HIV infection is transmitted mostly through mucosal tissues. One of the major factors responsible for the difficulties in developing a vaccine against HIV has been the virus utilizing three major receptors on mononuclear cells: CD4, CCR5 and CXCR4. HIV tropism is generated largely by co-receptor selection, but initial transmission is by the M-tropic HIV, which utilizes the CCR5 co-receptor [1–3]. The ligands of CCR5 are the three CC chemokines, regulated upon activation: normal T-cell expressed/secreted (RANTES; CCL5), macrophage-inflammatory protein (MIP) 1α (CCL3) and MIP-1β (CCL4), which may block or downmodulate the co-receptors, thereby inhibiting HIV transmission . The most striking resistance to HIV infection is in the naturally occurring homozygous Δ32 CCR5 mutation in approximately 1% of caucasians [2,5]. These individuals lack cell-surface expression of CCR5, have increased concentrations of the three CC chemokines , may develop antibodies to CCR5 , and do not suffer from ill-health.
We have developed a simian model of immunization against SIV/HIV infection that attempts to reproduce some of the functional aspects of the Δ32 CCR5 mutation. The novel vaccine strategy utilizes both the CCR5 co-receptor and HIV/SIV molecules in preventing HIV transmission. The CCR5 co-receptors can be blocked by the three CC chemokines that downmodulate the cell-surface expression of CCR5 and may elicit protection in vivo [8–11]. The 70 000 Mr heat shock protein (HSP70) acts as a mucosal and systemic adjuvant by virtue of stimulating dendritic cells and monocytes to generate the three CC chemokines . HSP70 translocates molecules from outside the cell into the HLA class I pathway , thereby acting as a T helper cell type 1 polarizing adjuvant. However, as the increased CC chemokine concentration does not completely downmodulate the CCR5 from the cell surface a complementary antibody mechanism has been developed by immunization with the extracellular peptides of CCR5 [14,15]. These CCR5 antibodies are found in individuals with the homozygous Δ32 CCR5 mutation , in xeno- immunized macaques , allo-immunized humans [16,17], seronegative women at risk of HIV infection , and in CCR5 DNA immunization of macaques . A dual mechanism of blocking CCR5 may thus operate, in which the three CC chemokines and antibodies to CCR5 may bind different or the same extracellular domains of CCR5, block the receptor function and prevent HIV or SIV transmission. However, preliminary experiments suggest that neither SIV gp120 and p27 linked to HSP70 nor CCR5 will protect macaques from SIV infection, if the co-receptor and SIV antigens are administered separately (W.M.J.M. Bogers, L.A. Bergmeier, H. Oostermeijer, J.L. Heeney, T. Lehner, unpublished data).
In this paper we have employed the above three principal observations in a vaccination strategy that uses HIV/SIV subunit antigens and CCR5 peptides, with HSP70 as an adjuvant. We administered this vaccine by the vaginal mucosal route or by a para-mucosal route, targeting the proximity of the draining iliac lymph nodes (TILN). This elicited significant serum and vaginal antibodies and T-cell responses, as shown by generating helper function, IFN-γ, IL-2, IL-4 and CC chemokines. Challenge with SHIV89.6P derivative by the vaginal route infected all macaques, but the clearance of plasma viraemia to undetectable levels (< 40 copies per ml of plasma), as ascertained by RNA-polymerase chain reaction (PCR) of the virus, resulted in five of the eight immunized and none of the four unimmunized control macaques. This was associated with significant differences in the CD4 cell counts and immune parameters in macaques, which showed clearance of SHIV89.6P, compared with those that failed to clear the virus.
Materials and methods
Preparation of the vaccine antigens and peptides
HSP70 derived from Mycobacterium tuberculosis was prepared in Escherichia coli as described previously . It was purified by Q-sepharose followed by ATP affinity chromatography. The Q-sepharose chromatography was repeated to remove endotoxin, which was tested by the limulus amebocyte lysate assay and showed 1.2 pg endotoxin per 1 μg HSP70 protein. Recombinant SIV mac251 gp120 was expressed in baculovirus-infected cells and recombinant SIV p27 was generated in pGEX-3X as a glutathione S-transferase fusion protein. HIV (clade IIIB) gp120 was kindly supplied by Dr R. Doms. An N-terminal polypeptide of the gp120 extending up to and including the KEYAL sequence within the V2 loop was expressed as a fusion protein, with the Cγ2 and Cγ3 domains of a murine IgG2a heavy chain in transgenic Nicotiana tabacum and purified by affinity chromatography (J. Ma, D. Chargelegue, P. Drake and P. Obregon, in preparation). The fragment has a relative molecular mass of approximately 30 000 and when fused to the Ig domains migrates with a molecular mass of approximately 55 000 on sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE) (Fig. 1a). This sequence includes at least six potential N-glycosylation sites and the difference in the predicted molecular mass of the polypeptide sequence and the mobility on SDS–PAGE suggests that glycosylation had taken place . The proteins were analysed for purity using SDS–PAGE, Coomassie blue staining and Western blot analysis. As shown in Fig. 1a, the HSP70 preparation migrated as a doublet with a molecular mass of approximately 70 000 with a minor component at approximately 35 000. The HIV-gp120-Ig preparation migrated with a molecular mass of 55 000 Mr (as above) and a diffuse band of approximately 30 000 Mr. Both bands were recognized by anti-Ig antibodies. SIV p27 migrated as a doublet of 27 000 Mr.
CCR5 peptides derived from the sequences of the N-terminal, loop 1 and loop 2 were synthesized to a purity greater than 85%, as determined by high-pressure liquid chromatography and purchased from Neosystem Laboratories (Strasbourg, France). The sequences of the peptides are shown below:
N-terminal (aa1–20): Met – Asp – Tyr – Gln – Val – Ser – Ser – Pro – Ile – Tyr Asp – ILe – Asp – Tyr – Tyr – Thr – Ser – Glu – Pro – Cys
Loop 1 (aa 89–102): His – Tyr – Ala – Ala – Ala – Gln – Trp – Asp – Phe – Gly Asn – Thr – Met – Cys – Gln
Loop 2 (aa178–197): Cys – Ser – Ser – His – Phe – Pro – Tyr – Ser – Gln – Tyr Gln – Phe – Trp – Lys – Asn – Phe – Gln – Thr – Leu– Lys
Conjugation of HSP70 to the antigens and peptides
The HSP70 was conjugated to HIV gp120, SIV p27, N-terminal and the second loop of CCR5 by means of the N-succinimidyl (3[2-pyridyl]-dithio) propionate reagent, which is less likely to alter the immunogenicity of the vaccine components than glutaraldehyde . The first loop was non-covalently linked to HSP70, as this peptide binds directly to the peptide binding groove, demonstrated both by surface plasmon resonance and by immunization in mice . At each stage of the SPDP substitution and conjugation the HSP70/protein or peptide complex was subjected to SDS–PAGE and Western blot analysis. Under non-reducing conditions, high molecular weight complexes were observed that stained for both HSP70 and the conjugated protein, indicating successful conjugation. Under reducing conditions the complexes were broken down into their constituent molecules. HSP70-SIV p27 complex contained a small amount of free SIV p27.
Animals and virus stock
Mature outbred female rhesus monkeys (Macaca mulata) were housed at the Biomedical Primate Research Center, The Netherlands. During the entire duration of the experiment, the animals were checked for appetite, general behaviour, stools, weight and body temperature. The Animal Care and Use Committee of the institute approved the study protocol, according to international ethical and scientific standards and guidelines.
SHIV89.6P was constructed with SIV mac239 expressing the HIV-1 env of a macrophage primary isolate (89.6) and the associated auxillary genes tat, vpu and rev, as described elsewhere . After in-vivo passage this virus became pathogenic [24,25]. One ampoule of this virus was propagated on rhesus peripheral blood mononuclear cells (PBMC) in a feeder system as described recently . The supernatant from this culture represents a rhesus PBMC-derived SHIV89.6P stock; it was filtered (0.22 μm), and frozen in aliquots. The stock contained 247 ng/ml p27 antigen and 790 000 TCID50 per ml (50% tissue culture infective doses) as determined by endpoint titration on C8166 cells . Our stock of SHIV89.6 derivative used for vaginal challenge was tested for co-receptor usage using the Ghost cell assay . SHIV89.6 infected a very high percentage of CCR5 and the Hi5 clone of CCR5 cells (> 50%) and to a lesser extent CXCR4 cells.
Vaginal and iliac lymph node immunization
Three groups of four macaques were investigated, of which four were unimmunized controls and eight were immunized by the vaginal or TILN route (Fig. 1). Aliquots of 150 μg HIV gp120 and 100 μg SIV p27 were linked with the same concentrations of HSP70, respectively. The concentrations of these antigens were initially determined in mice and then calculated for the weight difference of macaques. The concentration of each of the three peptides was, however, at a molar ratio of 1 : 10, with 50 μg HSP70 linked to 20 μg of each peptide. The total amount of HSP70 was thus 400 μg linked to 150 μg HIV gp120, 100 μg SIV p27 and 60 μg of the three CCR5 peptides. The vaccine was applied in a solution of approximately 1 ml directly into the vagina, by means of a lubricated paediatric nasogastric tube to four macaques, as described previously . TILN immunization was carried out by injecting 0.5 ml of the solution on each side of the groin in the proximity of the external and internal iliac lymph nodes in four macaques, as described before . The vaccination was carried out on weeks 1, 4 and 16. Blood and vaginal washings  were taken before and approximately one month after each immunization.
Serum and vaginal IgG and IgA antibodies
Specific serum and vaginal IgG and IgA antibodies to SIV gp120, p27 and the extracellular peptides were assayed using enzyme-linked immunosorbent assay as described previously [8,28]. Briefly, plates were coated with a pre-determined optimal concentration of antigen (1 μg/ml), peptides (10 μg/ml) and a random 20mer peptide (R20) as a control, and were then incubated with double dilution of serum or vaginal washings. Bound antibody was detected by incubation with rabbit IgG anti-monkey IgA (8 μg/ml) (Nordic Immunological Laboratories, Tilburg, The Netherlands) or IgG (2 μg/ml; Sigma-Aldrich, Poole, Dorset, UK), followed by affinity-purified goat anti-rabbit IgG-alkaline phosphatase conjugate (Sigma). The IgG and IgA antibody titres are presented as reciprocals before and after each immunization.
To determine neutralizing activity in the sera from vaccinated animals, Ghost cells were used as the target cells in co-receptor specific neutralization assays . Ghost cells expressing the co-receptors CXCR4 and CCR5 were used; reagents were obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, National Institute of Allergy and Infectious Diseases, National Institutes of Health from Dr Vineet, N. Kweal Ramani and Dr D.R. Littman.
T-cell proliferative responses
T-cell cultures were set up by separating mononuclear cells from defibrinated blood by Lymphoprep (Nycomed, Oslo) density gradient centrifugation before and after each immunization from all macaques . The cells were cultured with and without 1 and 10 μg/ml HIV gp120, SIV p27, three extracellular peptides of CCR5, a control peptide (R20), and concanavalin A in 96-well round-bottomed plates (Costar, Cambridge, MA, USA), containing RPMI 1640 (Invitrogen, Paisley UK), as described before . The results were expressed as stimulation indices (SI; ratio of counts with and without antigen) before and after each immunization, for cultures stimulated with the optimum concentration of antigen. All cultures yielded high SI and counts with concanavalin A, and no significant increase in counts was seen with the control peptide.
Generation and assay of CC chemokines
PBMC were separated and CD8 cells were enriched by the depletion of CD4 cells, B cells and monocytes by panning with CD4 (OKT4 hybridoma culture supernatant) and antibodies to human immunoglobulin (Serotec, Oxford, UK). The enriched CD8 cells were cultured with 10 μg/ml phytohaemagglutinin for 3 days. Preparation of the CD8 culture supernatant was then carried out according to the method described previously . Phytohaemagglutinin-stimulated CD8 cells were cultured at a concentration of 3 × 106 cells/ml in RPMI-1640 containing 10% fetal calf serum and 10% human IL-2 preparation (Biotest, Solihull, UK). Phytohaemagglutinin has been used, as it stimulates maximum concentrations of RANTES, MIP-1α, MIP-1β, and the assay is reproducible, unlike the variations when specific antigens are used. The chemokines were assayed in the CD8 cell culture supernatant using a specific enzyme immunoassay (R&D Systems Europe Ltd., Abingdon, Oxon, UK).
Evaluation of the proportion and cell-surface expression of CCR5 by flow cytometry
The cell-surface expression of CCR5 was evaluated by incubating 100 μl blood with 10 μl Mab 2D7 (Pharmingen, Woerden, The Netherlands), mixed in a 5.0 ml polystyrene, round-bottom tube (Falcon 2058; Becton Dickinson, Lincoln Park, NY, USA), at 4°C for 30 min, followed by washing with FACS buffer. Goat anti-mouse antibody (10 μl) was incubated for 30 min at 4°C. After staining, red blood cells were lysed by adding 2 ml FACS-lysis buffer (Becton Dickinson) for 20 min at room temperature. The cells were then washed, fixed with 2 ml freshly prepared 2% w/v paraformaldehyde overnight. Flow cytometry was performed on a FAC sort using Cell Quest software (Becton Dickinson), and a minimum of 5000 events were analysed.
Sequential CD4 cell counts
CD4 antibody conjugated to allophycocyanin (Becton Dickinson, Etter-Leur, NL) was used to determine the proportion of CD4 cells (percentage of PBMC) as described above for CCR5.
MHC typing of macaques
The Mamu A*01 antigen was assayed in these macaques by Dr D. Watkins’ laboratory and were found to be negative for this allele.
Assays for IL-2, IFN-γ and IL-4
The enzyme-linked immunospot (ELISPOT) assay was used to enumerate antigen-specific cytokine-secreting cells of IL-2, IL-4 and IFN-γ as reported elsewhere . Briefly, 1 μg recombinant SIV p27, 1 μg Env HIVIIIB (gp120, EVA-607; MRC) or 1 μg of the loop 1, 2 or N-terminal CCR5 peptides was used as antigen. Medium alone was used as a negative control, whereas concanavalin A (5 μg/ml; Sigma) was used as a mitogen (positive control) to test the responsiveness of PBMC.
Viral infection and kinetics
Vaginal inoculation with SHIV89.6P derivative
The macaques were first made more susceptible to mucosal infection by treatment with progesterone . One week after progesterone implantation, the animals were inoculated intravaginally with 10 MID50 of SHIV89.6P in 1 ml (× 2) over a 4-h period . Intravaginal inoculations were carefully performed to avoid any trauma to the mucosal surface one month after the last immunization (week 20) and inoculations were avoided during menses.
Plasma viral loads
Plasma viral loads were evaluated using the RNA PCR (QC-RNA-PCR) assay . Briefly, RNA was extracted from 200 μl ethylene diamine tetraacetic acid plasma by guanidine isothiocyanate-mediated lysis, followed by propan-2-ol precipitation. A known amount of internal standard RNA was added before the RNA extraction and was co-purified to monitor the efficiency of the purification. The RNA was reverse transcribed and amplified in a single reaction protocol using recombinant Tth DNA polymerase and biotinylated primers. The amplified fragments were then denatured and hybridized to an immobilized capture probe and detected by an avidin-enzyme conjugate-mediated colorimetric reaction. For the SHIV89.6P two nested PCR procedures were performed on each RNA sample using, an ENV-V3 nested primerset for HIV-1 and a GAG nested primerset for SIV.
The non-parametric Kruskal–Wallis and Mann–Whitney tests were used for the three groups of macaques. To evaluate the sequential changes in the viral load, the areas under the graphs from 4 to 24 weeks were determined for each macaque, and these were then analysed by the above non-parametric test .
Vaginal mucosal and TILN immunization was carried out with HSP70 linked to HIV gp120, SIV p27 and the three CCR5 peptides (Fig. 1). Moderate IgG and IgA antibody titres were elicited (Fig. 2), and statistical analysis of the three groups of macaques using the Kruskal–Wallis non-parametric test showed significant variation for serum IgG, as well as IgA antibodies to HIV gp120, SIV p27, CCR5 N-terminal (P ≤ 0.02) and to HSP70 (P < 0.01). Analysis of each immunized group with the control group using the Mann–Whitney test showed that the novel plant-generated HIV gp120 elicited IgG and IgA antibodies by both routes (P < 0.02) (Fig. 2a,b). IgG antibodies to SIV p27 were elicited only by the TILN route, but IgA antibodies were elicited by both the vaginal and TILN routes of immunization (P < 0.02). The higher levels of serum IgG antibodies elicited by the TILN than the vaginal route of immunization might be the result of differences in the way the antigens were processed and presented by antigen presenting cells, or possibly by the antigen entering the blood stream. Immunization with the three extracellular CCR5 peptides elicited both IgG (P = 0.02) and IgA (P = 0.01) antibodies, predominantly to the N-terminal peptide (P ≤ 0.02) and to a limited extent to loop 1 and 2 CCR5 peptides (Fig. 2a,b). These antibodies recognized intact cell-surface CCR5 expressed on HEK293 cells, inhibited SIV infectivity, and were complementary to HIV inhibition by CC chemokines, as shown elsewhere . HSP70 elicited significant serum antibody titres of the IgG (1 : 2000 ± 400) and IgA (1 : 450 ± 126) classes by the TILN route of immunization (P < 0.01), and no IgG, with negligible IgA antibodies by the vaginal route (data not shown). Sera collected 2 weeks before challenge failed to show neutralizing antibodies to the challenge virus (SHIV89.6P) and to HIV-1IIIB, using Ghost cells expressing CCR5 or CXCR4 co-receptors (data not presented).
Significant variation between the three groups of macaques was found (P = 0.05), with IgG antibodies to HIV gp120 and IgA antibodies to SIV p27 (Fig. 2c,d). As with serum, antibodies were elicited mostly to the N-terminal of CCR5, and significantly higher IgG antibody titres to HIV gp120 (P = 0.05) were reached only after TILN (8 ± 2.4), but IgA antibodies after vaginal (4 ± 1.4) immunization. In contrast to serum antibodies, higher IgA and IgG antibody titres to SIV p27 were elicited by the vaginal route than TILN immunization (Fig. 2c,d).
The T-cell proliferative responses were elicited mostly in TILN immunized macaques (Table 1). The SI to the plant generated HIV gp120 (9.1 ± 6.7), N-terminal (5.5 ± 3.3), loop 1 (8.2 ± 5.6) and loop 2 (3.8 ± 2.1) CCR5 peptides were all above, unlike those after vaginal immunization or in the unimmunized controls.
Antigen-specific IL-2 and IFN-γ-producing cells (ELISPOTs) were assayed to evaluate T helper type 1 cytokines and IL-4 representing T helper type 2 cytokines (Fig. 3). Significant differences between the three groups of macaques were found in IL-2, IFN-γ and IL-4-producing cells when stimulated with SIV p27, HIV gp120 or the CCR5 loop 2 peptide (Kruskal–Wallis P = 0.006–0.017). The number of ELISPOTs stimulated with any one of these antigens was significantly greater for the TILN and vaginally immunized groups of macaques, compared with the control animals (Mann–Whitney test; P < 0.05). Surprisingly, although vaginal immunization elicited higher cytokine-producing cells when stimulated with HIV gp120 or CCR5-loop 2 (except for IL-4 stimulated by loop 2) compared with TILN immunization, the converse was found with SIV p27 (Fig. 3). These differences reached the 5% level of significance with IL-2, stimulated by HIV gp120 or SIV p27. The N-terminal and loop 1 peptides induced significant variations between the three groups of macaques only with IL-2 (P < 0.01) but the numbers of cells producing IFN-γ or IL-4 were very small (data not presented).
Mucosal and TILN immunization induced significantly greater RANTES, MIP-1α and MIP-1β in the two groups of immunized macaques, compared with unimmunized controls (Fig. 4a). This reached the 5% level of significance with MIP-1α (P = 0.03) and MIP-1β (P = 0.02). As with some of the cytokines, vaginal immunization elicited higher concentrations of MIP-1β (P = 0.03), MIP-1α (P = 0.03) and to a limited extent RANTES than TILN immunization. The combined concentration of the three chemokines was also significantly higher in macaques immunized by the vaginal (3505 ± 237 pg/ml) than by the TILN route (1647 ± 267 pg/ml; P < 0.01). Studies of stromal cell-derived factor 1, however, showed no difference between the immunized and control animals (Fig. 4a).
CCR5 co-receptor assay
The proportion and cell-surface expression of CCR5 showed significant variations in the three populations of vaginal, TILN immunization and control macaques (P < 0.05) (Fig. 4b). A lower proportion of CCR5 cells was found in the vaginally immunized (25.2 ± 2.2%) or TILN immunized animals (33.0 ± 2.7%) compared with the unimmunized macaques (42.5 ± 5.5%; P = 0.03). However, the CCR5 subset of cells that contribute to the differences was not identified. The proportion of cells expressing CCR5 showed a significant inverse correlation with the concentration of three CC chemokines (r = 0.767, P < 0.01) (Fig. 4c). The mean fluorescence intensity showed no significant differences between the three groups of macaques (not shown).
Vaginal challenge with 10 MID50 of SHIV89.6P derivative infected all four unimmunized and eight immunized macaques. To evaluate sequential changes in the viral load, the areas under the graphs from 0 to 24 weeks were determined for each macaque, and these were then analysed using the Kruskal–Wallis test. A significant variation was found between the three groups of macaques over the 24-week period (P = 0.05). At the end of the experiment, 24 weeks after virus challenge, the mean viral load was approximately 3.7 logs lower in the TILN and 2.3 logs lower in the vaginally immunized than in the control animals (Fig. 5e). The SHIV89.6P derivative remained at a steady level in three of the four control macaques during the entire 24-week period after challenge, although in one animal the viral load fell to below 103 RNA copies, but another had to be killed at week 12 (Fig. 5a). In contrast, in five of the eight immunized animals the virus fell to undetectable levels (Fig. 5b); three of the four TILN and two of the four vaginally immunized macaques. It should be noted that in two of the five macaques the virus became undetectable at weeks 8 and 12 and in the other three at week 24. Although statistical analysis reached the 5% level of significance for the virus load between the three groups of macaques, the results must be viewed with caution in view of the small number of animals.
CD4 cell counts
The CD4 cell counts were monitored throughout the experimental period (Fig. 5c,d,f). Significant variation was found between the three groups of macaques (P = 0.037). Of the four control macaques one had to be killed at week 12, with a CD4 cell count of 10%, and in another the count fell to an undetectable level 24 weeks after challenge with SHIV89.6P (Fig. 5c). Indeed, the CD4 cell count was higher in macaques immunized by the vaginal (32.3 ± 4.2) or TILN (25.2 ± 3.2) route than the unimmunized controls (12.1 ± 12.8) (Fig. 5f). It is noteworthy that whereas the TILN immunized animals had the lowest mean copies of SHIV89.6P derivative, the vaginally immunized macaques retained the highest CD4 cell count.
Analysis of immune markers of protection
Immune markers of protection were evaluated by dividing the eight immunized macaques into two groups; five macaques in which SHIV89.6P was cleared or became undetectable by RT–PCR (< 40 copies/ml) and three macaques that failed to clear the virus. All immune parameters were analysed in the two groups after the third immunization and before the macaques were challenged by the live virus. Higher serum IgG antibodies to the three CCR5 peptides and to a lesser extent to HIV gp120 and SIV p27 were found in the five macaques in which the virus was cleared, compared with the three persistently infected immunized animals (Table 2). Similar results were recorded with vaginal fluid IgA antibodies to the three CCR5 peptides, as well as IgG antibodies to HIV gp120. The analysis of T-cell functions showed that T-cell proliferation to the three CCR5 peptides, HIV gp120 and IL-2 ELISPOTs to HIV gp120 and SIV p27 were higher in the macaques in which SHIV was cleared, compared with those in which the infection persisted (Table 2). However, IFN-γ ELISPOTs showed little difference between the two immunized groups of macaques, although they were significantly increased when compared with the control animals (P < 0.02). All five macaques in which virus was cleared developed significant T-cell proliferative responses (SI > 2) to one, two or all three CCR5 peptides, whereas only one out of three unprotected macaques showed an SI greater than 2, and even then only with one peptide. HIV gp120 showed significant stimulation of T cells in three out of five protected but none out of three unprotected macaques. However, only one protected macaque showed T-cell proliferation to all five antigens or peptides, but the other four showed T-cell responses to one or two peptides. In contrast, only one out of three unprotected macaques responded, and then to one peptide only (loop 1 of CCR5). An inverse correlation between the concentration of the three CC chemokines and IgG antibody titres to the three CCR5 peptides was found (not presented), and this suggests that the chemokines and antibodies may have a complementary function in preventing HIV infection. Neutralizing antibodies were not elicited with the heterologous HIV gp120 used, and did not play a role in containing SHIV infection.
The results are consistent with a broadly based T-cell and antibody mechanism clearing the SHIV infection, on account of increased titres in serum and vaginal antibodies, raised T-cell proliferative responses, T helper type 1 (IL-2)-producing cells, and the complementary effect of the three CC chemokines and IgG antibodies.
The resistance to HIV infection in the naturally occurring homozygous Δ32 CCR5 mutation [2,5] led us to develop the HIV-receptor immunization strategy. We targeted both the virus, by using HIV gp120 and SIV p27 to elicit specific immunity, and the CCR5 co-receptor, to block it with CC chemokines and antibodies. The vaginal and TILN routes of immunization elicited significant immune responses to HIV gp120, SIV p27 and CCR5 (especially its N terminus). However, TILN immunization was more effective in eliciting serum and vaginal IgG antibodies, T-cell proliferative responses and cytokine-producing cells to SIV p27, whereas vaginal immunization using the same vaccine elicited higher vaginal IgA antibody titres, MIP-1β and MIP-1α concentrations, a lower proportion of CCR5 cells and higher T helper type 1 cytokine-producing cells (IL-2 and IFN-γ) stimulated by HIV gp120 or loop 2 CCR5 peptide. Furthermore, a significant inverse correlation was found between the concentrations of the three CC chemokines and the proportion of cells expressing cell-surface CCR5 (P < 0.05), as reported in another series of macaques . No single immunological correlate of protection was, however, identified in this investigation. A recent report that HSP70 is incorporated into the virions of HIV and SIV  raises the possibility that HSP70 antibodies may affect HIV or SIV replication. The contribution of each of the vaccine components has not been controlled and will need to be determined. However, preliminary results suggest that HSP70 alone, linked with CCR5 peptides or SIV gp120 and p27, does not elicit protection unless the three components are combined (Bogers et al., in preparation). In the latter experiment the challenge virus was SIV mac8980, so it does not exclude the possibility that challenge with SHIV89.6P would have elicited a different result. Immunization with DNA of the CCR5 peptide alone also failed to protect macaques from intravenous challenge with SIV mac .
Vaginal challenge with SHIV89.6P derivative, after progesterone treatment , resulted in infection of all macaques. However, whereas all four unimmunized macaques remained infected over the 24 weeks of observation, and one of the animals had to be killed at week 12, a significant variation in viral load was found between the three groups of macaques over the 24-week period (P = 0.05). At the end of the experiment (week 24) the mean viral load was approximately 3.7 logs lower in the TILN and 2.3 logs lower in the vaginally immunized, compared with the control animals. Indeed, in five out of eight immunized macaques the virus could not be detected in peripheral blood. Although the RT–PCR method used was very sensitive, it is not clear whether SHIV89.6P derivative was contained or the virus was completely cleared in these macaques. Furthermore, although in two of the five macaques the virus became undetectable at weeks 8 and 12, in the other three animals this was found only by week 24. In view of this, the small number of macaques and variation in the viral load of the controls, the results must be viewed with caution. The mean CD4 cell count was lowest in the control animals (12.1 ± 12.8), and was significantly higher in the vaginally (32.3 ± 4.2) as well as the TILN (25.2 ± 3.2) immunized macaques (P < 0.05). As all macaques were negative for Mamu A*01 antigen, the decrease in viral load in some of the macaques could not be accounted for by this allele.
In most of the recent vaccine studies in macaques, the endpoint of vaccine-induced protection was prevention of the development of AIDS by decreasing the viral load or the clearance of SHIV89.6 generally by DNA priming with DNA-carrying virus boosting [35–40] or by live attenuated virus vectors [41,42]. These cytotoxic T-lymphocyte-based HIV vaccine strategies are susceptible to viral escape , and superinfection by mutated strains of HIV have been reported in patients with potent cytotoxic T-lymphocyte responses to a number of HIV epitopes [44,45]. The outcome of the novel virus-receptor vaccination strategy and challenge with SHIV89.6P by the vaginal route also failed to prevent infection, but showed a significant decrease in viral load, maintained the CD4 cell count and health of the macaques. However, the immune response was broadly based, integrating innate and adaptive immunity, in addition to eliciting IFN-γ ELISPOTs as markers of cytotoxic T-lymphocytes. It is noteworthy that unlike the above SHIV89.6P experiments, neither HIV Env nor SIV gag in the present study were homologous with SHIV89.6P. Indeed, an HIV gp120 N-terminal fragment was produced in genetically modified plants, and this is the first evidence that a plant-generated HIV antigen elicited antibodies and T-cell responses in primates. The efficacy of mucosal immunization was demonstrated in another study of rectal administration of four HIV/SIV peptides mixed with E. coli heat-labile toxin , in which rectal challenge with SHIV-KU2 resulted in lower plasma viral loads or clearance of the virus in the immunized compared with the control macaques.
In the present study, the viral load was significantly decreased, in spite of SHIV89.6P utilizing CXCR4 to a greater extent than CCR5 [24,25,47]. Only CCR5 co-receptor usage was targeted, with the upregulation of the three CC chemokines and antibodies to CCR5. However, antibodies to CCR5 may inhibit not only R5 but also R5X4 viruses in vitro, and dual tropic isolates may use CXCR4 on transfected cells but not necessarily on primary cells .
Quite apart from blocking and downmodulating CCR5 and inducing specific serum and vaginal IgA and IgG antibodies, IL-2 and IFN-γ are specifically stimulated by the immunizing antigens, and these may be involved in T helper type 1 cell-generated protection  These results are consistent with a combined antibody, T-cell-mediated, and innate concept of immune protection to mucosal HIV or SIV infection . The innate and adaptive immune repertoire may operate not only in central or systemic immunity, as is required in blood-induced HIV infection, but also in the peripheral mucosal–lymph node interaction, which in sexual transmission is the initial site of exposure to HIV.
The authors would like to thank Dr Sue Chinn for valuable statistical advice, and wish to acknowledge the support received from the European Union (grant no. BMH4-CT97-2345 and QLK2CT-1999-01321).
1. Alkhatib G, Combadiere C, Broder CC, Feng Y, Kennedy PE, Murphy PM, Berger EA. CC CKR5: a RANTES, MIP-1α and MIP-1β receptor as fusion cofactor for macrophage-tropic HIV-1. Science
2. Deng H, Liu R, Ellmeier W, Choe S, Unutmaz D, Burkhart M, et al
. Identification of a major co-receptor for primary isolates of HIV-01. Nature
3. Dragic T, Litwin V, Allaway GP, Martin SR, Huang Y, Nagashima KA, et al
. HIV-1 entry into CD4+ cells is mediated by the chemokine receptor CC-CKR-5. Nature
4. Cocchi F, DeVico AL, Garzino-Demo A, Arya SK, Gallo RC, Lusso P. Identification of RANTES, MIP-1α and MIP-1β as the major HIV-suppressive factors produced by CD8+ T cells. Science
5. Dean M, Carrington M, Winkler C, Huttley GA, Smith MW, Allikmets R, et al
. Genetic restriction of HIV-1 infection and progression to AIDS by a deletion allele of the CKR5 structural gene. Science
6. Paxton WA, Martin SR, Tse D, O'Brien TR, Skurnick J, VanDevanter NL, et al
. Relative resistance to HIV-1 infection of CD4 lymphocytes from persons who remain uninfected despite multiple high-risk sexual exposures. Nat Med
7. Ditzel HJ, Rosenkilde MM, Garred P, Wang M, Koefoed K, Pedersen C, et al
. The CCR5 receptor acts as an alloantigen in CCR5 Δ32 homozygous individuals: identification of chemokine and HIV-1 blocking human antibodies. Proc Natl Acad Sci U S A
8. Lehner T, Wang Y, Cranage M, Bergmeier LA, Mitchell E, Tao L, et al
. Protective mucosal immunity elicited by targeted iliac lymph node immunization with a subunit SIV envelope and core vaccine in macaques. Nat Med
9. Heeney JL, van Gils M, Teeuwsen VJP, Dubbes R, Koornstra W, Teeuwsen V, et al
. β-Chemokines and neutralizing antibody titres correlate with heterologous protection of HIV-1 vaccinated macaques. Proc Natl Acad Sci U S A
10. Ahmed RKS, Nilsson C, Wang Y, Lehner T, Biberfeld G, Thorstensson R. β-Chemokine production in macaques vaccinated with live attenuated virus correlates with protection against SIVsm challenge. J Gen Virol
11. Lehner T, Wang Y, Cranage M, Tao L, Mitchell E, Bravery C, et al
. Up-regulation of β-chemokines and down-modulation of CCR5 co-receptors inhibit simian immunodeficiency virus transmission in non-human primates. Immunology
12. Lehner T, Mitchell E, Bergmeier L, Singh M, Spallek R, Cranage M, et al
. Heat shock proteins generate β-chemokines which function as innate adjuvants enhancing adaptive immunity. Eur J Immunol
13. Castellino F, Boucher PE, Eichelberg K, Mayhew M, Rothman JE, Houghton AN, Germain RN. Receptor-mediated uptake of antigen/heat shock protein complexes results in major histocompatibility complex class I antigens presentation via two distinct processing pathways. J Exp Med
14. Lehner T, Wang Y, Doyle C, Tao L, Bergmeier LA, Mitchell E, et al. Induction of inhibitory antibodies to the CCR5 chemokine receptor and their complementary role in preventing SIV infection in macaques. Eur J Immunol
15. Lehner T, Doyle C, Wang Y, Babahmady K, Whittall T, Tao L, et al
. Immunogenicity of the extracellular domains of C-C chemokine receptor 5 and in vitro effects on simian immunodeficiency virus or HIV infectivity. J Immunol
16. Wang Y, Tao L, Mitchell E, Bravery C, Berlingieri P, Armstrong P, et al
. Allo-immunization elicits CD8+ T cell-derived chemokines, HIV suppressor factors and resistance to HIV infection in women. Nat Med
17. Wang Y, Underwood J, Vaughan R, Harmer A, Doyle C, Lehner T. Allo-immunization elicits CCR5 antibodies, SDF-1 chemokines, and CD8-suppressor factors that inhibit transmission of R5 and X4 HIV-1 in women. Clin Exp Immunol
18. Lopalco L, Barassi C, Pastori C, Longhi R, Burastero SE, Tambussi G, et al
. CCR5-reactive antibodies in seronegative partners of HIV-seropositive individuals down-modulate surface CCR5 in vivo and neutralize the infectivity of R5 strains of HIV-1 in vitro . J Immunol
19. Zuber B, Hinkula J, Vodros D, Lundholm P, Nilsson C, Morner A, et al
. Induction of immune responses and break of tolerance by DNA against the HIV-1 coreceptor CCR5 but no protection from SIVsm challenge. Virology
20. Mehlert A, Young DB. Biochemical and antigenic characterization of the Mycobacterium tuberculosis 71 kD antigen, a member of the 70 kD heat-shock protein family. Mol Microbiol
21. Faye L, Johnson KD, Strum A, Chrispeels MJ. Structure, biosynthesis and function of asparagine-linked glycans on plant glycoproteins. Physiol Plant
22. Czerkinsky C, Russell MW, Lycke N, Lindblad M, Holmgren J. Oral administration of a streptococcal antigen coupled to cholera toxin B subunit evokes strong antibody response in salivary glands and extramucosal tissues. Infect Immun
23. Wang Y, Kelly CG, Singh M, McGowan E, Carara A-S, Bergmeier L, Lehner T. Stimulation of TH1-polarizing cytokines, C-C chemokines, maturation of dendritic cells, and adjuvant function by the peptide binding fragment of HSP70. J Immunol
24. Karlsson GB, Halloran M, Li J, Park IW, Gomila R, Reimann KA, et al
. Characterization of molecularly cloned simian– human immunodeficiency viruses causing rapid CD4+ lymphocyte depletion in rhesus monkeys. J Virol
25. Reimann KA, Li JT, Voss G, Lekutis C, Tenner-Racz K, Racz P, et al
. A chimeric simian/human immunodeficiency virus expressing a primary patient human immunodeficiency virus type 1 isolate env causes an AIDS-like disease after in vivo passage in rhesus monkeys. J Virol
26. Bogers WM, Dubbes R, ten Haaft P, Niphuis H, Cheng-Mayer C, Stahl-Hennig C, et al
. Comparison of in vitro and in vivo infectivity of different clade B HIV-1 envelope chimeric simian/human immunodeficiency viruses in Macaca mulatta . Virology
27. Cecilia D, KewalRamani VN, O'Leary J, Volsky B, Nyambi P, Burda S, et al
. Neutralization profiles of primary human immunodeficiency virus type 1 isolates in the context of coreceptor usage. J Virol
28. Lehner T, Bergmeier LA, Tao L, Panagiotidi C, Klavinskis LS, Hussain L, et al
. Targeted lymph node immunization with simian immunodeficiency virus p27 antigen to elicit genital, rectal and urinary immune responses in nonhuman primates. J Immunol
29. Van der Meide PH, Groenestein RJ, de Labie MC, Heeney J, Pala P, Slaoui M. Enumeration of lymphokine-secreting cells as a quantitative measure for cellular immune responses in rhesus macaques. J Med Primatol
30. Marx PA, Spira AI, Gettie A, Dailey PJ, Veazey RS, Lackner AA, et al
. Progesterone implants enhance SIV vaginal transmission and early virus load. Nat Med
31. Miller CJ, McChesney MB, Lu S, Dailey PJ, Chutkowski C, Lu D, et al
. Rhesus macaques previously infected with simian/human immunodeficiency virus and protected from vaginal challenge with pathogenic SIV mac239. J Virol
32. Ten Haaft P, Verstrepen B, Uberla K, Rosenwirth B, Heeney J. A pathogenic threshold of virus load defined in simian immunodeficiency virus- or simian–human immunodeficiency virus-infected macaques. J Virol
33. Matthews JNS, Altman DG, Campbell MJ, Royston P. Analysis of serial measurements in medical research. BMJ
34. Gurer C, Cimarelli A, Luban J. Specific incorporation of heat shock protein 70 family members into primate lentiviral virions. J Virol
35. Barouch DH, Santa S, Schmitz JE, Karoda MJ, Fu TM, Wagner W, Bilska M. Control of viraemia and prevention of clinical AIDS in rhesus monkeys by cytokine-augmented DNA vaccination. Science
36. Amara RR, Villinger F, Altman JD, Lydy, SL, O'Neil, SP, Staprans SI, et al
. Control of a mucosal challenge and prevention of AIDS by a multiprotein DNA/MVA vaccine. Science
37. Amara RR, Smith JM, Staprans SI, Montefiori DC, Villinger F, Altman JD, et al
. Critical role for env as well as gag-pol in control of a simian–human immunodeficiency virus 89.6P challenge be a DNA prime/recombinant modified vaccinia virus ankara vaccine. J Virol
38. Tang Y, Villinger F, Silvja I, Amara RR, Smith JM, Herndon JG, Robinson HL. Slowly declining levels of viral RNA and DNA/Recombinant modified vaccinia virus Ankara-vaccinated macaques with controlled simian–human immunodeficiency virus SHIV-89.6P challenges. J Virol
39. Shiver JW, Fu T-M, Chen L. Replication-incompetent adenoviral vector elicits effective anti-immunodeficiency-virus immunity. Nature
40. Matano T, Kano M, Nakamura H, Takeda A, Nagai Y. Rapid appearance of secondary immune responses and protection from acute CD4 depletion after a highly pathogenic immunodeficiency virus challenge in macaques vaccinated with a DNA prime/Sendai virus vector boost regime. J Virol
41. Enose Y, Ui M, Miyake A, Suzuki H, Uesaka H, Kuwata T, et al
. Protection by intranasal immunization of a nef-deleted, nonpathogenic SHIV against intravaginal challenge with a heterologous pathogenic SHIV. Virology
42. Rose NF, Marx PA, Luckay A, Nixon DF, Moretto WJ, Donahue SM, et al
. An effective AIDS vaccine based on live attenuated vesicular stomatis virus recombinants. Cell
43. Barouch DH, Kuntsman J, Kuroda MJ, Schmitz JE, Santra S, Peyerl FW, et al
. Eventual AIDS vaccine failure in a rhesus monkey by viral escape from cytotoxic T lymphocytes. Nature
44. Jost S, Bernard MC, Kaiser L, Yerly S, Hirschel B, Samri A, et al
. A patient with HIV-1 superinfection. N Engl J Med
45. Atfield M, Allen TM, Yu X, Johnson MW, Agrawal D, Kerber BT, et al
. HIV-1 superinfection despite broad CD8+ T-cell responses containing replication of the primary viruses. Nature
46. Belyakov IM, Hel Z, Kelsall B, Kuznetsov VA, Ahlers JD, Nacsa J, et al
. Mucosal AIDS vaccine reduces disease and viral load in gut reservoir and blood after mucosal infection in macaques. Nat Med
47. Glushakova S, Yi Y, Grivel JC, Singh A, Schols D, De Clercq E, et al
. Preferential coreceptor utilization and cytopathicity by dual-tropic HIV-1 in human lymphoid tissue ex vivo . J Clin Invest
48. Trkola A, Ketas TJ, Nagashima A, Zhao L, Cilliers T, Morris L, et al
. Potent, broad-spectrum inhibition of human immunodeficiency birus type 1 by the CCR5 monoclonal antibody PRO 140. J Virol
49. Clerici M, Shearer GM. A Th1–Th2 switch is a critical step in the etiology of HIV infection. Immunol Today
50. Lehner T, Bergmeier L, Wang Y, Tao L, Mitchell E. A rational basis for mucosal vaccination against HIV infection. Immunol Rev