lysed (Sigma, R7757). 2.Math.10.sup.5 cells per well were cultured in triplicate for 40 h in Multiscreen plates (Millipore, MSHA S4510) coated with an anti-mouse IFNg monoclonal antibody (BD Biosciences; 10 g/ml, 551216) in (MEM culture medium (Gibco, 22571) supplemented with 10% FCS (Sigma, F7524 or JRH, 12003-100M), 80 U/mL penicillin/80 g/mL streptomycin (PAN, P06-07-100), 2 mM L-glutamine (Gibco, 25030), 1 non-essential amino acids (Gibco, 11140), 10 mM Hepes (Gibco, 15630), 1 mM sodium pyruvate (Gibco, 31350) and 50 M -mercaptoethanol (Gibco, 31350) and in presence of 10 units/ml of recombinant murine IL2 (Peprotech, 212-12), alone as negative control, or with: 10 M of one HLA-A2 restricted peptide present in HBV antigens encoded by Ad vectors (SLY in Pol, FLP, ILC for Core, VLQ, FLG and GLS for Env) or an irrelevant one; a pool of peptides at a final concentration of 5 g/ml per peptide 5 g/ml of Concanavalin A (Sigma, C5275) for positive control.

(68) IFNg-producing T cells were quantified by Elispot (cytokine-specific enzyme linked immunospot) assay as previously described (Himoudi et al., 2002, J. Virol. 76:12735). The number of spots (corresponding to the IFNg-producing T cells) in negative control wells was substracted from the number of spots detected in experimental wells containing HBV peptides. Results are shown as the mean value obtained for triplicate wells. An experimental threshold of positivity for observed responses (or cut-off) is determined by calculating a threshold value which corresponds to the mean value of spots observed with medium alone+2 standard deviations, reported to 10.sup.6 cells. A technical cut-off linked to the CTL Elispot reader was also defined as being 50 spots/10.sup.6 cells (which is the value above which the CV of the reader was systematically less than 20%). The highest value between the technical cut-off and the experimental threshold calculated for each experiment is taken into account to define the cut-off value of each experiment. Statistical analyses of Elispot responses were conducted by using a Mann-Whitney test. P value equal or inferior to 0.05 was considered as significant.

(69) 1.2.5. Intracellular Cytokine Staining (ICS) Assays

(70) ICS was performed on splenocytes from each animal of each group. Following red blood cells lysis with lysis buffer (Sigma, R7757), 210.sup.6 cells per well in flat-bottom 96-well plate were incubated in complete alpha (MEM culture medium (Gibco BRL, 22571) in presence of 10 units/ml of murine recombinant IL-2 (Peprotech, 212-12) alone as negative control or with 1 M of specific HBV peptide or with a pool of peptides at a final concentration of 5 g/ml per peptide or with 1 M of an irrelevant peptide. The GolgiPlug (BD Biosciences, 555029) was immediately added at a 1 l/ml final concentration for 5 h. Then, cells were harvested in V-bottom 96-well plates and washed with 1% FCS-PBS. Staining was performed using monoclonal antibodies against CD3 (hamster MAb anti-CD3e-PE, dilution 1/200), CD8 (rat MAb anti CD8a-APC, dilution 1/600) and CD4 (rat MAb anti-CD4-PerCP, dilution 1/600) (all from BD Biosciences, 553063, 553035 and 553052 respectively) in 50 l of 1% FCS-PBS for 15 min at room temperature. After washing, cells were fixed and permeabilized with Cytofix/Cytoperm and washed with Perm/Wash solution (BD Biosciences, 554714). Then, the anti-mouse IFNg-PE antibodies (BD Biosciences, 554412557724) and anti-mouse TNFa-Alexa488 antibodies (BD Biosciences, 557719) or the anti-mouse IFNg-PE antibodies (BD Biosciences, 554412557724) and anti-mouse IL2-Alexa488 antibodies (BD Biosciences, 557719) were added for 15 min at room temperature and after washing with Perm/Wash, cells were resuspended in 1% FCS-PBS and analysed by flow cytometry using a FacsCalibur (Becton Dickinson). CD3e+, CD8a+ cells or CD3e+, CD4+ cells were gated to determine percentages of IFNg+CD8+ or IFNg+CD4+ T or TNFa+CD8+ or TNFa+CD4+ T or IL2+CD8+ or IL2 CD4+ T or IFNg+ TNFa+CD8+ or IFNg+ TNFa+CD4+ or IFNg+ IL2+CD8+ or IFNg+ IL2+CD4+ T cell population. The percentage obtained in medium only was considered as background.

(71) 1.2.6. In Vivo CTL Assays

(72) In vivo CTL assays were performed as described (Fournillier et al., 2007). Splenocyte suspensions were obtained from syngenic mouse spleens and adjusted to 2010.sup.6 cells/ml after lysis of red blood cells. One third of the cells was incubated with one of the HBV specific peptide, the second third of the cells was incubated with another HBV peptide, all at 10 M final concentration for 1 hour at 37 C., whereas the last fraction was left unpulsed. 5(6)-carboxyfluorescein diacetate succinimidyl ester (CFSE) (Molecular probes, C1157) was added at 16 M (CFSE-high) to unpulsed cells, at 4 M (CFSE-medium) to ILC or VLQ peptide pulsed cells and at 1 M (CFSE-low) to SLY or FLP peptide pulsed cells, for 10 min. After washing with PBS, the three populations (unpulsed, ILC and SLY peptide pulsed cells or unpulsed, FLP and ILC peptide pulsed cells) were mixed and 30.10.sup.6 total cells were injected to anaesthetized mice via the retro-orbital vein (using ketamine-xylazine-PBS mix (Ketamine Virbac, Centravet KET204, final concentration 25 mg/ml; Xylazine hydrochloride Rompun Bayer, Centravet, final concentration 5 mg/ml)). Thus, CFSE-low and medium population represented specific targets supposed to be lysed by cytotoxic T cells and CFSE-high population was an internal reference allowing assay normalisation. Splenocytes from recipient mice were analyzed 24 h later by flow cytometry to detect the CFSE-labeled cells. Following gating on lymphocytes (SSC/FSC), a second gating was performed based on the number of events/CFSE fluorescence (FL1) which reveals 3 peaks, a 1rst one corresponding to CFSE-low cells, the 2nd one to CFSE-medium cells and the 3rd one to CFSE-high cells. For each animal, ratio between CFSE+ peptide-pulsed targets and CFSE+ unpulsed targets was calculated (R=Number CFSE-low cells/Number CFSE-high cells). Two naive mice were used to determine R reference. The percentage of specific lysis for each animal was determined by the following formula: % lysis=(1R.sub.mouse/R.sub.reference)100. A response was considered positive if the percentage of specific lysis was higher than 10% (cut-off).

2. Results

(73) 2.1 Antigen Expression by Viral Vectors

(74) 2.1.1 Expression of Antigens from Adenovirus Constructs AdTG17909 and AdTG17910

(75) Expression of the core-env1-env2-env4 fusion protein was evaluated by Western-blot. A549 cells (10.sup.6 cells) were infected at MOI of 10 or 50 for 48 hours with AdTG17909 or an empty adenovirus as negative control. The cell pellets were collected and probed with an anti-Core mouse monoclonal antibody (C1-5, sc-23945, Santa Cruz). As illustrated in FIG. 1A, a major band having the expected molecular weight (31.6 kDa) was revealed in the sample collected from cells infected with AdTG17909.

(76) Expression of the Pol* polypeptide was evaluated by Western-blot following AdTG17910 infection of A549 cells. The cell pellets were then collected and probed with an anti-Pol mouse monoclonal antibody (8D5, sc-81591, Santa Cruz). As illustrated in FIG. 1A, a band having the expected molecular weight (88.5 kDa) was revealed in the sample collected from cells infected with AdTG17910 together with some sub-products (partial polymerase proteins).

(77) 2.1.2 Expression of Antigens from MVA Constructs

(78) Analysis of Pol*, Pol*TMR, Core*t, Core*tEnv1, Core*t-Env1-Env2 expression was performed by Western-blot. A549 cells or CEF were infected at MOI of 1 or 0.2 respectively with MVATG17842, MVATG17843, MVATG17971, MVATG17972, MVATG17993, and MVATG17994 respectively in presence or in absence of proteasome inhibitor MG-132 (10 M) added to growth medium for MVATG17842 and MVATG17843. After 24 hours, cells were harvested.

(79) For MVATG17842, Western-blot analysis was performed using commercial monoclonal anti-Pol antibody Hep B Pol (8D5, Santa Cruz, #sc-81591). As shown in FIG. 1B, expression of a protein with an apparent molecular weight of 88.5 kDa was detected only in presence of MG-132. This band has the expected molecular weight for the Pol* protein.

(80) For MVATG17843, Western-blot analysis was performed using commercial monoclonal anti-Pol antibody Hep B Pol (8D5, Santa Cruz, #sc-81591). As shown in FIG. 1B, expression of a protein with an apparent molecular weight of 98.2 kDa was detected in presence or in absence of MG-132. This band has the expected molecular weight for the Pol*-TMR protein. It should be noticed that in presence of MG132 more product and an additional product of high molecular weight, over 200 KDa, were detected.

(81) For MVATG17971, Western-blot analysis was performed using a commercial monoclonal anti-core antibody Hep B cAg (13A9) (Santa Cruz, #sc-23946). As shown in FIG. 1B, expression of Core* was detected with an apparent molecular weight of 21 kDa which corresponds to the expected molecular weight.

(82) For MVATG17972, Western-blot analysis was performed using a commercial monoclonal anti-core antibody Hep B cAg (13A9) (Santa Cruz, #sc-23946). As shown in FIG. 1B, expression of Core*t was detected with an apparent molecular weight of 15.8 kDa which corresponds to the expected molecular weight.

(83) For MVATG17993 and MVATG17994, Western-blot analysis was performed using a commercial monoclonal anti-core antibody Hep B cAg (13A9, Santa Cruz, #sc-23946). As shown in FIG. 1B, expression of a protein with an apparent molecular weight of 19.9 and 23.4 kDa respectively was detected. This band has the expected molecular weight for the Core*t-Env1 protein and Core*t-Env1-Env2

(84) 2.2. Immunogenicity of Antigens Expressed from Adenovirus Vectors AdTG17909 and AdTG17910

(85) The immunogenicity of the HBV polypeptides expressed by adenovirus vectors was assessed in HLA-A2 transgenic mice immunized with either AdTG17909 or AdTG17910 alone or with a mixture of the 2 adenoviruses. Specific T cell responses induced following one subcutaneous injection were evaluated by Elispot IFNg, ICS and in vivo cytolysis assays using known (described as being the target of specific T cell responses in patients) HLA-A2 epitopes present in Polymerase, Core or the envelope domains or/and pools of overlapping peptides covering the HBV antigens of interest.

(86) 2.2.1. HBV Specific IFN Producing Cell Evaluation by Elispot Assays

(87) Elispot IFNg assays showed that AdTG17910 is able to induce IFNg producing cells specific of an HLA-A2 restricted epitope (SLYADSPSV) (SEQ ID NO: 55 located within the HBV polymerase at positions 816-824) (FIG. 2A). Immunization with AdTG17909 also resulted in high frequency induction of IFNg producing cells specific for 2 Core HLA-A2 restricted epitopes (FLPSDFFPSV at position 18-27 (SEQ ID NO: 56) and ILCWGELMTL at position 99-108 (SEQ ID NO: 57) as well as for 3 envelope HLA-A2 restricted epitopes (VLQAGFFLL (SEQ ID NO: 58) at positions 14-22 and FLGGTTVCL (SEQ ID NO: 59) at positions 41-49 both present in Env1, and GLSPTVWLSV (SEQ ID NO: 60) at positions 185-194 present in Env2) (FIGS. 2B and C). Immunization with the mixture of AdTG17909 and AdTG17910 also induced a comparable level of specific IFNg producing cells targeting the same epitopes in the 3 antigens, i.e. the SLY epitope present in Pol, the FLP and ILC epitopes in the Core protein, and the 3 epitopes of the envelope domains (VLQ, FLG and GLS) (FIGS. 2A, B and C). Frequency of T cell responses detected following immunization with a single Ad or the mixture of the two was comparable, showing that there is no major immunodominance between the 3 antigens expressed from the described vectors.

(88) 2.2.2. HBV Specific IFNg/TNFa Producing Cell Evaluation by Intracellular Staining Assays

(89) The number of CD8 T cells able to produce either IFNg alone or IFNg+ TNFa targeting HLA-A2 restricted epitopes present in polymerase (SLY) in Core (FLP and ILC) and in envelope domains (VLQ, FLG and GLS) were evaluated by ICS assay. All these epitopes were the target of double and simple secreting cells. The results are shown in FIGS. 3A-3C. Animals immunized with AdTG17909 alone or in combination with AdTG17910 mounted roughly equivalent Core- and Env-specific CD8 T cell responses (same percentages of specific CD8 T cells producing IFNg or IFNg+TNFa following restimulation with FLP, ILC, VLQ, FLG and GLS peptides as shown in FIGS. 3B and 3C). On the other hand, concerning the polymerase specific CD8 T cell response (SLY epitope), a very high percentage of CD8+ cells producing IFNg or IFNg+TNFa was detected in mice treated with AdTG1710 expressing Pol* (FIG. 3A) as well as in those immunized with the mixture of AdTG17910 and AdTG17909 although at a lower level (FIG. 3C).

(90) 2.2.3. HBV Specific IFNg/TNFa Producing CD8 and CD4 T Cell Evaluation Following Immunization with a Mix of Adenovirus Vectors, by Intracellular Staining Assays

(91) The percentages of CD8 and CD4 T cells able to produce either IFNg alone or IFNg+TNFa or IFNg+IL2 targeting HLA-A2 restricted epitopes present in polymerase (SLY) in Core (FLP and ILC) and in envelope domains (VLQ, FLG and GLS) or pools of overlapping peptides covering Core protein and Env domains were evaluated by ICS assay. All the tested HLA-A2 restricted epitopes were the target of single and double secreting cells (IFNg and IFNg+TNFa) and some pools of overlapping peptides were the target of single and double producing cells too (IFNg and IFNg+TNF and IFNg+IL2). The results are shown in FIGS. 4A-4C. Five HLA-A2 transgenic mice were immunized with a mix of AdTG17909 and AdTG17910 and 3 HLA-A2 transgenic mice were immunized with AdTG15149 (negative control). Animals immunized with AdTG15149 displayed no HBV-specific T cell responses (data not shown). Animals immunized with AdTG17909 combined with AdTG17910 displayed a strong CD8 T cell response specific of HBV targeted antigens (FIG. 4A), with high percentage of single (IFNg) and double (IFNg+TNFa) producing cells specific of the HLA-A2 epitopes present in Polymerase, Core and Env domains and specific of the core 1 pool of peptides and the pools of peptides covering Env1 and Env2 domains. As illustrated in FIGS. 4B and 4C, these vaccinated mice also displayed CD4 T cell responses specific of HBV antigens, in particular single (IFNg) and double (IFNg+TNFa and IFNg+IL2) producing cells specific of the core 2 pool of peptides and the Env2-covering pool of peptides.

(92) 2.2.4. Induction of In Vivo Cytolysis Measured by In Vivo CTL Assays

(93) The ability of adenovirus vectors AdTG17909 and AdTG17910 to induce in vivo cytolysis against cells presenting HBV HLA-A2 epitopes was assessed by in vivo CTL assays. Four HLA-A2 epitopes were tested, respectively SLY (Pol), FLP and ILC (Core) and VLQ (Env 1 domain). Six animals were immunized with a mix of AdTG17909+AdTG17910 and 2 animals were immunized with AdTG15149 (negative control). The half of each group (three AdTG17909+AdTG17910 immunized mice and one AdTG15149 immunized mouse) was tested for its ability to lyse in vivo cells pulsed with SLY peptide and cells pulsed with ILC peptide. The other half was tested for the ability of the vaccinated animals to lyse in vivo cells pulsed with FLP peptide and cells pulsed with VLQ peptide. The results are shown in FIGS. 5A and 5B. As expected, no HBV-specific in vivo cytolysis could be detected with AdTG15149 immunized mice (data not shown). In vivo cytolysis against the 2 core epitopes FLP and ILC was weak in AdTG17909+AdTG17910 immunized mice. However, in contrast, animals immunized with the mixture of AdTG17909 and AdTG17910 displayed a strong in vivo cytolysis against the polymerase epitope SLY (FIG. 5A) and Env1 epitope VLQ (FIG. 5B), reaching more than 50% of specific lysis in both cases.

(94) Interestingly, the combination of Ad vectors expressing pol, core and env moieties allows the induction of specific T cell responses targeting the 3 HBV antigens when co-injected. Induced T cells are able to produce one or 2 cytokines and to lyse in vivo cells loaded with some HBV peptides. All together these data demonstrate the immunogenic activity of the compositions described and their ability to induce CD8 and CD4 T cell responses when vectorised by Ad.

(95) 2.3. Immunogenicity of antigens expressed from MVA vectors MVATG17842, MVATG17843, MVATG7971, MVATG179 72, MVATG17993 and MVATG17994 The immunogenicity activity of the MVA-based compositions was assessed in HLA-A2 transgenic mice immunized with one of the MVA vectors described in Examples 1.1.2 to 1.1.7 (MVATG17842, MVATG17843, MVATG17971 or MVATG17972 alone) or with a mixture of 2 MVA (MVATG17843+MVATG17972, MVATG17843+MVATG17993 or MVATG17843+MVATG17994). Mice were immunized with three subcutaneous injections at one week interval and specific T cell responses were evaluated by Elispot IFNg and ICS using the above-described HLA-A2 epitopes present in Polymerase, Core or the envelope domains or/and pools of overlapping peptides covering the HBV antigens of interest.
2.3.1. HBV Specific IFN Producing Cell Evaluation by Elispot Assays Following Immunization with Polymerase Expressing MVA.

(96) Three mice were immunized with either MVATG17842 expressing a truncated and mutated polymerase antigen or MVATG17843 expressing a membrane-targeted version of the same truncated and mutated polymerase or MVA N33.1 (negative control). Polymerase-specific T cell responses were evaluated by IFNg Elispot assays using the SLY HLA-A2 restricted epitope and pools of peptides covering the polymerase. No HBV-specific T cell response was detected for mice immunized with MVA N33.1 and with MVATG17842 (data not shown). However, as shown in FIG. 6A, IFNg producing cells were induced following immunization with MVATG17843 which are specific of the HLA-A2 restricted epitope SLY and of the peptide pool 8 covering the C-terminal portion of the polymerase (No specific response could be detected against the other peptide pools 1-7 under the tested experimental conditions). These data highlight the benefit of expressing polymerase as a membrane-anchored antigen at least in MVA-based compositions.

(97) 2.3.2. HBV Specific IFN Producing Cell Evaluation by Elispot Assays Following Immunization with Core Expressing MVA.

(98) Eighth mice were immunized with either MVATG17971 expressing a core moiety deleted of residues 77-84 or MVATG17972 expressing a truncated version thereof (C-terminal truncation from residue 149) or MVA N33.1 (negative control). Core specific T cell responses were determined by IFNg Elispot assays using HLA-A2 restricted epitopes (FLP and ILC peptides) and the above-described core 1 and core 2 pools of peptides. As illustrated in FIG. 6B, immunization with MVA TG17971 is able to induce sporadic T cell responses specific against HLA-A2-restricted FLP and ILC peptides and the two peptide pool covering core antigen (FIG. 6B). No core specific T cell responses could be detected in mice immunized with MVATG17972 using the tested peptides and under the tested experimental conditions (data not shown).

(99) 2.3.3. HBV Specific IFN Producing Cell Evaluation by Elispot Assays Following Immunization with Combination of MVA Vectors.

(100) Three mice were immunized with a mixture of MVA TG17843 and either MVA TG17972, MVA TG17993 and MVA TG17994 and HBV specific T cell responses were evaluated by Elispot IFNg assay using the above-described peptides. Polymerase specific T cell responses were detected in the vast majority of animals vaccinated with the combination MVATG17843+MVATG17972 (positive responses in animals as shown in FIG. 7A), MVATG17843+MVATG17993 (positive responses in 3/3 animals as shown in FIG. 7B) and MVATG17843+MVATG17994 (positive responses in animals as shown in FIG. 7C). Frequency of IFNg producing cells seems to be comparable with the one observed when MVA TG17843 was injected alone (FIG. 6A) which demonstrates that vector combination is not detrimental to the induced immune response. no core-specific response could be detected under the experimental conditions tested in any of the vaccinated animals using HLA-A2 or pools of peptides (data not shown). no env-specific response could be detected under the experimental conditions tested in any of the animals vaccinated with the MVATG17993-comprising combination using Env1 HLA-A2 or pools of peptides (data not shown). Env 2-specific T cell responses were detected in 2 out of 3 animals immunized with the MVATG17994-comprising combination as illustrated in FIG. 7C which were directed against both the HLA-A2-restricted GLS epitope and the Env2-covering pool of peptides (no detection of T cell response against Env1 domain could be observed under these experimental conditions; data not shown).

(101) ICS assays performed in the same conditions confirmed the results observed in Elispot assays (data not shown).

(102) Interestingly, the combination of MVA vectors expressing pol, core and env moieties allows the induction of specific T cell responses targeting the pol and env2 antigens when co-injected.

(103) All together, these data demonstrate the immunogenic activity of the compositions described and their ability to induce T cell responses against the major HBVantigens.

3. Cross Reactivity of T Cells Induced by HBV AdTG17909 and AdTG17910 Vaccine Candidates

(104) At present time, 10 genotypes and many subtypes have been defined for HBV based on the natural variability existing within HBV proteins. As discussed in the above description, the divergence in the complete viral genomic sequence is more than 8% between genotypes and from 4% to 8% between subtypes. The geographic distribution of these HBV genotypes is different depending on the regions of the world (Lin et al, 2011, J. Gastro Hepatol. 26, 123-130). The genotype A is mainly prevalent in Africa and in northern Europe. Genotypes B and C are highly prevalent in Asia, in particular in China and the genotype D is mostly represented in Europe and mediterranean countries. The genotype F is found in South America as well as Central America where the genotype H is also represented. The genotype G is found in France and Germany for Europe and in the United States. Although the distribution of the newly identified genotypes I and J remains to be sharpened, they have been mainly found in Vietnam, Laos and Japan.

(105) As described above (e.g. in example 2.2) AdTG17909 and AdTG17910 induced potent IFNg T cell responses against six HLA-A2 restricted epitopes. It is interesting to document the cross-reactive potential of T cell responses induced by such genotype D-based vaccine candidates against B and C genotypes HBV viruses. Only few publications addressed the question of the T cell cross-reactivity in the HBV context (Liu et al., 2007, Clin. Immunol. 125, 337-345; Riedl et al., 2006, J. Immunol. 176, 4003-4011).

(106) The cross-reactivity study described hereinafter was conducted in 2 steps: The 1.sup.rst step was an in silico analysis of sequences from genotypes B, C and D in order to determine the theoretical variability both intra- and inter-genotype at the global antigen level and at the T cell epitope level by identifying major, minor and rare variants of known T cell epitopes within HBV antigens of interest. A 2.sup.nd step was an in vivo analysis in a preclinical mouse model in order to determine whether T cells induced by the genotype D based HBV vaccine candidates are able to recognize epitope variants from genotype B, C and D.

3.1 Materials and Methods

(107) 3.1.1 in Silico Study of HBV Sequences

(108) The natural variation of sequences of HBV antigens and T cell epitopes within one viral genotype and among different viral genotypes, was evaluated by conducting various queries in GeneBank Nucleotide database. For illustrative purposes, a preliminary query based on Hepatitis B virus and whole genome and Genotype B, C or D, and not fulminant key words resulted in 455, 827 and 369 entries for genotypes B, C and D respectively; Entries with indication non-functional protein for either of proteins of interest were skipped. Generator of random numbers were then used for selection of a given entry, for which accession number and amino acid sequences of each protein of interest (Polymerase, Core and Envelope) were downloaded in Excel spreadsheet program as text variables. Selection process was renewed until the limit of 100 acceptable sequences for each of B, C and D genotypes that fit with the defined criteria was reached.

(109) 3.1.2 Sequences Alignment and Definition of a Consensus Sequence Per Genotype

(110) For each protein of interest, alignments of the 100 selected sequences/genotypes were performed using ClustalW2 program on EMBL site (www.ebi.ac.uk/Tools/msa/clustalw2/) using default options. Results were downloaded in MS Word file. The obtained consensus sequences for genotypes B, C and D were then aligned with the genotype D prototype sequence (Y07587) encoded by AdTG17909 and AdTG17910 HBV vaccine candidates.

(111) 3.1.3 Study of Epitope Variant Distribution

(112) The study of epitope variant distribution was focused on class I epitopes restricted by HLA haplotypes that are mainly represented in Caucasian (HLA-A2, HLA-B7) and Asian (HLA-A24 and -A11) populations and 49 different epitopes were selected on this basis.

(113) Their natural variants were searched through protein sequences selected as previously described. The total number of sequences per genotype that include every given variant was obtained. Since exactly 100 sequences per genotype were studied, these numbers are equivalent to variant frequencies (expressed in %). Epitope variants existing in more than 50% of sequences of one genotype were called major variants. Epitope variants existing in 5% to 50% of sequences of one genotype were called minors variants. Epitope variants existing in less than 5% of sequences of one genotype were referred as rare variants. Major variant were identified for each epitope and each genotype and then aligned and compared.

(114) 3.1.4. Epitope Variants Selected for Ex Vivo Testing

(115) As illustrated in the following Table 1, major and minor variants of the 6 HLA-A2-restricted epitopes against which AdTG17909 and AdTG17910 induced potent IFNg T cell response were selected and synthesized as peptides to be tested in ex-vivo experiments. Rare variants (representing <5% of sequences) were not considered in the study.

(116) TABLE-US-00001 TABLE1 ThesixHLA-A2restrictedepitopesselectedforex-vivotesting,majorand minorvariantsandfrequencies. HBV Percentage protein epitope Name sequence GenotypeD GenotypeC GenotypeB Core FLP FLP FLPSDFFPSV 85 12 7 (SEQIDNO:56) FLP4 FLPSDFFPSI 1 84 79 (SEQIDNO:62) ILC ILC ILCWGELMTL 69 3 4 (SEQIDNO:57) ILC2 ILCWGELMNL 4 68 71 (SEQIDNO:63) ILC3 IVCWGELMNL 0 6 15 (SEQIDNO:64) ILC4 ILCWGDLMTL 11 0 0 (SEQIDNO:65) ILC5 IVCWGELMTL 0 6 0 (SEQIDNO:66) ILC6 ILCWVELMNL 0 5 1 (SEQIDNO:67) ILC7 VLCWGELMTL 1 5 0 (SEQIDNO:68) Env VLQ VLQ VLQAGFFLL 97 93 78 (SEQIDNO:58) VLQ2 VLQAGFFSL 0 2 11 (SEQIDNO:69) FLG FLG FLGGTTVCL 84 4 0 (SEQIDNO:59) FLG2 FLGGTPVCL 3 2 72 (SEQIDNO:70) FLG3 FLGGAPTCP 1 68 5 (SEQIDNO:71) FLG4 FLGETPVCL 0 0 10 (SEQIDNO:72) GLSP GLSPTVWLSV 78 94 96 (SEQIDNO:60) GLSP2 GLSPTVWLLV 9 0 1 (SEQIDNO:73) GLSP3 GLSPIVWLSV 6 0 1 (SEQIDNO:74) Pol SLY SLY SLYADSPSV 89 5 87 (SEQIDNO:55) SLY2 SLYAVSPSV 4 81 6 (SEQIDNO:75)

(117) 3.1.5 Immunization and Read Out

(118) The HLA-A2.1 mice were immunized by subcutaneous route with a mixture of AdTG17909 and AdTG17910 containing 10.sup.8 iu of each Ad preparation or with 210.sup.8 iu of the empty AdTG15149 as a negative control. Thirteen to 15 days after immunization, animals were euthanized and spleens were aseptically removed for further ex-vivo Elispot IFN-gamma assay and Intracellular Cytokine Staining (ICS) analysis.

(119) IFNg Elispot assays were performed as described in Example 1.2.4 except the peptides used for stimulation: Irrelevant peptide (10 M, DLMGYIPLV HLA-A2 peptide from HCV Core, synthesized by Eurogentec; SEQ ID NO: 76) or, Concanavalin A (5 g/ml Sigma, C5275) as a positive control or, HBV Pol, Core and Env selected peptides described in Table 1 (10 M, synthesized by Eurogentec).

(120) Results are shown as the mean value obtained for triplicate wells, related to 10.sup.6 cells. The technical cut-off linked to variability of the CTL ELISpot reader is 10 counted spots/well as it has been determined that 10 counted spots is the threshold above which the variation coefficient is never more than 20%. According to this technical cut-off, in this study, a response was then considered positive if the number of spots was higher than 50 spots per 10.sup.6 cells. The experimental cut-off defined as the mean number of spots observed with medium alone+2 Standard Deviations (SD) was also calculated. Only specific responses displaying a number of spots higher than the 2 cut-offs are considered as positive.

(121) Double Intracellular Cytokine Staining Assays (IFNg/TNFa were performed as described in Example 1.2.5. CD8.sup.+ CD4.sup. cells were gated and presented on IFNg/TNFa dot-plot. Four quadrants were defined to gate positive cells for either one cytokine (IFNg-SP or TNFa-SP) or both cytokines simultaneously (IFNg/TNFa-DP) or cells negative for both cytokines. Numbers of events found in these quadrants were divided by total number of events in CD8.sup.+ CD4.sup. parental population thus giving percentages of responding CD8.sup.+ T cells. For each mouse, the percentage obtained in medium alone was considered as background and subtracted from the percentage obtained with peptide stimulations. Response was considered positive if the percentage of stained cells was higher than 3 times the SD of background values (obtained with medium alone) found for all animals in the particular protocol and higher than 0.244% of CD8.sup.+ T cells (technical cut-off defined for HLA-A2 mice and based on ability of FacsCalibur to detect rare events).

(122) Given that the percentage of TNFa-SP CD8.sup.+ cells were never found above cut-off, analyses were done for total IFNg-positive CD8.sup.+ cells accounting for both IFNg-SP and IFNg/TNFa-DP cells.

(123) Statistical analysis of the resultswere performed using Statistica 9.0 (2009; StatSoft).

(124) 3.2 Results

(125) 3.2.1. In Silico Analysis ofHBV Sequence Variability

(126) Analysis of the Whole Antigen Sequences

(127) Consensus sequences of HBV polymerase, core and env antigens were obtained from multiple alignments of HBV sequences of genotypes B, C and D (100 sequences/genotype). These consensus sequences were aligned as described in Materials and Methods with the HBV genotype D near-consensus prototype sequence (Y07587) encoded by AdTG17909 and AdTG17910. The Y07587 sequence differs from the genotype D consensus sequence, by only 6 amino acids in the polymerase and one in Core. No amino acid difference was identified for the Envelope antigen. The sequence homologies and variations between the Y07587 sequence was also studied with respect to the consensus sequences of genotypes B and C. The results are presented in Table 2.

(128) TABLE-US-00002 TABLE 2 sequence conservation between HBV consensus sequences from genotypes B and C and Y07587 genotype D sequence present in AdTG17909 and AdTG17910 vaccine candidates Antigens or domains .Math. Pol Core Env Env1 Env2 Env4 Total number of amino acids .Math. 843* 183 226 38 30 25 Number (and %) of amino B = C = D 700 (83%) 174 (95%) 199 (88%) 33 (87%) 29 (97%) 23 (92%) acid positions within the B = C D 57 (7%) 9 (5%) 7 (3%) 1 (2.5%) 1 (3%) 1 (4%) antigen or domain for {open oversize bracket} C = D B 39 (5%) 0 12 (5%) 1 (2.5%) 0 1 (4%) which conservation and B = D C 26 (3%) 0 8 (4%) 3 (8%) 0 0 differences are: B C D 21 (2%) 0 0 0 0 0 *accounting for 11 AA insertion observed in B and C genotypes.

(129) In this alignment, when the amino acid is identical for the 3 sequences, it was referred as B=C=D. When differences were observed among the 3 genotypes for an amino acid position, they were classified as follow: B=CD means that the amino acid is identical between genotype B and C but different for the genotype D, B=DC means that the amino acid is identical between genotype B and D but different for the genotype C, C=DB means that the amino acid is identical between genotype C and D but different for the genotype B, BCD means that the amino acid is different in all 3 genotypes. Number and percentage of conserved amino acid position as well as number and percentage of positions where differences were detected within each antigen and domain were also indicated. Of note, as shown in Table 2, the HBV proteins appeared highly conserved between genotypes B, C and D with very high percentages of homology, ranging from 83% for the polymerase protein to 95% for the Core protein. For the polymerase protein, although most of amino acid positions are identical among the 3 genotypes (700 out of 843), distinct differences were observed among the 3 genotypes for the 143 other amino acid positions. For example, 57 amino acid positions are identical between genotypes B and C but different in genotype D (notation B=CD). This is mainly due to the insertion of 11 amino acids in the polymerase sequence of genotypes B and C compared with the sequence of genotype D. 39 amino acid positions are identical between genotypes C and D but different in genotype B (notation C=DB), etc and 21 positions are different in all genotypes (notation BCD). (see Table 2). For the Core protein, all amino acids are identical between genotypes B and C sequences (meaning that the consensus sequences from genotype B and genotype C are identical for the Core protein) and only 9 amino acids differ from the genotype D sequence. For the envelope protein, as for the polymerase protein, although most of amino acid positions are identical among the 3 genotypes (199 out of 226), differences are seen for 27 amino acids positions (B=CD; B=DC; C=DB). The most frequent case of difference was found when amino acids were identical between genotype C and D but different for genotype B (notation C=DB). Concerning the specific domains Env1, Env2 and Env4, the Env2 domain is mainly conserved with only one different amino acid among the 3 genotypes, which is conserved between genotypes B and C but differs from genotype D. The Env1 domain is slightly more variable (5 amino acids that differ among the 3 genotypes) with amino acids that are identical in 2 genotypes and differ in the 3rd one (B=CD; B=DC; C=DB). The Env4 domain is in an intermediary situation with 2 different amino acids among the 3 genotypes, which are conserved in 2 genotypes but differ from the 3rd one (B=CD; C=DB).

(130) Considering the full length sequences for each antigen, alignments confirm a very high conservation of HBV sequences among genotypes B, C and D. Various examples of differences and conservation among the 3 genotypes were observed for polymerase and HBsAg showing that genotypes B, C and D are similarly different for these antigens. The case of the core protein is quite different as consensus sequences of genotype B and C are strictly identical, suggesting a high identity of core protein sequences between these 2 genotypes. In addition, these Core consensus sequences are also very close to the Core genotype D sequence present in AdTG17909 as there is 95% of identity at the amino acid level. Overall, these results confirmed the very high conservation of the Core protein.

(131) Analysis of the T Cell Epitope Sequences

(132) The study of T cell epitope sequences was focused on class I epitopes restricted by HLA haplotypes that are mainly represented in Caucasian population (HLA-A2, HLA-B7) and in Asian population (HLA-A24, -A3, -A11). Forty-nine epitopes were studied for their variants in the different genotype sequences (summarized in Table 1). Epitope variants existing in more than 50% of sequences of one genotype were called major variants. Epitope variants existing in 5% to 50% of sequences of one genotype were called minors variants. Epitope variants existing in less than 5% of sequences of one genotype were referred as rare variants. The major variant of each epitope for each genotype was defined and major variants of the 3 genotypes were then compared (Table 3).

(133) TABLE-US-00003 TABLE 3 sequence conservation between HBV major epitope variants of genotypes B, C and D. Antigens and domains .Math. Pol Core Env Env1 Env2 Env4 Total number of analysed class 1 epitopes .Math. 18 17 14 4 4 5 Number (and %) of B = C = D 16 (89%) 9 (53%) 8 (57%) 2 (50%) 4 (100%) 1 (20%) epitopes for which major B = C D 0 8 (47%) 0 0 0 0 variants are classified as: {open oversize bracket} C = D B 0 0 4 (29%) 1 (25%) 0 3 (60%) B = D C 2 (11%) 0 0 0 0 0 B C D 0 0 2 (14%) 1 (25%) 0 1 (20%)

(134) Major variants of studied epitopes (see Table 1) were obtained as described in Materials and Methods for each HBV genotypes (C, B and D). When the amino acid sequence was the same for major variants of all 3 genotypes, major variants of the epitope were referred as B=C=D. When the amino acid sequence was the same for major variants of genotypes B and C but different from the one of genotype D, major variants of the epitope were referred as B=CD. When the amino acid sequence was the same for major variants of genotypes B and D but different from the one of genotype C, major variants of the epitope were referred as B=DC. When the amino acid sequence was the same for major variants of genotypes C and D but different from the one of genotype B, major variants of the epitope were referred as C=DB. When the amino acid sequence was different for major variants of the 3 genotypes, major variants of the epitope were referred as BCD. Number and percentage of epitopes with the same major variants among the 3 genotypes as well as number and percentage of epitopes with different major variants among genotypes are presented in Table 3 for the each antigen and domains.

(135) As shown in Table 3, 18 class I epitopes of the polymerase protein were analysed: the major variants of 16 epitopes are identical between the 3 genotypes and the major variants of the 2 other epitopes are identical for genotypes B and D but differ from the major variant of genotype C. For the Core protein, 17 class I epitopes were analysed: the major variants of 9 epitopes are identical for the 3 genotypes and the major variants of 8 epitopes are identical for genotype B and C but differ from genotype D. For the envelope protein, 14 class I epitopes were analysed: major variants of 8 of them are identical among the 3 genotypes, 2 are different in the 3 genotypes and 4 are identical for genotypes C and D but different for the genotype B.

(136) In conclusion the class I epitopes are mostly conserved among the 3 genotypes B, C and D. For 33 out of the 49 analysed epitopes, the main variant is identical for the 3 genotypes, for 14 epitopes the main variant is identical for 2 genotypes out of the 3 and for 2 epitopes only, the main variant is different for the 3 genotypes. These in silico results confirmed the high conservation of the amino acid sequences of HBV proteins at the T cell epitope level.

(137) 3.2.2. Cross Reactivity Studies Following AdTG17909 and AdTG17910 Immunization of HLA-A2 Transgenic Mice

(138) The ability of the AdTG17909 and AdTG17910 to induce IFNg producing cross-reactive T cells against the HBV Core, Envelope and Polymerase antigens was assessed in HLA-A2 transgenic mice immunized with a mix of 10.sup.8 iu of each Ad vector (210.sup.8 iu in total) by subcutaneous route. Splenocytes of vaccinated mice were collected 2 weeks after immunization and in vitro stimulated with peptides that are homologous to the sequence encoded by the AdTG17909 (Core and Env domains) or the AdTG17910 (Polymerase) or with the major and minor variants of genotype D, B and C (see Table 1). IFNg Elispot and ICS assays were performed as described in Materials and Methods

(139) Cross-Reactivity of T Cell Responses Targeting the HBV Core Antigen

(140) Two core epitopes were tested including the homologous genotype D FLP and ILC and their variants listed in Table 1 representative of genotypes B and C.

(141) FLP epitope and variants: 13 mice were immunized by the AdMix and splenocytes of the immunized mice were stimulated with either the homologous (genotype D) FLP epitope (FLPSDFFPSV; SEQ ID NO: 56) or the major variant FLP4 representative of genotypes B and C (FLPSDFFPSI; SEQ ID NO: 62).

(142) As illustrated in FIG. 8A, Elispots IFNg showed a similar frequency of responders against FLP and FLP4 (54% versus 46% respectively) and a similar frequency of specific IFNg producing cells recognizing FLP or FLP4. No statistical difference was observed between FLP and FLP4

(143) As illustrated in FIG. 8B, IFNg ICS assay displayed a similar profile with the same responder frequency (92% in both cases) and the same level of responses comparing the FLP peptide and its FLP4 variant.

(144) Elispots and ICS results lead to the conclusion that T cells induced by the Ad mix are mainly cross-reactive and able to recognize the major variants of FLP for genotype D (homologous sequence) and for genotype B and C (FLP4 sequence).

(145) ILC Epitope and Variants:

(146) The ILC peptide (ILCWGELMTL; SEQ ID NO: 57), homologous to the genotype D core sequence encoded by AdTG17909 was tested as well as six ILC variants (ILC2 to ILC7 identified in Table 1 and corresponding to SEQ ID NO: 63 to 68, respectively). ILC2 is the major variant of both genotypes B and C whereas ILC3 to 7, are minor variants of the 3 genotypes.

(147) As shown in FIG. 8A, ELISpots IFNg assays showed a high frequency of responding mice against ILC as well as ILC2 and ILC5 variants. Lower frequency was observed for variants ILC3, 4 and 7 and no responding mouse for ILC6 variant. The level of responses observed against ILC2 and ILC5 reaches 62 and 43% respectively of the response against the ILC peptide. The level of responses against ILC3, 4, 6 and 7 is even lower, being less than 20% for ILC3, 4 and 7 and 2% for ILC6.

(148) In ICS assay (FIG. 8B), a high frequency of responding mice (from 85% to 100%) was found for ILC and all variants, except the ILC6 (only one mouse out of 13 responding to ILC6). When comparing with the response observed against the homologous ILC peptide (100%), high level of responses was obtained after ILC2 and ILC5 stimulation (78% and 87% of the ILC-induced response). Level of responses targeting ILC3 (51%), ILC4 (28%) and ILC6 (1%) was statistically lower than the one targeting the ILC peptide. However, ILC7-induced response detected by ICS was much higher than the one measured in ELISpot.

(149) The combined results of ELISpot and ICS assays lead to the conclusion that T cell responses induced by the AdTG17909 are mainly cross-reactive, recognizing the major ILC2 variant of genotype B and C as well as most of the minor variants of the 3 genotypes B, C and D. Number of responding mice is similar for all the variants (except ILC6), even if the level of responses displayed by these mice is statistically lower than the one targeting the homologous peptide ILC.

(150) In conclusion for the Core protein, AdTG17909-induced T cell responses are mostly cross reactive against the major variants of the 3 genotypes and even some minor variants.

(151) Cross-Reactivity of T Cell Responses Targeting the HBV Envelope

(152) Three epitopes of the Envelope antigen (VLQ and FLG contained in Env1 domain and GLS contained in Env2 domain) and their variants were tested.

(153) VLQ Epitope and Variants:

(154) The VLQ peptide (VLQAGFFLL; SEQ ID NO: 58) is the major variant of the 3 genotypes B, C and D whereas VLQ2 (VLQAGFFSL; SEQ ID NO: 69) is a minor variant of genotype B (and rare variant of genotype C).

(155) As shown in FIGS. 9A and 9B, Elispots IFNg assays (FIG. 9A) showed a high frequency of responding mice (100%) against VLQ itself whereas a significantly lower number of responding mice was detected against the VLQ2 (38%). ICS assays (FIG. 9B) confirmed this trend with lower frequency of responding mice against VLQ2 (77%) than against VLQ (100%) and a level of induced T cells which is significantly lower against VLQ2 than against VLQ (p=0.0028).

(156) These results indicate that T cell response is cross-reactive against the minor VLQ2 variant of genotype B but less potent than the one against the VLQ peptide.

(157) FLG Epitope and Variants:

(158) ELISpot and ISC assays were performed using the FLG peptide (FLGGTTVCL; SEQ ID NO: 59), the major variant of genotype D and three other variants. FLG2 (SEQ ID NO: 70) is the major variant of genotype B, FLG3 (SEQ ID NO: 71) is the major variant of genotype C and FLG4 (SEQ ID NO: 72) is a minor variant of genotype B.

(159) ELISpots IFNg assays (FIG. 9A) showed a high frequency of responding mice (85%) and a high level of T cell responses against the homologous peptide FLG. The frequency of responding mice for FLG2, 3 and 4 variants is drastically lower (respectively 15%, 54% and 8%) and the level of detected T cell responses is also lower than for the FLG peptide (respectively 4%, 21% and 3% of the level of T cell responses detected for FLG). Statistical analyses confirmed that these differences between FLG and the 3 variants are significant.

(160) ICS assays (FIG. 9B) confirmed the same trend. A high frequency of responding mice (100%) and a high level of IFNg producing cells were detected for the FLG peptide and frequency of responding mice and level of T cell responses are significantly lower for FLG2 (8% of responding mice and 3% of the FLG-response), FLG3 (62% of responding mice and 29% of the FLG-response) and FLG4 (no responding mouse). Statistical analyses on these assays also confirmed that T cell responses observed against the 3 variants FLG2, 3 and 4 are significantly lower than the one against the FLG epitope.

(161) Thus, Elispots IFN and ICS assays demonstrated that AdTG17909-induced T cell responses are poorly cross-reactive against the major variants of genotype B and C of the FLG epitope.

(162) GLSP Epitope and Variants:

(163) GLSP (GLSPYVWLSV; SEQ ID NO: 60) is the major variant of the 3 genotypes B, C and D and GLSP2 (SEQ ID NO: 73) and 3 (SEQ ID NO: 74) are minor variants of genotype D.

(164) Elispots IFNg assays (FIG. 9A) showed a high frequency of responding mice (100%) as well as high frequency of specific T cells producing IFNg against the homologous peptide GLSP. Frequency of responding mice and of specific IFNg producing cells targeting the variants GLSP2 and GLSP3 are lower than those targeting GLSP. This difference is statistically significant comparing GLSP with GLSP2 or GLSP3.

(165) ICS assays (FIG. 9B) showed the same trend with high frequencies of both responding mice and IFNg-producing T cells against the homologous GLSP peptide. A high frequency of responding mice was also observed against GLSP2 and GLSP3 (respectively 100% and 85%) but the frequency of IFNg-producing cells was lower than the one observed for GLSP (78% and 36% of the GLSP-specific response for GLSP2 and GLSP3). Statistical analyses showed that the difference between GLSP and GLSP3 is statistically significant but not between GLSP and GLSP2.

(166) The T cell response specific of the GLSP peptide was strong, whatever the genotype, as the major variants of the GLSP epitope have the same amino acid sequences for the 3 genotypes. T cell responses targeting the GLSP2 and GLSP3 minor variants are cross reactive but to a poor extend for GLSP3

(167) In conclusion, 2 out of the 3 HLA-A2 tested epitopes in the HBsAg domains encoded by AdTG17909 are highly conserved among the 3 genotypes. Thus, T cell responses targeting these epitopes are strong in all cases. For the FLG epitope which is not conserved between the 3 genotypes, the induced T cells targeting the major variants of genotype B and C are cross-reactive to some extent with a high frequency of responding mice but the level of T cell responses is significantly lower than levels observed against the homologous peptide.

(168) Cross-Reactivity of T Cell Responses Targeting the HBV Polymerase

(169) SLY Epitope and Variants:

(170) The SLY peptide (SLYADSPSV; SEQ ID NO: 55) is the major variant of genotypes B and D. SLY2 (SLYAVSPSV; SEQ ID NO: 75) is the major variant of genotype C (other identified variants being only rare variants thus not analysed in the in vivo study).

(171) As expected, ELISpot assays showed a high percentage of responding mice (100%) and a high frequency of induced IFNg producing T cells targeting the homologous SLY peptide. However, the ability of inducing SLY2-recognizing T cells was dramatically low with only 1 responding mouse out of the 10 tested mice. The same trend was observed in ICS with 100% of responding mice against the SLY peptide and 0% against the SLY2 peptide (significant difference).

(172) Overall, our data show that induced SLY-specific T cell responses recognized the major variant of genotype B and D as this major variant is the same but are not able to recognize the major variant of genotype C. A global conclusion regarding cross-reactive responses targeting the polymerase cannot be fully derived as analysis was limited to a single epitope for this antigen.

3.3 Conclusion

(173) The in silico study showed at the global antigen level that the amino acid sequence of Polymerase, Core and Envelope proteins is highly conserved within the same genotype but also among genotypes B, C and D. This study also highlights that the sequence is also highly conserved at the T cell epitope level within these 3 proteins between European (genotype D) and Asian haplotypes (genotypes B and C). It was assumed that the overlapping open-reading frames of the HBV genome provide a high constraint that may limit sequence variability and, thus, appearance of escape mutations in T cell epitopes. This is in contrast to what has been described for other viruses such as HCV for which effective T cell responses are thought to be responsible for the selection of escape mutants (Petrovic et al., 2011, Eur. J. Immunol. 42, 1-10).

(174) The in vivo study focused on 6 HLA-A2 epitopes targeting T cells and identified in core (FLP and ILC) and polymerase (SLY) antigens and env domains (VLQ, FLG and GLSP). The ELISpot and ICS results showed that the T cell responses induced following AdTG17909 and AdTG17910 immunization are generally cross reactive and able to recognize the major variants of genotype B, C and D as well as some minor variants. This good level of cross-reactivity of induced T cells is due in some cases to the high conservation of sequences (major variants such as GLSP and VLQ epitopes are identical between the 3 tested genotypes and identical to the Y07587 sequence included in AdTG17909 or AdTG17910) but also to the capacity of induced T cells to recognize heterologous sequences (FLP and ILC epitopes). For 2 out of the 6 tested epitopes, SLY and FLG epitopes, the cross reactivity of T cell responses appeared to be weaker. The induced T cell responses are not able to recognize respectively the SLY major variant of genotype C (SLY2) and the FLG major variant of genotype B (FLG2) as no responding mice were detected. For the FLG major variant of genotype C, responding mice were detected, suggesting that cross reactivity exists but the level of T cell response is significantly lower than the one observed with the FLG epitope. It was also shown that the induced T cells are able to recognize some minor variants of these epitopes since mice are responding to these variants although the level of induced T cells is generally lower than against the homologous peptide. Overall, vaccination with a genotype D sequence in this mouse model allows inducing T cell responses recognizing 5 out of 6 genotype B major variants of the tested HLA-A2 epitopes as well as 5 out of 6 genotype C major variants of tested HLA-A2 epitopes. Even if this study is limited to HLA-A2 epitopes, these results are the proof of concept that a HBV genotype D prototype sequence encoded by a vectorized vaccine candidate is able to induce T cell responses that are mostly cross-reactive with HBV sequences of genotype B, C and D. The cross reactivity potential provided by genotype D-based AdTG17909 and AdTG17910 is in favour of a large use of such vaccine candidates not only restricted to genotype D-infected patients but also broadened to genotype B or C-infected patients.