Adenovirus vaccine vectors

Abstract

The invention relates to recombinant adenovirus displaying one or more heterologous epitope(s) on their fiber protein. These recombinant adenovirus are useful as vaccines for generating an immune response against said epitope(s) in individuals having a pre-existing anti-Ad immunity.

Claims

1. A method of generating a humoral immune response against one or more target B-cell epitope(s) in a subject having a pre-existing humoral immunity against an adenovirus of a given serotype resulting from a previous exposure to a wild-type or recombinant adenovirus, comprising: administering to the subject a recombinant replication-defective adenovirus of the same serotype, wherein the subject has pre-existing humoral immunity to the same adenovirus serotype of said replication-defective adenovirus; said replication defective adenovirus has a heterologous polypeptide of up to 50 amino acids containing one or more target B-cell epitope(s) from a foreign antigen of interest inserted into the HI loop of the knob of a fiber protein of the replication-defective adenovirus, and an IgG humoral immune response against the target B-cell epitope(s) in the subject is enhanced by a pre-existing humoral immunity against said adenovirus serotype.

2. A method of generating a humoral immune response against one or more target B-cell epitope(s) in a subject comprising: (a) administering to the subject a recombinant replication-defective adenovirus having a heterologous polypeptide of up to 50 amino acids containing one or more target B-cell epitope(s) from a foreign antigen of interest inserted into the HI loop of the knob of a fiber protein of the recombinant replication deficient adenovirus, and (b) re-administering the recombinant replication-defective adenovirus at least once to the subject, wherein IgG humoral immune response against the target B-cell epitope(s) is enhanced by humoral immunity generated against the recombinant replication-defective adenovirus.

3. The method of claim 1, wherein said recombinant replication-defective adenovirus further comprises one or more heterologous polypeptide(s) containing the same target B-cell epitope(s) inserted into a capsid protein other than the fiber protein.

4. The method of claim 1, further comprising re-administering at least once said recombinant replication-defective adenovirus to said subject.

5. The method of claim 3, further comprising re-administering at least once said recombinant replication-defective adenovirus to said subject.

6. The method of claim 1, wherein the subject is human.

7. The method of claim 2, wherein the subject is human.

8. The method of claim 3, wherein the subject is human.

9. The method of claim 4, wherein the subject is human.

10. The method of claim 5, wherein the subject is human.

Description

FIGURE LEGENDS

(1) FIG. 1 Epitope detection on capsid-modified Ad. (A) Silver staining of capsid-modified Ad. 10.sup.10 vp (viral particles) of either a control Ad (AdWT) or a capsid-modified Ad (AdH-3OVA2, AdH-OVA, AdF-3OVA2, AdF-OVA) were separated on a 10% polyacrylamide gel. Major capsid components are identified (black arrow) and the difference between modified or native fiber is indicated (white arrow). (B and C) Detection of OVA- and 3OVA2-epitopes on virions. ELISA plates were coated with 100 ng of native (B) or denaturated (C) viruses and incubated with a rabbit polyclonal antibody against the ovalbumin protein. The binding was detected with HRP-conjugated secondary antibody. One of two experiments is shown, n=6; means+SD of sixtuplates.

(2) FIG. 2 In vitro gene transfer by capsid-modified Ad. CHO-CAR were mock-infected (PBS) or infected with increasing MOI of IacZ recombinant AdWT, AdH-OVA, AdH-3OVA2, AdF-OVA and AdF-3OVA2. ?-Gal activity, expressed as RLU/?g of protein, was measured in cell lysates 24 h later. Experiments run in duplicate were performed twice and representative results are shown.

(3) FIG. 3 Kinetic of anti-ovalbumin humoral response. C57Bl/6 mice were immunized intra-peritoneally with 10.sup.10 vp of a control Ad (AdWT) or one capsid-modified Ad (AdH-3OVA2, AdH-OVA, AdF-3OVA2 or AdF-OVA). Anti-ovalbumin IgG titers were determined by ELISA at different days p.i. (post-injection). Ab titers below 100 were plotted as 50. One of two experiments is shown, circles and bars represent results of individual mice (n=6) and means, respectively.

(4) FIG. 4 Characterization of anti-ovalbumin Abs isotypes. C57Bl/6 mice were immunized intra-peritoneally with 10.sup.10 vp of a control Ad (AdWT) or one capsid-modified Ad (AdH-3OVA2, AdH-OVA, AdF-3OVA2 or AdF-OVA). Anti-ovalbumin IgG1, IgG2a and IgG2b titers were determined by ELISA at day 21 p.i. Ab titers below 100 were plotted as 50. One of two experiments is shown, n=6; means+SD.

(5) FIG. 5 Dose-dependence of anti-ovalbumin humoral response. C57Bl/6 mice were immunized intra-peritoneally with 10.sup.9 or 10.sup.10vp of AdH-3OVA2 or AdF-3OVA2. Anti-ovalbumin IgG titers were determined by ELISA at day 14 p.i. One of two experiments is shown, circles and bars represent results of individual mice (n=6) and means, respectively.

(6) FIG. 6 Repeated administration of capsid-modified Ad. C57Bl/6 mice were immunized twice two weeks apart intra-peritoneally with 10.sup.10 vp of AdWT or capsid-modified Ad and sera were collected at day 14 p.i. Anti-ovalbumin, anti-?gal and anti-Ad IgG titers were determined by ELISA at day 14 p.i. Ab titers below 100 were plotted as 50. One of two experiments is shown, circles and bars represent results of individual mice (n=6) and means, respectively.

(7) FIG. 7 Anti-ovalbumin responses after several virus administrations. C57Bl/6 mice were injected four times at two week intervals intraperitoneally with 10.sup.10 vp of AdWT or capsid-modified Ad and sera were collected at day 14 after each injection. Ab titers below 100 were plotted as 50. One of two experiments is shown, circles and bars represent results of individual mice (n=6) and means, respectively.

(8) FIG. 8 Anti-ovalbumin responses after subcutaneous administration of capsid-modified Ad. C57Bl/6 mice were injected subcutaneously twice two weeks apart with 10.sup.10 vp of AdWT or capsid-modified Ad and sera were collected at day 14 p.i. Ab titers below 100 were plotted as 50. One of two experiments is shown, circles and bars represent results of individual mice (n=6) and means, respectively.

(9) FIG. 9 Influence of anti-Ad immunity on the anti-ovalbumin humoral response. (A) C57Bl/6 mice injected with AdWT or mock-injected were injected two weeks later with AdH-3OVA2 or AdF-3OVA2 (10.sup.10 vp). Anti-ovalbumin IgG titers were determined by ELISA at day 14 p.i. Ab titers below 100 were plotted as 50. One of two experiments is shown, circles and bars represent results of individual mice (n=6) and means, respectively. Splenocytes (B), CD4.sup.+ (C), CD8.sup.+ (D) lymphocytes or serum (E) from mock-injected mice or from mice injected with AdWT were adoptively transferred into naive mice before intraperitoneal injection of AdH-3OVA2 or AdF-3OVA2 (10.sup.10 vp). Anti-ovalbumin IgG titers were determined by ELISA at day 14 p.i. Ab titers below 100 were plotted as 50. Circles and bars represent results of individual mice (n=6) and means, respectively.

(10) FIG. 10 Neutralization of capsid-modified Ad by anti-Ad Abs. AdWT or capsid-modified viruses (AdH-3OVA2 or AdF-3OVA2) were mixed with serial dilutions of anti-Ad serum and then incubated with 293A cells. ?gal activity was measured one day later and expressed as percentage of transduction relative to cells infected with virus alone.

(11) FIG. 11 Masking of 3OVA2 epitope by anti-Ad Abs. AdWT, AdH-3OVA2 or AdF-3OVA2 coated on 96-well plates were incubated with serum from naive or Ad-immune mice and 3OVA2 epitope accessibility was assessed using anti-ovalbumin Abs. 3OVA2 epitope detection on virions in different conditions of incubation was expressed relative to detection of 3OVA2 on virus incubated with control serum. One of two experiments is shown, histograms represent means?SD (n=10).

(12) FIG. 12. Comparison of capsid-modified Ad to a recombinant Ad expressing ovalbumin as a transgene. C57BL/6 mice were immunized twice two weeks apart intra-peritoneally with 10.sup.10 vp of capsid-modified Ad (AdH-3OVA2 or AdF-3OVA2) or AdWT-Ovalb, expressing ovalbumin as a transgene. Sera were collected at day 14 p.i. and anti-ovalbumin IgG titers were determined by ELISA. Ab titers below 100 were plotted as 50. One of two experiments is shown, circles and bars represent results of individual mice (n=6) and means, respectively.

(13) FIG. 13. Anti-ovalbumin responses after priming and boosting with different capsid-modified Ad. C57BL/6 mice were immunised twice two weeks a part intra-peritoneally with 10.sup.10 vp of capsid-modified Ad (AdH-3OVA2 or AdF-3OVA2) or were immunised first with AdH-3OVA2 and second with AdF-3OVA2. Sera were collected at day 14 p.i. and anti-ovalbumin IgG titers were determined by ELISA. Ab titers below 100 were plotted as 50. One of two experiments is shown, circles and bars represent results of individual mice (n=6) and means, respectively.

MATERIALS AND METHODS

(14) Adenoviral Vectors

(15) LacZ-recombinant control AdWT (AE18 in ref. (VIGNE et al., J Virol, 73, 5156-61, 1999)) was derived from Ad5 with E1 and E3 regions deleted. AdH-OVA and AdH-3OVA2, derived from AdWT, contain respectively a short (OVA.sub.323-339: ISQAVHAAHAEINEAGR (SEQ ID NO: 1), referred in the text as OVA peptide) or a long (OVA.sub.320-342: SLKISQAVHAAHAEINEAGREV (SEQ ID NO: 2), referred in the text as 3OVA2 peptide) peptide derived from ovalbumin. These peptides were inserted into the hexon protein in place of .sub.269TTEAAAGNGDNLT.sub.281 (SEQ ID NO: 3) of hypervariable region 5. AdF-OVA and AdF-3OVA2 contain respectively OVA or 3OVA2 peptide inserted in place of T.sub.539QETGDTTPS.sub.548 (SEQ ID NO: 4) into the HI loop of the fiber protein. All modified viruses (Table 1) were constructed using recombinational cloning in E. coli (CROUZET et al., Proc Natl Acad Sci USA, 94, 1414-9, 1997).

(16) TABLE-US-00001 TABLE1 Modified Insertedpeptide Titer Virus protein Adaptor (SEQIDNO:#) Adaptor (?10.sup.12vp/ml) AdWT 7.5? 0.6 AdH-OVA hexon G ISQAVHAAHAEINEAGR(1) LGG 5.9 AdH-3OVA2 hexon G SLKISQAVHAAHAEINEAGREV(2) LGG 7.9? 1.4 AdF-OVA fiber SS ISQAVHAAHAEINEAGR(1) GSS 9.5 AdF-3OVA2 fiber SS SLKISQAVHAAHAEINEAGREV(2) GSS 8.1? 0.4

(17) All viruses were obtained using standard procedures as described (MARTIN et al., Mol Ther, 8, 485-94, 2003), stored at ?80? C. in PBS-7% glycerol and titrated by spectrophotometry (1 OD.sub.260=1.1?10.sup.12 viral particle (vp)/ml).

(18) Oligonucleotides

(19) Oligonucleotides used for insertion and PCR detection of nucleotide sequences encoding 3OVA2 and OVA epitopes are described in Table 2.

(20) TABLE-US-00002 TABLE2 Name Nucleotidesequence(SEQIDNO:#) Target Hex 5-atgggatgaagctgctactg-3 (5) Hexon IM21A 5-ggcatatctcaagctgtccatgcagcacatgcagaaatcaatgaagcaggcagacttggcggccc- OVA 3 (6) into hexon IM21B 5-ttagggccgccaagtctgcctgcttcattgatttctgcatgtgctgcatggacagcttgagatatgcc- OVA 3 (7) into hexon IM22A 5-ggcagcctgaagatatctcaagctgtccatgcagcacatgcagaaatcaatgaagcaggcagagaggtgc 3OVA2 ttggcggccc-3 (8) into hexon IM22B 5-ttagggccgccaagcacctctctgcctgcttcattgatttctgcatgtgctgcatggacagcttgagat 3OVA2 atcttcaggctgcc-3 (9) into hexon HigU1 5-cagctccatctcctaactgtagactaaatg-3 (10) Fiber IM23A 5-gtaaccctaaccattacactaaacggttctagcatatctcaagctgtccatgcagcacatgcagaaatca OVA atgaagcaggagc-3 (11) into fiber IM23B 5-gctagaacctctgcctgcttcattgatttctgcatgtgctgcatggacagcttgagatatgctagaaccg OVA tttagtgtaatggttagg-3 (12) into fiber IM24A 5-gtaaccctaaccattacactaaacggttctagcagcctgaagatatctcaagctgtccatgcagcacatg 3OVA2 cagaaatcaatgaagcaggcagagaggtgggttctagc-3 (13) into fiber IM24B 5-gctagaacccacctctctgcctgcttcattgatttctgcatgtgctgcatggacagcttgagatatcttc 3OVA2 aggctgctagaaccgtttagtgtaatggttagg-3 (14) into fiber
PCR Detection of Epitope Coding Sequences

(21) Adenoviral DNA was extracted from purified AdWT and capsid-modified viruses and DNA concentrations were determined by OD260. The primers used in PCR reactions (Table 2) are as follows: Hex and IM21B for AdH-OVA, Hex and IM22B for AdH-3OVA2, HigU1 and IM23B for AdF-OVA, HigU1 and IM24B for AdF-3OVA2. The amplification mixture contained 200 ?M dNTPs, 0.5 ?M of each primer, 1.5 U of Taq Polymerase (Biolabs, Ipswich, Mass.), 1? TaqPol buffer, and 150 ng of total DNA. The reaction was initiated by a 4 min denaturation step at 94? C. Amplification occurred during 30 cycles, each cycle consisting of 1 min at 94? C., 1 min at 53? C., and 2 min at 72? C. PCR products were analyzed by gel electrophoresis.

(22) SDS-PAGE and Silver Stain Analysis

(23) Purified viruses (10.sup.10 vp) were resuspended in Laemmli lysis buffer, boiled for 5 min and loaded onto a 10% NuPage gel (Novex, Invitrogen, CA). After electrophoresis, the gel was stained with a silver staining kit (Invitrogen, Carlsbad, Calif.).

(24) Detection of OVA- and 3OVA2-Epitopes on Virions

(25) To assess whether the epitopes were present or accessible on the capsid surface, denaturated or native viruses were coated on 96-well plates (Nunc, Roskilde, Denmark). Viruses were inactivated at 56? C. during 30 minutes and 0.1% SDS was added. Spectrophotometric reading was performed at 215 and 225 nm to determine virus concentrations and 100 ng were coated on 96-well plates. Alternatively, same quantities of native viruses were coated. After overnight incubation at 4? C., non-specific sites were blocked with 5% milk PBS-Tween, then plates were washed and incubated with a rabbit polyclonal antibody against the ovalbumin protein (AB1225, Millipore, Mass.) for 1 hour. After washing, an anti-rabbit IgG peroxydase-linked Ab (NA934, Amersham Biosciences, Saclay, France) was added for 1 hour and peroxidase activity was revealed by incubation with the substrate o-Phenylenediamine dihydrochloride (Sigma-Aldrich, Lyon, France) for 30 min. The reaction was stopped by addition of 3N HCl and spectrophotometric reading was performed at 490 nm.

(26) The masking of ovalbumin-derived epitopes by anti-Ad Abs was investigated as follows. Viruses (10.sup.9 vp) were coated on 96-well plates (Nunc) and dilutions of serum from mice injected twice with AdWT were added. Sera from pre-immune mice were used as controls. Following a one hour incubation, plates were washed, and incubated with a rabbit polyclonal anti-ovalbumin Ab (AB1225) overnight at 4? C. The binding of anti-ovalbumin Ab was revealed as described above.

(27) Cells

(28) 293A were maintained as recommended by Invitrogen. CHO-CAR, kindly provided by Dr Bergelson, were described previously (BERGELSON et al., Science, 275, 1320-3, 1997).

(29) CHO-CAR Transduction

(30) All the experiments were performed in 12-well dishes (Corning Glass Works, Corning, N.Y.) and run in duplicate. Confluent cell monolayers of CHO-CAR cells were infected with increasing MOI (10, 10.sup.2 and 10.sup.3 vp/cell) of AdWT or capsid-modified viruses in 400 ?l of serum-free medium. One hour later, 2 ml of complete medium were added. After 24 h, cells were lysed and ?-galactosidase (?-gal) activity was measured using a chemiluminescent assay (BD Biosciences Clontech, Palo Alto, Calif.). Protein content was determined using the Bio-Rad Protein Assay (Bio-Rad Laboratories, Hercules, Marnes-la-Coquette, France). Results are expressed as relative light units (RLU) per ?g of protein.

(31) Mice

(32) 7-week-old C57BL/6 female mice were purchased from Janvier (Le Gesnest Saint Isle, France) or Harlan (Gannat, France). They were conditioned at least for one week in the animal facilities before beginning of the experiments. All animal experiments were approved by the IGR Institutional Animal Care and Use Committee.

(33) In Vivo Experiments

(34) Control or capsid-modified viruses (10.sup.9 or 10.sup.10 vp) in PBS (200 ?L) were injected intra-peritoneally. Repeated injections were performed at intervals of 2 weeks (total injection between 2 and 4). Blood samples were collected before virus injection and at different intervals thereafter. Mice sera were prepared and analyzed for the presence of anti-ovalbumin, anti-?galactosidase (?gal) and anti-Ad antibodies by ELISA as described below.

(35) Serum and Cell Transfer Experiments

(36) For sera and splenocytes transfer experiments, mice were injected intra-peritoneally twice two weeks apart with AdWT (10.sup.10 vp in 200 ?L of PBS) or PBS (200 ?L). At day 15 after the second injection, mice were sacrificed and spleens were removed. Spleens were crushed in RPMI medium supplemented with 5% SVF, 1% non essentials amino acids, 1% glutamine, 1% pyruvate and 5?10.sup.?5 M ?-mercaptoethanol, and filtered through a 100 ?m cell strainer. After removal of blood cells by ACK Lysing Buffer (Invitrogen, Cergy-Pontoise, France), the cells were resuspended and the concentration was adjusted at 5.Math.10.sup.7 cells/mL. In parallel, sera were prepared from blood. Serum (diluted to the third in PBS) or splenocytes (10.sup.7 splenocytes in 200 ?L of RPMI medium) were injected intravenously into mice retro-orbital plexus. The following day, mice were injected intra-peritoneally with either AdH-3OVA2 or AdF-3OVA2 (10.sup.10 vp). Blood were collected at day 15 after Ad injection. Mice sera were prepared and analyzed for the presence of anti-ovalbumin, anti-?gal and anti-Ad antibodies by ELISA as described below.

(37) For CD4.sup.+ and CD8.sup.+ T cells transfer experiments, splenocytes were prepared as described above and resuspended in 1 mL of complemented RPMI medium. Then, PE-Cy7 anti-CD3 (eBioscience, CA, USA), APC anti-CD19 (eBioscience) and APC anti-NK-1.1 (BD Bioscience) Abs were added during 30 minutes at 4? C. After washing, cells were resuspended, filtered through a 40 ?m cell stainer and 7-AAD (BD Bioscience) was added. CD3.sup.+ cells were sorted on a DakoCytomation MoFlo cytometer (Beckman Coulter, Villepinte, France) and collected. After centrifugation, cells were resuspended in 1 mL of complemented RPMI medium and incubated with PE anti-CD4 and FITC anti-CD8 Abs (BD Bioscience) during 30 minutes at 4? C. After cells sorting, two fractions corresponding to CD4.sup.+CD3.sup.+ and CD8.sup.+CD3.sup.+ T cells were obtained. CD4.sup.+ (2.Math.10.sup.6) or CD8.sup.+ (10.sup.6) T cells in 200 ?L of RPMI medium were injected intravenously into mice retro-orbital plexus. The following day, mice were injected intra-peritoneally either with AdH-3OVA2 or with AdF-3OVA2 (10.sup.10 vp/mouse). Blood were collected at day 15 after Ad injection,

(38) Determination of Specific Antibodies

(39) Ovalbumin-specific antibodies in the sera were determined by ELISA. After coating of 96-well plates (Nunc) with 1 ?g of ovalbumin protein (Calbiochem, Merck chemicals, Nottingham, England), serial dilutions of the sera in 5% milk PBS-Tween were added. Bound antibody was detected with peroxidase-conjugated anti-mouse IgG, IgG1, IgG2a or IgG2b isotypes goat antibodies (Southern Biotechnology Associates, Birmingham, Ala.). The peroxidase was revealed by incubation with the substrate o-Phenylenediamine dihydrochloride (Sigma-Aldrich) for 30 min. The reaction was stopped by addition of 3N HCl and spectrophotometric reading was performed at 490 nm. ?gal- and Ad-specific antibodies in the sera were determined by ELISA as described previously (BENIHOUD et al., Gene Ther, 14, 533-44, 2007). Titers were calculated as reciprocal dilutions 2-fold above background values.

(40) Anti-Adenoviral Neutralizing Antibody Assay

(41) LacZ-recombinant wild-type or capsid-modified viruses were mixed with serial dilutions of serum samples from pre-immune or Ad-immune mice decomplemented for 30 min at 56? C. Then, after a one-hour incubation at 37? C., the mixture was incubated with 293A cells (40,000 cells/well, multiplicity of infection of 50) in 96-well plates for 1 h at 37? C. Then, 100 ?l of complete medium was added, and the cells were cultured for 19 h. Cells were washed and incubated with 100 ?l of lysis buffer (6 mM Na2HPO4, 10 mM KCl, 0.1 mM MgSO4, 50 mM 2-mercaptoethanol, and 0.5% Triton X-100) containing 0.1 mg of ?gal substrate (4-methyl-umbelliferylb-D-galactoside; Sigma) for 30 min at 37? C. After excitation at 360 nm, the resulting fluorescence was measured at 460 nm with a Wallach Victor 2 (Perkin-Elmer, Waltham, Mass.). For each serum dilution, the percentage of transduction was calculated as follows: (experimental value?background value [without virus])/(positive control [without serum]?background value)?100.

(42) Statistical Analyses

(43) A Mann-Whitney test, recommended for groups fewer than 30 mice, was conducted. Differences were considered significant when P<0.05.

(44) Results

(45) Production and Characterization of Ad Displaying Ovalbumin-Derived Epitopes

(46) In order to analyze whether the site of epitope insertion may influence vaccination efficacy of Ad displaying epitopes on their capsid, ovalbumin-derived epitopes (OVA and 3OVA2) were genetically inserted into either hexon (AdH-OVA and AdH-3OVA2) or fiber (AdF-OVA and AdF-3OVA2) proteins of a IacZ recombinant Ad. All capsid-modified Ad were produced and purified at titers comparable to AdWT presenting a wild-type capsid (Table 1). SDS-PAGE analyses confirmed no differences in virus composition and integrity between capsid-modified Ad and AdWT (FIG. 1A). In addition, there was no difference in the ability of capsid-modified Ad to tranduce CHO-CAR cells, a cell line overexpressing Ad primary receptor (FIG. 2).

(47) The presence of OVA or 3OVA2 epitope coding sequence in Ad genome was confirmed by PCR performed on purified virions (data not shown). For AdF-OVA and AdF-3OVA2, the presence of the epitope within the fiber was demonstrated by a modification of the fiber migration pattern in SDS-Page (FIG. 1A). Using a polyclonal anti-ovalbumin antibody, OVA or 3OVA2 epitope were detected by ELISA performed on native (FIG. 18) or denaturated (FIG. 1C) purified virions. 3OVA2 epitope was detected on native or denaturated AdH-3OVA2 and AdF-3OVA2. OVA epitope was detected on either native or denaturated AdH-OVA but only on native AdF-OVA (FIG. 18).

(48) Anti-Ovalbumin Humoral Responses Triggered by One Injection of Capsid-Modified Adenovirus

(49) To assess the ability of capsid-modified Ad to mount an anti-ovalbumin humoral response, C57Bl/6 mice were injected intra-peritoneally with AdWT or capsid-modified Ad (10.sup.10 viral particle (vp)). Sera were collected until 68 days post-injection (p.i.) and anti-ovalbumin IgG Ab titers were measured. Most of mice injected with capsid-modified Ad displayed anti-ovalbumin Abs whereas AdWT-injected mice did not. Production of Abs was detected as soon as day 7 p.i., peaked between day 14 and day 28 p.i. and was still detectable 68 days p.i. for all capsid-modified Ad (FIG. 3). Interestingly, at all time points, anti-ovalbumin Ab titers were higher when the epitope (OVA or 3OVA2) was inserted into hexon rather than into fiber protein. Thus, AdH-3OVA2-injected mice displayed much higher anti-ovalbumin Ab titers than AdF-3OVA2-injected mice at all time points (p<0.05 for all the kinetic, and notably, p<0.01 at day 14 and 28). Anti-ovalbumin titers were also stronger at day 14 p.i. in sera of AdH-OVA-injected mice than AdF-OVA (FIG. 3, p<0.05). Careful analysis of IgG subisotypes revealed a high production level of IgG2a and IgG2b anti-ovalbumin Abs, reminiscent of the Th1 bias of anti-Ad humoral responses (BENIHOUD et al., J Virol, 72, 9514-25, 1998). Interestingly, AdH-3OVA2-injected mice displayed significantly higher titers of anti-ovalbumin IgG2a and IgG2b subisotypes compared to AdF-3OVA-2-injected mice (FIG. 4, p<0.01).

(50) To better compare the efficiency of vaccination by hexon- or fiber-modified Ad, mice were injected with the two viruses giving the best humoral responses (AdH-3OVA2 or AdF-3OVA2) at the dose of 10.sup.10 or 10.sup.9 vp and anti-ovalbumin antibodies were measured at day 14 p.i. Anti-ovalbumin antibody levels increase with viral dose for both AdH-3OVA2 and AdF-3OVA2, however whatever the dose examined they were higher in AdH-3OVA2-injected mice than in AdF-3OVA2 (FIG. 5). Anti-ovalbumin titers were comparable in mice treated with 10.sup.9 vp of AdH-3OVA2 and in mice treated with 10.sup.10 vp of AdF-3OVA2, thus underlining a 10-fold better efficacy of AdH-3OVA2 compared to AdF-3OVA2 (FIG. 5).

(51) Altogether, these results pointed out that following one capsid-modified Ad injection, a better humoral response is obtained when the epitope is inserted into hexon protein.

(52) Anti-Ad Pre-Existing Immunity Shapes Anti-Ovalbumin Humoral Responses Induced by Capsid-Modified Ad

(53) To determine to what extent anti-ovalbumin humoral responses could be boosted, C57BL/6 mice were injected intra-peritoneally at days 0 and 14 with AdWT or different capsid-modified Ad. Sera were collected at day 14 after the first and the second injection. Compared to mice receiving one virus administration, anti-ovalbumin Ab titers present an 11- and 1.5-fold increase in mice injected twice with AdH-OVA and AdH-3OVA2, respectively (FIG. 6). In sharp contrast, after the second injection, AdF-OVA and AdF-3OVA2-injected mice exhibit a 314- and 110-fold increase in anti-ovalbumin Ab titer, respectively, compared to the ones observed after the first viral administration (FIG. 6). This better efficacy of fiber-modified Ads was still observed following 4 intraperitoneal virus injections (FIG. 7) and remarkably was also found using another mode (subcutaneous) of virus administration (FIG. 8). It should be emphasized that such a dramatic increase in humoral responses after the second injection of fiber-modified Ad was only observed for anti-ovalbumin Abs. Indeed, anti-?gal and anti-Ad Abs were not differentially increased after challenge with fiber-modified Ad compared to hexon-modified Ads or AdWT (FIG. 6).

(54) Our results stress that after the second virus administration, a most powerful humoral response is obtained with epitope insertion into fiber protein, thus reversing the hierarchy observed after one virus injection (FIGS. 3 to 5). Since we ruled out a difference in the kinetic of anti-ovalbumin Ab responses between hexon and fiber-modified Ad (FIG. 3), we examined a potential role of anti-Ad immunity in controlling anti-ovalbumin Abs levels. Mice were either injected with PBS or with AdWT (10.sup.10 vp), and then two weeks later, mice received either AdH-3OVA2 or AdF-3OVA2 injection. Mice injected with AdWT followed by AdH-3OVA2 displayed a 21-fold decrease in serum anti-ovalbumin Ab titers compared to mice injected with PBS followed by AdH-3OVA2 (FIG. 9A), thus suggesting that pre-existing Ad immunity dampens the capacity of the immune system to mount a humoral response against ovalbumin epitope inserted into hexon protein. In contrast, mice injected successively with AdWT and AdF-3OVA2 displayed a 55-fold increase in serum Ab titers compared to mice injected with PBS and AdF-3OVA2 (FIG. 9A).

(55) Anti-Ad Humoral but not Cellular Responses Potentiate Anti-Ovalbumin Humoral Responses Induced by Fiber-Modified Ad

(56) To examine how the immune response to Ad influences on anti-ovalbumin humoral responses, we studied the respective roles of anti-Ad lymphocytes and antibodies. First, we analysed anti-ovalbumin humoral responses in mice adoptively transferred with na?ve or anti-Ad splenocytes before administration of capsid-modified Ad. FIG. 9B indicated that compared to their na?ve counterparts, splenocytes from Ad-immune mice dramatically reduced anti-ovalbumin response in mice injected with AdH-3OVA2 but potentiated this response in AdF-3OVA2-injected mice (FIG. 9B). To further define the role of anti-Ad cellular immunity, CD4.sup.+ or CD8.sup.+ T lymphocytes from mice injected with PBS or AdWT were sorted and adoptively transferred into na?ve recipients before administration of AdH-3OVA2 or AdF-3OVA2. FIG. 9D and FIG. 9E ruled out a major influence of anti-Ad CD4.sup.+ or CD8.sup.+ lymphocytes in mounting anti-ovalbumin Ab responses. In sharp contrast, transfer of serum from Ad-immune mice into naive mice before administration of AdH-3OVA2 or AdF-3OVA2 led either to a strong inhibition or to a dramatic increase in anti-ovalbumin Ab responses, respectively.

(57) Altogether, these results underline that humoral anti-Ad pre-existing immunity strongly shapes the humoral response against an ovalbumin-derived epitope inserted into Ad capsid. Most importantly, our data underline that, in Ad immune mice, fiber constitutes the best site of peptide insertion to mount an efficient anti-epitope humoral response.

(58) Masking of Ovalbumin-Derived Epitope Inserted into Hexon Protein by Anti-Ad Antibodies

(59) In order to understand how anti-Ad Abs affect anti-ovalbumin humoral responses, we first examined whether there was a difference in sensitivity of AdH-3OVA2 or AdF-3OVA2 to antibodies raised against AdWT. First, AdH-3OVA2, AdF-3OVA2 and AdWT were incubated without or with different anti-Ad serum dilutions, then, their ability to transduce CAR-expressing cells was analysed. The results indicated for all viruses a comparable inhibition of cell transduction by anti-Ad antibodies as documented by measurement of ?gal activity (FIG. 10). Thus, the difference in mounting anti-ovalbumin Ab responses between hexon-modified and fiber-modified Ads was not linked to a differential neutralization by anti-Ad antibodies.

(60) As an alternative hypothesis, we postulated that anti-Ad Abs could induce a steric hindrance and inhibit in vivo the recognition of ovalbumin epitope by B cells, leading to a decrease of anti-ovalbumin humoral responses in AdH-3OVA2-injected mice. This hypothesis is supported by previous reports showing that hexon protein was the main target of anti-Ad antibodies (ROBERTS et al., Nature, 441, 239-43, 2006). To get insight into ovalbumin epitope accessibility, capsid-modified Ad were immobilized on ELISA plates and incubated with serum from naive or AdWT-injected mice, then, 3OVA2 epitope was detected by an anti-ovalbumin polyclonal Ab. FIG. 11 showed an anti-Ad-Ab-mediated inhibition of 3OVA2 epitope detection on both AdH-3OVA2 and AdF-3OVA2. However, it should be emphasized that detection of ovalbumin epitope was strongly reduced at high serum concentration for AdH-3OVA2 compared to AdF-3OVA2, thus suggesting that anti-Ad Abs reduced accessibility of the epitope when inserted into hexon protein.

(61) Vaccine Efficiency of Hexon Modified or Fiber-Modified Ads Compared to Recombinant Ad Expressing the Antigen of Interest.

(62) To compare the vaccine efficiency of the epitope display strategy to the classical approach based on Ad driving the expression of the antigen of interest, C57BL/6 mice were injected two times either with capsid-modified Ad (AdH-3OVA2 or AdF-3OVA2) or with AdWT-Ovalb, bearing an expression cassette for ovalbumin protein. After the first injection, anti-ovalbumin Ab titers obtained were higher for mice injected with AdWT-Ovalb compared to AdH-3OVA2 or AdF-3OVA2. However, after the second virus administration, a comparable anti-ovalbumin humoral response was obtained in AdF-3OVA2 and AdWT-Ovalb-injected mice (FIG. 12), thus underlying the potency of the epitope display strategy of vaccination after two virus injections.

(63) Vaccine Efficiency after Priming with Hexon-Modified Ad and Boosting with Either Hexon-Modified or Fiber-Modified Ad.

(64) We analysed the anti-ovalbumin humoral response induced by a first administration of AdH-3OVA2 followed by a second administration of AdF-3OVA2, to compare it with those induced by two consecutive administrations of either AdH-3OVA2 or AdF-3OVA2. FIG. 13 shows that mice injected with AdH-3OVA2 followed by AdF-3OVA2 present after the second administration a level of anti-ovalbumin Ab response comparable to mice injected two times with AdF-3OVA2, and superior to mice injected two times with AdH-3OVA2.

(65) Discussion

(66) Our results indicated that after one Ad injection, the strongest anti-ovalbumin humoral responses were triggered when the epitope was inserted into hexon protein, thus underlining an important role of epitope number (240?3 for hexon-modified Ad versus 12?3 for fiber-modified Ads). In sharp contrast, after two or more injections, a remarkable increase in anti-ovalbumin humoral responses was obtained when the epitope was inserted into fiber protein. In addition, our data unravelled a role of anti-Ad immunity in controlling anti-epitope humoral responses.

(67) To study the influence of the insertion site on epitope immunogenicity, OVA.sub.323-339 alone (OVA) or surrounded by residues flanking this epitope in ovalbumin protein (3OVA2) was inserted into hexon or fiber protein. In vitro studies showed that this epitope was better recognized by polyclonal anti-ovalbumin Abs on AdH-3OVA2 and AdF-3OVA2 rather than on AdH-OVA and AdF-OVA. This suggests that OVA.sub.323-339 accessibility or conformation is improved by addition of flanking residues. After virion denaturation, OVA.sub.323-339 epitope was better detected on AdH-3OVA2 versus AdF-3OVA2 and on AdH-OVA versus AdF-OVA in strict correlation with the number of epitopes per capsid. However, the results were strikingly different for native virions. Indeed, OVA.sub.323-339 is better detected on AdF-3OVA2 than AdH-3OVA2, thus suggesting that fiber protein naturally protruding from the virion is a more suitable site for display of a B cell epitope.

(68) After one injection of capsid-modified Ad into C57Bl/6 mice, a higher anti-ovalbumin humoral response was obtained with AdH-OVA and AdH-3OVA2 compared to AdF-OVA and AdF-3OVA2. This hierarchy between hexon and fiber modified-Ads is maintained throughout the kinetic up to 68 days p.i. Moreover, a dose of 10.sup.9 vp of AdH-3OVA2 led to similar levels of anti-ovalbumin Abs than 10.sup.10 vp of AdF-3OVA2, demonstrating a 10-fold better efficacy of vaccination by AdH-3OVA2. These results were surprising given the better OVA.sub.323-339 accessibility into fiber protein. However, they could be linked to the number of epitopes per capsid (240?3 versus 12?3) and suggest a direct role of epitope number on the vaccination efficiency.

(69) Following the second injection, a boost of anti-ovalbumin Abs was observed in all mice groups. However, whereas hexon-modified Ads led only to a modest increase in anti-ovalbumin Ab responses, fiber-modified Ads triggered a dramatic increase. As a consequence, the second injection of capsid-modified Ads reverses the hierarchy observed after one injection, with fiber-modified Ads triggering the highest anti-ovalbumin Ab responses. Interestingly, this bias is maintained even after up to 4 injections. The boosting efficacy of anti-ovalbumin humoral responses by fiber-modified Ads was observed not only after intraperitoneal but also after subcutaneous Ad injection and thus was not dependent on the mode of virion administration. Remarkably, this boosting efficacy appears to be specific of the epitope inserted into the capsid since no difference in the degree of boosting of anti-?gal or anti-Ad Abs was observed between all viruses. To the best of our knowledge, this is the first report showing a difference in vaccination efficiency among capsid-modified Ads upon reinjection.

(70) The dramatic increase in anti-ovalbumin Ab responses observed after the second injection of fiber-modified Ads could not be explained by different kinetic compared to hexon-modified Ads. Therefore, we hypothesized a role of Ad immunity in shaping anti-epitope Ab responses. In fact, Ad-immune mice injected with AdF-3OVA2 displayed much higher anti-ovalbumin Ab levels than naive mice injected with AdF-3OVA2. To identify which component of anti-Ad immunity plays a role in potentializing anti-ovalbumin Ab responses, mice were injected with total splenocytes, purified CD4.sup.+ or CD8.sup.+ T cells or serum from Ad-immune mice and then challenged with AdF-3OVA2. Anti-ovalbumin humoral responses were boosted by anti-Ad serum or splenocytes but not by anti-Ad CD4.sup.+ or CD8.sup.+ T cells. From these results, we deduced that serum anti-Ad Abs were responsible for the enhanced anti-epitope humoral responses. In contrast, a role of anti-Ad cellular immunity was ruled out because anti-Ad CD4.sup.+ or CD8.sup.+ T cells did not modify anti-epitope humoral responses. Of note, the apparent contradiction with the boosting induced by total splenocytes could be resolved by the presence within splenocytes of plasma cells able to produce significant levels of anti-Ad Abs.

(71) In sharp contrast to fiber-modified Ads, hexon-modified Ads reinjection led to a modest boost of anti-ovalbumin response. Moreover, anti-ovalbumin responses triggered by hexon-modified Ads were reduced in Ad-immune mice as compared to naive mice, thus suggesting that anti-Ad immune responses limit the efficiency of vaccination against the epitope embedded within hexon protein. This is in agreement with serum transfer experiments showing reduced anti-ovalbumin Ab levels in mice receiving anti-Ad serum before injection of hexon-modified Ads. Altogether, these results demonstrate that anti-Ad Abs impair the induction of antibody responses against an ovalbumin epitope inserted into hexon protein. As discussed by Getahun et al. for other antigens, two hypotheses, antigen clearance or epitope masking could explain Ab-mediated reduction of anti-epitope humoral responses (GETAHUN & HEYMAN, Scand J Immunol, 70, 277-87, 2009). The first hypothesis was invalidated since there was no difference in anti-Ad antibody induction and sensitivity between fiber and hexon-modified Ads. Thus, in contrast to previous data, deletion of hexon HVR5 does not modify significantly sensitivity to Ad Abs (ABE et al., J Gene Med, 11, 570-9, 2009). The second hypothesis, epitope masking by anti-Ad Abs, is strongly supported by the observation that epitope detection within hexon protein was inhibited in the presence of anti-Ad Abs. Such an influence of these Abs is not surprising given the fact that hexon protein is the major target of Ad neutralizing Abs (SUMIDA et al., J Immunol, 174, 7179-85, 2005; ROBERTS et al., Nature, 441, 239-43, 2006). Furthermore, since most of these Abs are directed against hexon HVRs, they may limit by steric hindrance epitope recognition by B cells.

(72) It is also to be noted that following two administrations, anti-ovalbumin responses triggered by fiber-modified Ad were shown to be comparable to those induced by a recombinant Ad expressing ovalbumin as a transgene. This is noteworthy, given that the latter virus encodes the complete ovalbumin protein and not only 3OVA2 epitope.

(73) The present study shades a new light on the determinants controlling the recently developed vaccination strategy based on epitope display by Ad. Indeed, we unravelled an unexpected role of anti-Ad Abs in controlling the vaccination efficacy against a capsid-embedded epitope. Depending on the site of epitope insertion, anti-Ad Abs were shown either to dramatically enhance (fiber insertion) or to reduce (hexon insertion) the humoral response against the epitope. A major novelty of our findings is to provide clues for choosing the best Ad-based vaccine to be used in clinic. Hence, one can anticipate that hexon-modified Ads should be more efficient for triggering anti-epitope responses in Ad-na?ve patients. On the contrary, a better efficiency of Ad displaying epitope into fiber protein is expected in Ad5 seropositive patients or in case of repeated injections.