Tat complexes, and vaccines comprising them

Abstract

Complexes comprising HIV Tat and the V3 loop from gp120 Env provide novel epitopes and are immunogenic to prevent or inhibit infection by HIV.

Claims

1. A complex comprising a first peptide and a second peptide bound thereto, the first peptide comprising the V3 loop of gp120, wherein the V3 loop is exposed and thereby bound to a binding region on the second peptide to form the complex, the second peptide comprising said binding region which region comprises at least residues 21-40 and 46-58 of Tat (SEQ ID NO 1), or at least said residues with a further point mutation whereby Cys 22 of Tat is replaced by Glycine to produce a Tat Cys22 mutant, wherein the first peptide comprising the V3 loop is selected from the group consisting of: an isolated V3 loop peptide of gp120; a peptide consisting of residues 299-333 of SEQ ID NO. 2; and the trimeric gp140 form of Env, retaining part of gp41 and lacking the V2 loop of Env, and optionally wherein the complex further comprises a molecule or substance capable of interacting with Env to expose a functional V3 loop or the peptides are cross-linked.

2. The complex of claim 1, the binding region on the second peptide being derived from Tat and being recognisable by a monoclonal antibody directed against the CCR5 second extracellular loop.

3. The complex of claim 1, wherein the binding region comprises at least residues 21-58 of Tat as set forth in SEQ ID NO 1, or a TatCys22 mutant thereof.

4. The complex of claim 1, prepared with native Tat.

5. The complex of claim 1, wherein the peptide comprising the V3 loop comprises some of gp120 in addition to the V3 loop.

6. The complex of claim 1, wherein the first peptide consists of the V3 loop region of gp120.

7. The complex of claim 1, wherein the first peptide is the trimeric gp140 form of Env, retaining part of gp41 and lacking the V2 loop of Env.

8. The complex of claim 1, further comprising a molecule or substance capable of interacting with Env to expose a functional V3 loop, wherein said molecule or substance is CD4 and interacts with Env to expose a functional V3 loop.

9. The complex of claim 1, further comprising a heparin sulphate.

10. The complex of claim 1, further comprising a substance selected from the group consisting of integrins, basic fibroblast growth factor, CD26, VEGF receptors, and chemokine receptors.

11. The complex of claim 1, wherein the binding region is contained within a fragment of Tat generatable by proteasomes of human cells on exposure to Tat, wherein the Tat fragment is selected from the group consisting of fragments containing the cysteine, basic and RGD regions of Tat; fragments containing the cysteine and basic regions of Tat; fragments containing the basic and RGD region of Tat; and, fragments containing the basic region of Tat, alone.

12. The complex of claim 1, wherein said peptides are cross-linked.

13. A vehicle suitable for injection comprising the complex of claim 1.

14. A kit comprising at least two separate preparations of the components of the complex of claim 1.

15. A method for inducing antibodies to HIV which comprises administering the complex of claim 1 to a patient in need thereof.

16. A method for inducing antibodies to HIV in a patient in need, whereby the HIV expresses a molecule capable of forming a ternary complex between said molecule, CD4 and CCR5, comprising administering the complex of claim 1 to the patient.

17. A method to establish whether a sample from a patient contains antibodies against the complex of claim 1 which comprises contacting said complex with said sample and detecting the presence of antibodies bound to said complex.

18. A complex according to claim 1, wherein the second peptide comprises at least residues 21-60 of Tat (SEQ ID NO: 1) with a further point mutation whereby a residue corresponding to Cys 22 of Tat in SEQ ID NO: 1 is replaced by Glycine.

19. An immunogenic complex comprising a first peptide and a second peptide bound thereto, the first peptide comprising, at least the V3 loop of gp120 and lacks at least the majority of the V2 loop of gp120, wherein the first peptide forms a trimer, and wherein the V3 loop is exposed and thereby bound to a binding region on the second peptide to form the complex, MD the second peptide comprising said binding region which region comprises at least residues 21-40 and 46-58 of Tat (SEQ ID NO 1), or at least said residues with a further point mutation whereby Cys 22 of Tat is replaced by Glycine to produce a TatCys22 mutant; wherein the complex further comprises a molecule or substance capable of interacting with Env to expose a functional V3 loop or the peptides are cross-linked.

20. A complex according to claim 19, wherein said molecule or substance is CD4 capable of interacting with Env to expose a functional V3 loop.

21. A complex according to claim 19, wherein the molecule or substance is a heparin sulphate.

22. The complex of claim 1, wherein said molecule or substance is CD4 capable of interacting with Env to expose a functional V3 loop.

23. An immunogenic complex comprising a first peptide and a second peptide bound thereto, wherein the first peptide is a gp160 molecule comprising at least the V3 loop of gp120 and lacking at least the majority of the V2 loop of gp120, wherein the first peptide forms a trimer, and wherein the V3 loop is exposed and thereby bound to a binding region on the second peptide to form the complex, the second peptide comprising said binding region which region comprises at least residues 21-40 and 46-58 of Tat (SEQ ID NO 1), or at least said residues with a further point mutation whereby Cys 22 of Tat is replaced by Glycine to produce a Tat Cys22 mutant.

Description

EXAMPLE 1

(1) Molecular Interaction of HIV-1 Tat with the Gp120 V3 Loop

(2) The binding of HIV-1 Tat to the HIV-1 gp120 V3 loop was investigated using a molecular docking model, in which Tat (BH10 HIV strain) was allowed to interact with the V3 loop of the Env protein (Ba-L HIV strain). All structural models were calculated using, as template, all of the available structures of the Tat protein and of the Env protein deposited in the Protein Data Bank (Berman, H. M et al., Nucl. Acids Res. 28, 235-242, 2000) as of July 2003. The sequences of the various proteins were aligned using ClustalW (Thompson, J. D et al., NucL Acids Res. 22, 4673-4680, 1994) and the structural models were generated with Modeller6v2 (Sali, A. and Blundell, T. L. J. Mol. Biol. 234, 779-815, 1993). All of the calculated structural models were optimised through energy minimisation with AMBER-5 (Pearlman, D. A. et al., in AMBER 5.0, University of California, San Francisco, 1997). These structural models were then used to calculate the protein-protein adducts with the program BIGGER (Palma, P. N. et al., Proteins Struct. Funct. Genet. 39, 372-384, 2000). The latter program generates protein-protein complexes and ranks them on the basis of shape complementary and non-bonded (electrostatic and Van der Waals) interactions.

(3) Initial molecular docking calculations were made with the isolated V3 loop (a.a. 297-336) and gave rise to three types of low energy adducts, characterised by three unique interaction regions. These adducts are characterised by different Tat residues interacting with the V3 loop although, by contrast, the V3 loop showed only a single interaction region involving residues Thr300, Arg301, Ala331, His332, Asn334, and several amino acids of the 306-328 segment of the V3 loop.

(4) The interaction between Tat and a relatively large domain (291 a.a.) of HIV-1 Bal gp120 exposing the V3 loop was next calculated. As no structural information on the conformation of the V3 loop and its relative orientation with respect to the rest of the gp120 domain was available, the range of accessible conformations and the flexibility for V3 loop was sampled. The variability of the loop conformation can, in fact, have a sizable effect on the complex geometry. The conformation sampling was done through a long molecular dynamics simulation in explicit solvent. Docking calculations were performed, allowing Tat to interact with five different conformations of the gp120 of the Env protein including the two most different V3 loop conformations plus three intermediate conformations.

(5) These calculations identified an adduct interacting with Tat in a region involving the V3 loop, which was essentially the same as that found in one of the adducts found with calculations performed with the V3 loop alone. This adduct was, therefore, predicted to be the most stable and was subjected to molecular dynamics (MD) calculations to optimise its conformation and to estimate its stability when a complete force field (produced by the atoms of the two molecules) is effective. The calculations were performed on both the oxidised (i.e. with disulphide bridge on V3 loop) and the reduced states of the Env protein, and showed that the adduct is similarly stable in both oxidation states. In order to validate the interaction model found with the described procedure, and to analyse the protein-protein interface, docking calculations with the program Haddock were also performed. The five lowest energy adducts were found to have the same geometry as the model found with the BIGGER calculations.

(6) All these calculations, therefore, pointed at a unique mode of interaction. The final structural model of the adduct was found to be quite stable, with an average interaction surface of 2260?112 ?.sup.2 and an average protein-protein intermolecular energy of ?412?14 kcal mol.sup.?1. The largest contribution to the interaction energy was due to the electrostatic contribution, with an average energy of ?325?12 kcal mol.sup.?1: the interaction surface involves residues 1, 2, 4, 16, 19-22, 25, 26, 29, 34, 35, 45-47, 51, 55, 57, 59, 61 on Tat and residues 301, 316, 317, 318, 321, 322, 324, 325, 327, 328, 329, 331, 332, 405, 407, 412, 416-419 on the Env.

(7) Three intermolecular salt bridges between Tat and gp120 residues, respectively (Asp5-Arg316, Lys50-Glu407 and Lys19-Asp327) were found to be completely conserved in all of the various models and during the molecular dynamics simulations. The adduct was found to be further stabilised by additional intermolecular hydrogen bonds, which varied in number during the simulation from six to eleven, but always involving at least one residue of the common interaction region. A sizable contribution to the stabilisation of the adduct was also determined to be brought by 30 to 40 hydrophobic interactions. Twenty to thirty of these interactions were contributed by residues belonging to the V3 loop whereas about 20 of them by residues belonging to the 20-59 segment of Tat.

EXAMPLE 2

(8) Tat Binds the Gp120 V3 Loop in an ELISA Assay.

(9) Enzyme-linked immunosorbent assay (ELISA) tests were performed to determine whether Tat actually binds the gp120 V3 loop in vitro. To this purpose, ELISA plates were coated with a peptide encompassing the entire V3 loop, followed by extensive blocking with carrier bovine serum albumin (BSA), multiple washing steps, and additional incubation with biologically active Tat protein or, as a control, its buffer (PBS-BSA 0.1%) (Cafaro et al., Nat Med 1999; Fanales-Belasio et al., Immunology 2001). Monoclonal anti-V3 and polyclonal anti-Tat antibodies were used as primary antibodies for the detection of the bound protein.

(10) When uncoated (i.e. BSA-blocked) wells were used in the ELISA assay with anti-Tat or anti-V3 antibodies, a slight background signal was detected, ranging from about 0.1 to 0.4 OD. The results are shown in Table 1, below. However, when wells coated with the V3 peptide were incubated with Tat, the signal was increased to about 1 optical density (OD). In contrast, wells incubated with buffer alone yielded signals comparable to background levels (uncoated wells). As expected, V3 coated wells yielded high ELISA signals with anti-V3 antibodies. These experiments show that biologically active Tat binds to the gp120 V3 loop in vitro, confirming the data obtained with molecular docking calculations. Similar results were obtained by coating wells with Tat and by incubating coated wells with increasing amounts of the V3 loop peptide, which, as expected, showed a dose-dependent binding of V3 loop peptide to immobilised Tat (Table 1bis).

(11) TABLE-US-00001 TABLE 1 Antibodies Coating Incubation anti-Tat anti-V3 none buffer 0.132 OD 0.15 OD none Tat 0.431 OD 0.122 OD V3 (500 ng) buffer 0.277 OD 3 OD V3 (500 ng) Tat 1 OD 3 OD

(12) TABLE-US-00002 TABLE 1 bis V3 loop peptide amounts (ng) 0 50 100 200 500 Wells coated with Tat (100 ng) and blocked with bovine serum albumins (BSA) (100 ?g) anti V3 antibody 0.119 OD 0.181 OD 0.285 OD 0.435 OD 0.787 OD anti TAT antibody 1.438 OD 1.545 OD 1.51 OD 1.515 OD 1.567 OD Wells coated with BSA (BSA) (100 ?g) anti V3 antibody 0.103 OD 0.124 OD 0.148 OD 0.142 OD 0.179 OD

EXAMPLE 3

(13) Tat is Recognised by Antibodies Directed Against the CCR5 HIV Co-Receptor

(14) Since the gp120 V3 loop appears to be the major determinant for co-receptor choice and utilisation by HIV strains, experiments were performed to determine whether the capability of Tat to bind the V3 peptide was due to mimicry by Tat of co-receptor molecules. To this purpose, monoclonal antibodies directed against the major HIV-1 co-receptors (CCR5 and CXCR4) (Pharmingen) were used in an ELISA assay to determine whether they could recognise Tat, or Tat peptides consisting of specific Tat sequences and/or structural and functional domains. These monoclonal antibodies are known to recognise conformational epitopes present on HIV-1 co-receptors (Lee B et al., J Biol. Chem., 1999; Baribaud F et al., J Virol. 2001). Accordingly, any recognition of Tat by these antibodies would indicate that Tat shares structural similarity with the relevant co-receptor.

(15) ELISA plates were coated either with native Tat or one of the following Tat peptides (the region or regions to which these peptides essentially correspond is given in parentheses):

(16) Tat (1-20) (N-terminal domain);

(17) Tat (21-40) (cysteine-rich regiontransactivation domain);

(18) Tat (36-50) (core region);

(19) Tat (46-60) (basic regionnuclear localisation signal);

(20) Tat (56-70) (glutamine-rich region);

(21) Tat (65-80) (RGD sequence);

(22) Tat (73-86) (ROD sequence);

(23) Tat (83-102) (C-terminal domain); and

(24) Tat (21-58) (cysteine, core, and basic regions).

(25) Monoclonal anti-CCR5 or anti CXCR4 antibodies were used for the detection step. Anti-CCR5 specifically recognised the recombinant native Tat protein, the Tat (21-58) peptide and, although with a lower efficiency, the Tat (46-60) peptide. The results are shown in Table 2, below. In contrast, no recognition was observed with the antibodies directed against CXCR4.

(26) TABLE-US-00003 TABLE 2 Antibodies Coating anti-CCR5 anti-CXCR4 CTR isot. anti-Tat Tat 1.452 0.085 0.081 3.000 Tat 1-20 0.000 0.000 0.006 3.000 Tat 21-40 0.000 0.133 0.082 0.1 Tat 36-50 0.000 0.000 0.000 0.000 Tat 46-60 0.466 0.000 0.063 0.000 Tat 56-70 0.147 0.000 0.000 0.75 Tat 65-80 0.000 0.000 0.000 0.1 Tat 73-86 0.000 0.077 0.019 0.087 Tat 83-102 0.000 0.058 0.000 0.000 Tat 21-58 3.000 0.072 0.108 0.551

(27) The anti-CCR5 antibody used in these experiments is known to recognise a conformational epitope present in the CCR5 second extracellular loop (ECL2) and to be neutralising for HIV (Lee B et al., J Biol. Chem., 1999). In addition, RCL2 is known to be involved in the Env conformational changes leading to membrane fusion (Lee B et al., J Biol. Chem., 1999). Thus, these data indicated that Tat sequences encompassing both Tat transactivation domain and basic-rich region mimic at the structural level a region of the CCR5 involved in cell fusion upon recognition of CCR5 by gp120.

EXAMPLE 4

(28) Native, Biologically Active Tat is Required for CCR5 Recognition by Anti-CCR5 Antibodies

(29) The data described in Example 3, above, indicate that Tat sequence present in peptide 21-58 is capable of folding to mimic a conformational epitope of the CCR5 co-receptor. To determine whether a specific conformation of Tat is required for recognition by the anti-CCR5 antibody, the capability of native, biologically active Tat to be recognised by the anti-CCR5 antibody was compared to an oxidised Tat preparation, obtained by exposing the protein to the air and direct light, according to a procedure known to abrogate most of its biological activity (Fanales-Belasio, Immunology, 2001). This procedure results in the oxidation of SH groups and in the formation of intra- and intermolecular disulphide bounds, mediated by the cysteine residues present in the Tat transactivation domain. The transactivating properties of Tat, in turn, are known to activate the expression of host genes including HIV co-receptors (Huang 1998; Secchiero 1999). However, Tat transactivation properties are abolished in a transactivation mutant where cysteine 22 is substituted by a glycine (Tat-cys.sub.22) (Caputo A et al., Gene Ther. 1996). This Tat mutant, nevertheless, maintains its immunogenic properties, intact (Caselli E et al., J Immunol. 1999). Thus, Tat-cys.sub.22 was also included in this set of experiments.

(30) ELISA wells were coated with native Tat, oxidised Tat (Tat OX) or Tat-cys.sub.22 and the anti-CCR5, anti-CXCR4 antibody were used in the detection step. These experiments showed that the antibody specifically recognises the recombinant native Tat protein and the Tat-cys.sub.22 mutant, but not the oxidised Tat protein. The results are shown in Table 3, below. In contrast, and as a control, polyclonal (rabbit) anti-Tat antibodies recognised, as expected, all proteins with similar efficiency, demonstrating that all wells were equally coated.

(31) TABLE-US-00004 TABLE 3 Antibodies Coating anti-CCR5 anti-CXCR4 CTR isot. anti-Tat Tat 1.24 0.042 0.051 3 Tat OX 0.128 0.071 0.016 3 Tat-cys22 0.75 0 0.026 3

(32) These experiments showed, therefore, that Tat sequences encompassing the Tat transactivation domain, the core region and the basic region, fold to mimic a major epitope present on CCR5, and that a point mutation which abrogates the transactivating properties of Tat does not interfere with epitope formation.

EXAMPLE 5

(33) Extracellular Tat Enhances Infection of CD4+ Susceptible Cells by HIV-1 and Expands HIV-1 Tropism in CCR5 Low Expression Cell Lines.

(34) To determine whether Tat can mediate HIV-1 entry by mimicking CCR5, it was necessary to determine the effects of Tat on HIV entry in a CCR5-independent system. To this purpose, infection experiments were performed with a single cycle assay using a replication-defective recombinant HIV-1 encoding a cloramphenicol acetyltransferase (CAT) reporter gene and which was pseudotyped with the envelope glycoprotein of the CXCR4-tropic HXBc2 HIV isolate, or the CCR5-tropic ADA or YU2 HIV isolates. These replication-defective viruses (herein referred to as the R4-tropic HXBc2/HIV-CAT or the R5-tropic ADA/or YU2/HIV-CAT viruses) enter susceptible cells through CD4/CXCR4 or CD4/CCR5, integrate their cDNA's in the cell genome, and express the reporter gene CAT, but they can not produce progeny, i.e. they cannot support further infection of cells through subsequent cycles of virus production (Helseth E. J Virol 1990)). Thus, HIV-CAT viruses produce a single-round infection cycle of target cells, quantification of CAT acetylation levels allowing quantitative evaluation of the efficiency of HIV infection.

(35) Based on the data obtained in Examples 1 to 4, above, experiments were performed to determine whether Tat could assist infection by HIV, expand HIV tropism, and render TCLs susceptible to infection by R5-tropic (i.e. monocyte/macrophage-tropic) HIV strains, owing to molecular mimicry of CCR5 co-receptor by Tat. To this purpose, CEMss and Jurkat cells, two TCLs expressing both CD4 and CXCR4, but lacking CCR5 expression at the protein level in amounts detectable by standard flow cytometry or Western blot, or the CD4-negative 293 cell line, were plated on Tat-coated wells that had previously been incubated with HIV-CAT viruses pseudotyped with the envelope from the X4-tropic HXBc2 strain, or the CCR5-tropic ADA or YU2 strains. As expected, both CEMss and Jurkat cell lines were efficiently infected with the HXBc2/HIV-CAT, whereas no infection was detected with the CD4-negative 293 cells, due to the lack of the primary HIV-1 receptor. Strikingly, furthermore, both CEMss and Jurkat cells were also infected at high efficiency by the ADA or YU2 pseudotyped HIV-CAT, despite being known to be resistant to infection by R5-tropic HIV-1 strains. These data, therefore, confirmed the unexpected prediction that immobilised. Tat is capable of increasing HIV-1 cell tropism through molecular mimicry of specific CCR5 extracellular structural domains, i.e., of rendering CCR5-tropic strains capable of infecting TCLs expressing such low amounts of CCR5 to be not consistently infected in. the absence of Tat.

(36) Further experiments showed that Tat.sub.21-58, but not Tat.sub.21-40, was sufficient to assist infection of CEMss cells by the R5 tropic ADA Cat virus, showing that the region of Tat mediating binding to the gp120 V3 loop is the same as is required for HIV expanded tropism.

EXAMPLE 6

(37) Anti-CCR5 Antibodies, but not Anti-CXCR4 Antibodies, Block Tat-Assisted Infection of Low CCR5 Expression Cell Lines

(38) To further demonstrate that immobilised Tat expands the cell tropism of R5-tropic HIV-1 strains by mimicking CCR5, experiments were performed to determine whether active molecules capable of blocking CCR5 were also capable of blocking Tat-assisted infection of CCR5-negative cells. To this purpose, the CD4+/CCR5? (CCR5 RT-PCR-positive) CEMss cells were plated in the presence of antibodies directed against CCR5, CXCR4 or CCR3 on Tat-coated wells which were previously incubated with cell supernatants containing the R5-tropic ADA/HIV-CAT single infection round recombinant virus. Tat assisted infection was almost completely abolished by anti-CCR5 antibody, whereas no reduction in infectivity was observed with anti-CXCR4 or CCR3 antibody as compared to control. Since CEMss cells are CCR5-negative at the protein level, these data indicate that the blocking activity of the antibodies is due to their capability of recognising Tat structural motifs mimicking CCR5 conformational epitopes, as detailed in Examples 3 and 4. Further, these data confirmed that molecular mimicry of CCR5 by Tat is required for entry of CCR5-tropic HIV-1 strains in CCR5-negative cells.

EXAMPLE 7

(39) The Complexes Between Tat and Env are Novel Immunogens.

(40) To determine whether Tat/Env complexes represent novel immunogens, i.e. that cryptic epitopes were being exposed, mice were immunised with mixtures of: Tat and Env proteins known to expose the V3 loop; Tat and the V3 loop peptide; or with the single antigens, as controls. The rationale of these experiments is based on the prediction that the B cell epitope determinant repertoire of the two antigens, combined, will be different, as compared to that for the single antigens, since new epitopes are generated, cryptic epitopes are exposed, and/or pre-existing epitopes are hidden upon complex formation. Accordingly, some complexes are predicted to broaden and/or to increase the intensity of the humoral responses against Tat and/or Env, and others to narrow/decrease at least part of them, depending on the nature of the complex and the B cell epitopes generated, exposed or hidden.

(41) Three Env molecules were selected: wild type monomeric gp120 (wild type Env), which has been shown to generate more intense antibody responses against the V3 loop, owing to a better V3 loop exposure, as compared to the trimeric form present in the virus envelope (Fouts T R et al., J Virol 1997; Earl P L et al., J Virol 1994) a trimeric gp140 form of the Env molecule, which retains part of gp41 and is lacking the V2 loop (?V2Env) and is known to expose the V3 loop (Srivastava I K et al., J. Virol. 2003, Vol 77:11244-11259); and, a cyclic peptide corresponding to the V3 loop. To confirm the exposure of the V3 loop by the selected Env molecules, both monomeric wild type Env and trimeric ?V2Env were tested by ELISA for reactivity against polyclonal anti-V3 loop sera, with positive results.

(42) In a first 2-arm experiment, mice were immunised with Tat, wild-type Env, ?V2Env, or the combination of Tat and wild-type Env, or Tat and ?V2Env (in the presence of Alum as adjuvant) at days 0, 14 and 28. Humoral responses (IgG titres) were tested by ELISA on mouse sera obtained at day 38. Anti-Env IgG titres were strongly increased in mice vaccinated with Tat and ?V2Env combined as compared to mice immunised with ?V2Env alone. In contrast, they were comparable in mice immunised with wild type Env and Tat combined or with wild type Env alone. In addition, antibody titres against Tat were decreased upon combination with wild-type Env, but not upon combination with ?V2Env. These data confirm that the combination of Tat ?V2Env results in the formation of new molecular species (complexes) characterised by a new B cell epitope determinant repertoire. In addition, they show that the Tat/?V2Env complex has the capability to greatly increase anti-Env humoral responses, while protecting the elicitation of high anti-Tat antibody titres that, in contrast, are suppressed by wild type Env upon vaccination. Since monomeric wild type Env is shed in large amounts by HIV and infected cells, this is in agreement with the low frequency of anti-Tat antibodies in natural infection (Butta et al., J Infect Dis, 2002; Rezza et al., J Infect Dis, in press).

(43) Mice were also immunised with Tat, V3 loop peptide, or the combination of Tat and V3 loop peptide in Alum. Of note, the V3 loop peptide alone was not immunogenic, eliciting no or borderline antibody titres, whereas the combination of Tat and V3 loop peptide highly increased antibody titres against the V3 loop with no effects on antibody titres against Tat.

(44) Thus, these data demonstrate that Tat/Env or Tat/V3 loop complexes are novel immunogens, capable of eliciting higher and newer immune responses. In particular, the complex of Tat, together with ?V2Env or V3 loop peptide, induces better humoral responses against HIV Env, and that these are different as compared to those elicited by the corresponding single antigens or by the complex between Tat and wild type Env.

EXAMPLE 8

(45) Complexation of Tat and Env Changes Antibody Recognition of Individual Epitopes Present on Env

(46) To determine whether Tat/Env complexes induce humoral responses directed against Env epitopes different from those recognised upon immunisation with Env molecules alone, the sera from the same mice from the first immunisation protocol described in Example 7 were used to analyse reactivity to specific linear Env B cell epitopes. For this, 15-mer peptides spanning wild type Env or ?V2 Env (SHIV-1 SF162.P3) were mixed together to form pools of peptides composed of three contiguous 15-mers (i.e., covering 45 aminoacids of Env or ?V2 Env), and three 15-mers each overlapping the junction between two contiguous peptides.

(47) Sera from mice immunised with wild type Env, or ?V2 Env, combined with Tat, recognised linear epitopes present in between residues 77 to 132, i.e., spanning the first 14 amino acids of the HIV-1 V1 loop. In contrast, these epitopes were not recognised by sera from mice immunised with Env or ?V2 Env used as single antigens. Consistent with deletion of the V2 loop in ?V2 Env, only wild type Env elicited antibodies directed against the V2 loop, and reactivity was greatly increased upon combination with Tat. By contrast, immunisation with wild type Env alone elicited a strong reactivity against a region spanning residues 28 to 83 in HIV SF162 Env, that was completely lost upon co-immunisation with Tat. These data indicated, therefore, that the interaction between Tat and Env or DV2 Env exposes/hides linear Env/?V2 Env epitopes, consistent with complex formation.

(48) Significantly, sera from mice immunised with Tat combined to ?V2 Env, but not with ?V2 Env alone, showed a strong reactivity against a region spanning the N and C helix of gp41, indicating conformational modifications of gp41 which are known to occur upon binding of Env to CD4 and, in turn, of V3 loop to CCR5 or other co-receptors, and which are known to be required and to precede the fusion of the virus envelope and the cell membrane. Therefore, these data are consistent with binding of Tat to the V3 loop, and with mimicry of CCR5 binding to Env, which is required for virus entry. Most importantly, these data indicate that the complex between Tat and ?V2 Env can induce antibodies against HIV gp41, thereby being able to neutralise virus infectivity.

(49) These linear gp41 epitopes are not recognised by sera from mice immunised with ?V2 Env alone, indicating that the complex between Tat and ?V2 Env is a new immunogen.

(50) A similar change in epitope recognition is likely for conformational epitopes.

EXAMPLE 9

(51) The Increase of Anti-Env Antibody Titres Upon Immunisation with Tat/?V2Env or V3 Loop Peptide Complexes is Due to Formation of New/Stronger B Cell Epitopes Present in the Complex, and not to the Increase of Th2 Responses Against Env.

(52) It is well known in the art that T helper type 2 (Th2) responses, such as production of interleukin 4 (IL-4) by T cells, are key for generating humoral responses against antigens. Thus, at least part of the increase in anti-Env antibody titres elicited by immunisation with Tat combined with ?V2Env or the V3 loop peptide might be explained, in addition to the generation of new epitopes on the complex, by increased and/or broadened Th2 responses against Env. Therefore, we investigated the effects of immunisation with Tat/?V2Env complexes on Th2 responses.

(53) For this, mice immunised with a combination of Tat and ?V2Env, or with ?V2Env alone (see Example 7 for the protocol) were assessed for antigen-specific cellular responses against Env by IL-4 ELISPOT assay. This assay measures the production of IL-4, and is used in the art to evaluate Th2 responses against antigens or T cell-epitope peptides. Anti-Env cellular responses were assessed using matrices of pools of peptides containing Env 15-mers (overlapping by 11 amino acids) spanning the entire ?V2Env molecule. These experiments showed that immunisation with ?V2Env combined with Tat did not increase or broaden Th2 responses against Env, as compared to ?V2Env alone, but elicited Th2 responses directed against the same Env epitopes. Therefore, these data indicate that the increase of antibody titres against Env upon immunisation with Tat combined with ?V2Env or the V3 loop peptide is due to the generation/exposure of novel/stronger B cell epitopes upon complex formation.

EXAMPLE 10

(54) Tat Broadens Cellular Responses to Env in Mice Co-Immunised with Both Antigens.

(55) It is known that Tat can function as an adjuvant, increasing cell-mediated immune responses against antigens, and polarises the immune response toward a Th1 phenotype (Fanales Belasio et al., J Immunol 2002; Ensoli B., WO03/009867). In addition, Tat broadens Th1 responses against antigens by altering their processing by the proteasome (Ensoli et al., PCT/EP2004/11950). This results in the induction of responses against antigenic cytotoxic T cell (CTL) epitopes that are normally sub-dominant (Ensoli et al., PCT/EP2004/11950).

(56) To investigate the effects of immunisation with Tat and Env combined in a complex on Th1 responses and polarisation by Tat, production of ?IFN (a typical Th1 cytokine) by spleen cells from mice immunised with the combination of Tat and ?V2Env, or with ?V2Env alone, were analysed using matrices of pools of peptides containing Env 15-mers (overlapping by 11 amino acids) (see example 7 for the protocol). Anti-Env g-ELISPOTS displayed Env-specific T cell responses against a larger number of peptide pools and epitopes (not shown) in mice immunised with Tat+?V2Env, as compared to mice immunised with ?V2Env alone. These data indicate that Tat combined with ?V2Env maintains the capacity to broaden Th1 responses against Env in immunised mice, as already shown for Tat combined with wild-type Env (Ensoli et al., PCT/EP2004/11950). Therefore, Env complexes can be used as immunogens to induce effective/neutralising humoral responses and, at the same time, to broaden CTL responses against Env.