|Home | About | Journals | Submit | Contact Us | Français|
In this study we investigated the ability of a replication-competent Ad5hr-SIVenv/rev and Ad5hr-SIVgag recombinant priming/gp120 boosting regimen to induce protective immunity in rhesus macaques against pathogenic simian immunodeficiency virusmac251. Immunization of macaques by two sequential administrations of the same recombinants by the same route resulted in boosting and persistence of SIV-specific cellular immune responses for 42 weeks past the initial immunization. Anti-SIV gp120 immunoglobulin G (IgG) and IgA antibodies were induced in secretory fluids, and all macaques exhibited serum neutralizing antibody activity. After intrarectal SIVmac251 challenge, all of the macaques became infected. However, relative protection, as assessed by statistically significant lower SIV viral loads in plasma at both acute infection and set point, was observed in 8 out of 12 immunized non-Mamu-A01 animals. Elevated mean cellular immune responses to Gag and Env, neutralizing antibody activity, and IgG and IgA binding antibody levels were observed in the eight protected macaques. Statistically significant correlations with protective outcome were observed for cellular immune responses to SIV Env and Gag and for SIV gp120-specific IgG antibodies in nasal and vaginal fluids. Two macaques that exhibited the greatest and most persistent viremia control also exhibited strong CD8+ T-cell antiviral activity. The results suggest that a spectrum of immune responses may be necessary for adequate control of viral replication and disease progression and highlight a potential role for nonneutralizing antibodies at mucosal sites.
Despite extensive efforts made to combat human immunodeficiency virus (HIV) infection and AIDS since the discovery of the virus, the number of people infected with HIV and developing the disease worldwide is still increasing rapidly. The need for a vaccine against HIV is now one of the world's greatest public health problems; however, development of a safe and effective HIV vaccine has proved difficult due to several unique challenges presented by the virus. These include difficulty in eliciting broadly reactive neutralizing antibodies, the high variability of the virus, and integration of HIV proviral DNA into the host genome, resulting in latent infection and making achievement of sterilizing immunity nearly impossible (44). Considering recent reports associating either humoral or cellular immune responses with protection against HIV infection or disease progression, it is difficult to define requirements for protective immunity against HIV (1, 4, 9, 23, 24, 32, 38, 40). Accumulating evidence indicates that an ideal HIV vaccine should induce broad humoral, cellular, and mucosal immunity against multiple viral antigens in order to combat infectious viral particles and HIV-infected cells at any point during infection (19, 25, 33, 50). To achieve this goal, many strategies are being investigated, including recombinant viral proteins and peptides, naked DNA, live viral and bacterial vectors, and prime-boost combinations (19).
Adenovirus (Ad) is one of the live viral vectors being developed for use as an HIV vaccine. Ad infects a broad spectrum of human cells, including immature dendritic cells, leading to efficient antigen presentation and causing their maturation without polarizing the T-helper response (22, 53, 54). Because AIDS is mainly a sexually transmitted disease, vaccine-elicited mucosal immunity against HIV is critical. Ad vectors are therefore highly attractive, because they target epithelial cells at mucosal surfaces and can be administered orally and intranasally. Both replication-competent and replication-defective Ad recombinants have been investigated as potential AIDS vaccines.
Replication-defective Ad vectors, long used in gene therapy applications, have been adapted for use as HIV vaccines (5, 46, 51). Recent studies with an E1- and E3-deleted Ad5-SIVgag recombinant to immunize rhesus macaques elicited high-frequency SIV p11C-tetramer-positive cells. Following challenge with pathogenic SHIV89.6P the monkeys exhibited significantly reduced viral burdens and were protected against SHIV-induced disease (46).
We have taken a different approach, using replication-competent Ad recombinants with deletions only in the E3 region. Because of the inability of human Ad to replicate in most mammalian species, our studies initially were carried out with chimpanzees, which are permissive for Ad replication. Replication-competent, E3 region-deleted Ad-HIVenv and -HIVgag/pro recombinants were investigated and shown to elicit cellular immune responses, antibody responses in mucosal secretions, high-titer serum antibodies able to neutralize both T-cell-line-adapted and primary HIV isolates, and significant protective efficacy (20, 21, 30, 31, 43, 55). Chimpanzees immunized with an Ad-HIVenv priming/gp120 boosting regimen were protected against both low- and high-dose HIV challenges, including challenge with a heterologous primary HIV isolate. The protection elicited was shown to be long lasting.
To further develop this approach in a macaque model, we took advantage of an Ad5 host range mutant (Ad5hr) (41) and carried out experiments by using an Ad5hr-SIVsmH4env/rev recombinant shown to replicate in monkey cells in vitro (8). Again by using a recombinant priming/gp120-boosting regimen, we demonstrated that the Ad5hr recombinant also replicated in vivo and elicited SIV-specific cellular immunity and humoral immune responses in serum and secretory fluids (6, 7). This solely envelope-based vaccine achieved a reduction in acute-phase viral burden following intravaginal challenge with pathogenic SIVmac251; however, the viral load began increasing by 8 weeks postchallenge. Accumulating data have shown the potential benefit of incorporating additional viral antigens that elicit strong cellular immune responses into vaccine regimens. Therefore, we added an SIV gag component to the vaccination regimen. We have reported that coadministration of the Ad5hr-SIVsmH4env/rev plus Ad5hr-SIVgag recombinant elicits cellular immune responses to each of the encoded SIV gene products, Env, Gag, and Rev, and that the responses are boosted by a second administration of the same recombinants (35, 52). These recombinant immunizations also prime strong neutralizing antibodies and both IgG and IgA SIV-specific antibodies in secretory fluids (J. Pinczewski, J. Zhao, L. J. Patterson, N. Malkevitch, G. Alvord, and M. Robert-Guroff, unpublished results). Here we have investigated the ability of the dual immunization regimen to protect against an intrarectal challenge with pathogenic SIVmac251 and have explored whether particular immune responses are correlated with the protective outcome. We report improved protective efficacy against a mucosal SIVmac251 challenge, with viremia reduced at both the acute phase and set point of infection. Several immune parameters contributed to the protective outcome.
Twenty-four female rhesus macaques were used in this study; six were Mamu-A01 positive. All were negative for SIV, simian retrovirus type D, and simian T-cell leukemia virus prior to use.
An Ad5hr-SIVsmH4env/rev recombinant encoding the entire SIV gp160 envelope protein and Rev (8) and an Ad5hr-SIV239gag recombinant encoding rev-independent Gag (11, 52) were used as priming immunogens. Native SIVmac251 gp120 protein purified from a productive tissue culture medium (13) was used as the subunit booster immunogen.
Sixteen macaques, including four Mamu-A01-positive animals, were immunized orally and intranasally at 0 weeks and were similarly boosted at 12 weeks with the Ad5hr-SIVenv/rev and -SIVgag recombinants. Each macaque received 5 × 108 PFU of each recombinant orally in 500 μl of phosphate-buffered saline (PBS) and 5 × 108 PFU of each recombinant intranasally in 500 μl of PBS at each immunization. Oral immunizations were given by stomach tube following administration of bicarbonate solution. The macaques were subsequently boosted intramuscularly at 24 and 36 weeks with SIV gp120 (100 μg/macaque) in QS-21 adjuvant. Four control macaques, including one Mamu-A01-positive animal, were mock immunized with Ad5hr vector and adjuvant alone. Four additional controls, including one Mamu-A01-positive macaque, remained naïve.
All animals were challenged intrarectally with SIVmac251 at week 42. The challenge stock was derived by culturing peripheral blood mononuclear cells (PBMCs) from Mamu-A01-positive macaque 561L that had been infected with SIVmac251 by the vaginal route (34). A macaque exposed rectally to 0.5 ml of the third harvest of virus from the tissue culture exhibited a viral burden similar to that of six other macaques that all became infected following intrarectal exposure to 0.5 ml of the second harvest of virus, which was obtained 3 days earlier. Therefore, 0.5 ml of this third virus stock/macaque was used here for the intrarectal challenge.
Peripheral blood and serum samples were routinely obtained over the course of immunization and following challenge. PBMCs were separated from whole blood on Lymphocyte Separation Medium (ICN Pharmaceutical, Inc., Costa Mesa, Calif.), washed twice with PBS, and viably frozen in freezing medium until use. Serum samples were stored at −70°C.
Gamma interferon (IFN-γ) secretion in response to SIV Env, Rev, or Gag peptides was evaluated by ELISPOT assay as previously described (52). Briefly, Gag peptides included 50 20-mer peptides with 10-amino-acid overlaps that spanned the entire SIVmac239 Gag protein and were obtained from the AIDS Research and Reference Reagent Program (National Institute of Allergy and Infectious Diseases, National Institutes of Health). SIV Env and Rev peptides (Advanced BioScience Laboratories, Inc., Kensington, Md.) represented the SIVsmH4 gp160 protein minus the signal peptide and the SIVsmH4 Rev protein, respectively. These peptides were 15 mers with an 11-amino-acid overlap. The Env peptides were tested in three pools: pool A, peptides 1 to 71; pool B, peptides 72 to 142; and pool C, peptides 143 to 214. Data are reported here as the sum of the spot-forming cells (SFC) detected for each pool following subtraction of background spots. The Gag and Rev peptides were used in single pools of 50 and 22 peptides, respectively.
SIV-specific IFN-γ-secreting cells were enumerated by using an ELISPOT kit (U-Cytech, Utrecht, The Netherlands) according to the manufacturer's manual with slight modification. Briefly, 96-well flat-bottom plates were coated overnight with anti-IFN-γ monoclonal antibody MD-1 (U-Cytech), washed, and blocked as described by the manufacturer. Dilutions of thawed PBMC ranging from 1 × 105 to 1.25 × 104 per 100 μl of R-5 medium (RPMI 1640 [Invitrogen, Carlsbad, Calif.] containing 5% fetal calf serum [FCS], 2 mM l-glutamine, 25 mM HEPES, 100 U of penicillin/ml, and 100 μg of streptomycin/ml [all from Gibco BRL, Grand Island, N.Y.]) were transferred to triplicate wells, together with 2 μg of each peptide/ml in the Gag peptide pool or 1 μg of each peptide/ml in the Env or Rev peptide pools. Concanavalin A (5 μg/ml; Sigma) and medium alone were used as positive and negative controls, respectively. Following overnight incubation at 37°C in 5% CO2 the cells were removed, and the wells were washed and incubated with biotinylated rabbit anti-IFN-γ antibody (U-Cytech). After further washing, bound anti-IFN-γ antibody was detected with a gold-labeled anti-biotin IgG (U-Cytech). Incubating the plates with U-Cytech's activator mixture developed spots. The color reaction was stopped by washing with distilled water. The plates were air dried, and spots were counted visually by using an inverted microscope. A positive ELISPOT response was defined as equal to or greater than the number of background spots plus 2 standard deviations and at least 20 SFC/106 PBMC.
Serum neutralizing antibody activity was measured by using T-cell-line-adapted SIVmac251 grown in H9 cells (39). A fixed concentration of virus representing a 50% tissue culture infectious dose (TCID50) was added to eight threefold serial dilutions of heat-inactivated sera (56°C, 30 min) beginning at 1:25, and the mixtures were incubated at 4°C for 30 min. A 25-μl suspension of H9 cells (4 × 106 cells per ml) was added to the virus/serum mixtures, which were then incubated overnight at 37°C in 5% CO2. The H9 cells were washed four times and were resuspended in 125 μl of growth medium consisting of RPMI 1640 containing 10% FCS, 100 IU of penicillin/ml, 100 μg of streptomycin/ml, and 2 mM l-glutamine. Triplicate 25-μl suspensions of the H9 cells were then transferred to wells of a fresh 96-well plate containing 200 μl of growth medium per well. The samples were incubated for 7 days at 37°C in 5% CO2, after which supernatants were harvested and tested for SIV p27 by antigen capture (Beckmann-Coulter, Miami, Fla.). Neutralizing titers are reported as the reciprocal of the serum dilution at which 50% neutralization was achieved.
Neutralizing antibody activity against primary SIV isolates, SIVmac251, and molecularly cloned SIVmac239/nef-open was assessed by using a CEMX174 cell-killing assay as previously described (27). Neutralization was considered positive when at least 50% of cells were protected from virus-induced killing as measured by neutral red uptake. The primary isolate stocks were low-passage derivatives of previously described animal challenge stocks (18) and were produced on human PBMC.
Suppression of SIV replication by CD8+ T cells was evaluated in endogenous and acute CD8AA suppression assays adapted from those described previously (36). CD8+ effector cells were isolated from PBMC by using a CD8 Negative Isolation kit (Dynal Biotech, Inc., Lake Success, N.Y.) and were cultured for 72 h at 37°C in R-10 medium (RPMI 1640 containing 10% FCS, 2 mM L-glutamine, and antibiotics) supplemented with 10 U of human interleukin-2 (Invitrogen)/ml and 1% (vol/vol) rehydrated phytohemagglutinin A (PHA) (Gibco BRL). For the endogenous CD8AA suppression assay, allogeneic target cells were PHA-activated CD4+ T cells obtained by depleting CD8+ T cells from PBMC of an SIVmac251-infected macaque. For the acute CD8AA suppression assay, allogeneic targets were prepared by first depleting the CD8+ T cells from PBMC of a naïve macaque. The remaining CD4+ T cells were then infected in vitro with 2,500 TCID50 of SIVmac251/ml. Effectors and targets were washed four times in PBS and were cocultured at effector-to-target ratios of 4:1, 2:1, 1:1, 0.5:1, and 0.25:1 for 14 days at 37°C in 96-well plates containing 100 μl of R-10 supplemented with 10 U of interleukin-2/ml. In all cases, effectors and targets were separated by 0.2-μm-pore-size Anopore semipermeable membranes (Nunc). On days 4, 7, 10, and 14, 50 μl of supernatant was removed and replaced with fresh medium. SIV replication was assessed by p27 antigen capture assay. Percent suppression was calculated relative to p27 production by control CD4+ targets cultured in the absence of CD8+ effectors.
Virus isolations were carried out by coculturing macaque PBMC with PHA-activated human PBMC. Culture supernatants were collected weekly and were tested for virus expression with SIV p27 antigen capture kits. Virus isolation was scored positive if two or more successive supernatants were positive for p27 production.
SIVmac251 RNA in plasma was quantitated by nucleic acid sequence-based amplification (45). The sensitivity was <2,000 copies of SIV RNA per input volume of plasma (generally 100 μl). Values here are multiplied by 10 and are reported per milliliter of plasma. Viral burdens with <2,000 copies/input volume are recorded as 10,000/ml for purposes of calculation.
The two-tailed Wilcoxon rank sum test was used for comparison of plasma SIV RNA levels and immune responses between groups. The combined differences in two or three paired immune responses were assessed by the Wei-Johnson method. This method was also used in simultaneous comparisons of postchallenge immune responses between immunized and control macaques over successive time points. The Jonckheere-Terpstra test for trend was used in evaluating immune response trends across set point viremia groups. In this exploratory study with correlated tests between small groups no correction for multiplicity was applied.
We have previously reported that immunization with the Ad5hr-SIVenv/rev recombinant elicits both cytotoxic T-lymphocyte activity and T-cell proliferative responses (6, 7, 37). Further, we have shown that both the Ad5hr-SIVenv/rev and Ad5hr-SIVgag recombinants elicit SIV-specific IFN-γ-secreting cells and that this response can be boosted by a second immunization with the same recombinants (35, 52). Here we extend these findings by showing that the Ad recombinant priming/subunit boosting regimen results in cellular immune responses that persist at least 30 weeks past the second Ad recombinant immunization. As shown in Fig. Fig.1,1, IFN-γ secretion in response to both overlapping Env and Gag peptides was sustained out to the time of challenge at 42 weeks. The overall mean responses were low, attributable to the use of cryopreserved rather than fresh PBMC as shown earlier (52). Further, the overlapping peptides, either 15 or 20 mers, were not optimal for antigen presentation to CD8+ T cells, perhaps also contributing to the low level of measured responses. Nevertheless, a clear difference between the immunized and control macaques can be seen over the immunization course (Fig. (Fig.11).
The increase in response to Env peptides following the protein booster immunization at week 24 and a similar tendency for the Gag response to increase (Fig. (Fig.1)1) are of interest. These response patterns may reflect expansion of SIV-specific cells in vivo following induction of CD4+ T-helper cell responses by the protein and/or adjuvant administration. The increased Env responses may also result from SIV Env-specific CD4+ T cells responding to helper epitopes represented in the Env peptide pools.
The boosting regimen with SIV gp120 elicited strong antibody responses. Table Table11 summarizes the responses to SIV gp120 in serum and secretory fluids at week 42, the time of challenge. SIV-specific IgGs were evident in both the systemic and mucosal compartments. Both vaginal and nasal secretions had significant IgG antibody titers against the envelope protein, although antibody titers in serum were 60- to 190-fold higher. In order to avoid any possibility of damaging the rectal mucosa, rectal samples were not taken at the time of challenge. Strong neutralizing antibody activity against T-cell-line-adapted SIVmac251 was also present in serum; however, neutralization of the primary isolates, SIVmac251 and SIVmac239/nef-open, was not observed (data not shown).
Much lower levels of IgA binding antibodies compared to those of IgG were observed in secretory fluids. Nevertheless, SIV gp120-specific IgA antibodies were observed in both nasal secretions and saliva. Vaginal sample volumes were insufficient for analysis, and again rectal samples were not obtained at the time of challenge.
Six weeks following the last immunization, and 42 weeks overall after the initial Ad recombinant immunizations, the macaques were challenged intrarectally with SIVmac251. All animals developed persistent infection, as demonstrated by viral isolation from their PBMCs, monitored up to 20 weeks postchallenge (data not shown). To determine if the immunization regimen resulted in reductions in viral burdens, plasma viremia was evaluated over the year following challenge (Fig. (Fig.2).2). The SIVmac251 challenge stock readily infected the control macaques and resulted in high viral burdens of 3.28 × 107 SIV RNA copies/ml of plasma at the acute phase of infection and 1.65 × 106 at the 12-week set point (Table (Table2).2). The plateau in viral burden, or set point, following acute infection has been shown to be prognostic for eventual disease outcome. Various time points have been used for analysis of this set point, including 6 (49), 12 (47), and 20 weeks (29). Here, viral burdens were stable over 8 to 16 weeks postchallenge, and the week 12 value was the median within this period of time for over half of the macaques. Therefore, viral burden at 12 weeks was taken as the viremia set point. Overall, the immunized macaques had significantly reduced viral burdens (14-fold; P < 0.0001) compared to those of the controls during the acute phase of infection, but significant differences were not achieved at the set point.
The two control Mamu-A01 macaques appeared to have lower viral burdens at set point than the non-Mamu-A01 controls (Fig. 2A and B). This difference did not reach statistical significance (P = 0.071), but the P value was the lowest which could be reached in comparing the two Mamu-A01 controls with the six non-Mamu-A01 controls. When we compared the four immunized Mamu-A01 macaques with the six control non-Mamu-A01 macaques, significant differences of more than 20-fold in viral burdens at both acute infection (P < 0.01) and set point (P < 0.01) were observed (Fig. 2A and C). However, statistically significant differences in viral loads were not observed between immunized and control Mamu-A01 macaques at either the 2- or 12-week time points (Fig. 2B and C). Others have reported significant differences in susceptibility of Mamu-A01-positive and -negative macaques to SIV infection (28, 29, 34). Further, the challenge stock used was the one with which the Mamu-A01 effect was first described (34). With this in mind, and because we could not distinguish immunization effects from Mamu-A01 effects on viremia outcomes, we carried out further analyses on only the non-Mamu-A01 immunized and control macaques.
Analysis of plasma viral loads of the non-Mamu-A01 macaques revealed a significant difference (P < 0.01) between the immunized animals and the controls during the acute phase of infection but not at set point (Table (Table2).2). However, some macaques controlled virus more effectively than others; significant differences became apparent when the macaques were grouped according to their set point viremia. As illustrated in Fig. 2D to F and summarized in Table Table2,2, four of the non-Mamu-A01 immunized macaques exhibited low-set-point viremia with a statistically significant geometric mean 48-fold reduction compared to that of the controls (P < 0.01). Four additional macaques had a statistically significant intermediate reduction in set point viremia (threefold; P < 0.01), while the four remaining macaques had viral burdens similar to those of the control macaques. With further monitoring, of the eight macaques that exhibited reduced set point viremia, two (macaques 773 and 784; Fig. Fig.2D)2D) continued to control viremia a year following challenge.
In earlier studies in which only an envelope-based vaccine was employed, reduced viremia following a vaginal SIVmac251 challenge was observed during the acute phase of infection but the immunized macaques quickly began to lose control of viremia by 8 weeks postchallenge and displayed increased viral burdens over the early levels (6, 7). Here, as reductions in viremia below acute-phase levels were seen during the set point of infection in a subset of the immunized non-Mamu-A01 macaques, we initially asked if cellular responses to the additional SIV Gag immunogen contributed to the improved viremia control. As illustrated in Fig. Fig.3,3, IFN-γ secretion in response to Gag peptides was higher in the immunized macaques with low viral burdens than in other macaque groups. This was most evident when peak responses observed prechallenge were analyzed (Fig. (Fig.3A).3A). A similar trend was observed at the time of challenge (Fig. (Fig.3B).3B). Nevertheless, comparison of the mean Gag responses of the low-viral-load group with the high-viral-load group did not show a statistically significant difference for either peak or time-of-challenge responses.
The cellular immune response to Env peptides was particularly robust in the immunized macaques with intermediate viral burdens at the time of challenge, as were the mean peak responses in both the low and intermediate groups (Fig. (Fig.3).3). The mean peak and time-of-challenge responses to Env in the intermediate group were higher than those in the high-viral-load group, and these differences approached statistical significance (P = 0.057 for both). A comparison of the peak mean response to Env for the eight combined macaques in the low- and intermediate-viral-load groups versus the high-load group was statistically significant at the P = 0.024 level. Thus, the Env response may have contributed to the reductions in set point viral burden in these eight macaques. Comparison of the difference between the higher mean peak responses to both Gag and Env of the eight macaques in the low plus intermediate groups of macaques with the lower responses of the four macaques in the high-viral-load group was of borderline statistical significance (P = 0.051). Of interest, cellular immune responses to Env and Gag were not elevated in the immunized Mamu-A01 macaques compared to those of macaques in the high-viral-burden group and therefore cannot explain their lower viral burdens.
To further explore the role of cellular immune responses in reduction of set point viremia, we examined cellular responses to Gag and Env peptide pools following the intrarectal challenge. Simultaneous comparison of results at the time of challenge through week 4 postchallenge (Fig. 4A and B) showed that all immunized macaques exhibited significantly higher numbers of IFN-γ-secreting cells than controls in response to both Gag and Env peptides (P = 0.0083 and 0.0062, respectively). There was no Mamu-A01 effect on these postchallenge immune responses. We next examined postchallenge cellular immunity with regard to set point viremia, grouping the immunized non-Mamu-A01 macaques as before (Fig. 2D to F) and including the control non-Mamu-A01 macaques (Fig. (Fig.2A).2A). Analysis of trends across these set point viremia groupings revealed highly significant trends for both Gag and Env at weeks 2 and 4 postchallenge (P = 0.0039 and 0.0013 for Gag; P = 0.028 and 0.014 for Env). Figure 4C and D illustrates the elevated number of SIV Gag- and Env-specific IFN-γ-secreting cells, respectively, associated with the low-viremia set point and the decreased responses for the intermediate- and high-viremia groups. An overall summary of the statistically significant cellular responses pre- and postchallenge is given in Table Table33.
Viremia outcomes for the control and immunized Mamu-A01 macaques were low (Fig. 2B and C), placing these macaques in the lowest viremia groups. Nevertheless, a similar trend analysis that included all macaques continued to show a strong correlation of cellular immunity to Gag associated with set point viremia at both weeks 2 and 4 postchallenge (P = 0.0005 and 0.0014, respectively). The association with Env, however, became nonsignificant at both time points (data not shown) when the Mamu-A01 macaques were included.
The vaccine regimen employed incorporated booster immunizations with SIV envelope protein following priming with the Ad-SIV recombinants in order to elicit potent antibody responses. To explore a possible correlation between reduced viral load and SIV-specific antibody, serum samples collected on the day of challenge were first assayed for binding antibody to SIV gp120 and for neutralizing antibody activity. As shown in Fig. Fig.5A,5A, serum samples from macaques with low or intermediate viral burdens exhibited higher antibody binding titers than macaques with a high viral burden, but this difference was not significant (P = 0.099). However, immunized macaques with low viral burdens exhibited higher neutralizing antibody titers than the high-viral-burden macaques (Fig. (Fig.5B),5B), a difference approaching statistical significance (P = 0.057). Thus, while serum antibody to the SIV envelope protein may have contributed to reduction of set point viremia in the low-viral-load group, it is notable that none of the macaque sera at challenge was able to neutralize primary SIVmac251 or SIVmac239/nef-open challenge stocks. Both of these isolates are very difficult to neutralize (14, 26). Finally, as with the cellular responses, serum antibody does not appear associated with the reduced viral loads of the immunized Mamu-A-01 macaques.
Because the SIVmac251 challenge virus was administered mucosally, we also investigated whether SIV-specific antibody in secretions contributed to enhanced protection. As shown in Fig. Fig.6A,6A, although immunized macaques with low viral burdens exhibited the highest mean IgG binding titers to SIV gp120 in vaginal secretions, no significant differences were observed with the other groups of macaques. In contrast, SIV gp120-specific antibody in nasal secretions of macaques with low and intermediate viral burdens exhibited significantly higher antibody levels compared to those of macaques in the high-viral-load group (Fig. (Fig.6B;6B; low versus high, P = 0.029; low plus intermediate versus high, P = 0.028). With regard to SIV gp120-specific IgA antibody in saliva and nasal secretions, few samples were available from the immunized non-Mamu-A01 macaques. Overall, IgA absorbance values were low, and significant differences were not observed between low- and intermediate-viral-load groups and the high-viral-burden group. Analysis of paired nasal IgG and nasal IgA values revealed significant differences between low-viral-burden groups versus high-viral-burden groups (P = 0.029) and low- plus intermediate-viral-burden groups versus high-viral-burden groups (P = 0.032). The significantly elevated nasal antibodies in some of the immunized macaques perhaps reflect the initial intranasal priming by the Ad-SIV recombinants. SIV-specific IgG antibody levels in Mamu-A01 macaques were not significantly higher than those in macaques with the high viral burdens. Although the mean IgA responses in saliva and nasal secretions of the Mamu-A01 group of macaques appear higher than those of the other groups of macaques (Fig. 6C and D), the responses were very variable and no statistically significant differences were obtained.
Examination of additional paired antibody responses revealed that vaginal and nasal IgG responses in the low-viral-load group and in the low- plus intermediate-viral-load groups were significantly different from responses in the high-viral-load group (P = 0.029 and 0.042, respectively). In fact, all three IgG binding antibody responses in serum as well as in secretory fluids were significantly higher in the low-viral-burden group than in the high-burden group (P = 0.029), and the difference between the low plus intermediate group compared to the high group approached statistical significance (P = 0.057). A summary of the statistically significant differences in humoral immune responses is also provided in Table Table33.
Examination of Fig. Fig.2D2D shows that two of the immunized macaques, 773 and 784, controlled viremia better than the other macaques in the low-virus-load group for a year following challenge. In an effort to determine why disease progression was slower in these macaques, we examined CD8+ T-cell suppressor activity in the immunized and control macaques at 28 weeks postchallenge. Both macaques 773 and 784 exhibited strong suppressor activity in both assay systems. Macaque 773 exhibited 60 and 93% suppression by the endogenous and acute assays, respectively, while macaque 784 exhibited even stronger suppressive activities of 95 and 98% in endogenous and acute assays, respectively. The acute suppression values of these 2 macaques were the highest observed among all 12 non-Mamu-A01 macaques (P = 0.030). Overall, the immunized macaques exhibited higher mean endogenous and acute suppressor activity than the controls (Fig. (Fig.7A);7A); however, the difference for endogenous activity was not significant (P = 0.13) and only approached significance for the acute assay (P = 0.059). Further, in examining the macaques grouped by viral burden, no differences were seen in the endogenous assay, and although the low-viral-burden group had the highest mean acute activity, it was not significantly higher than that of the other groups of macaques (Fig. (Fig.7B7B).
For this study, by using an Ad5hr-SIVgag recombinant to prime cellular immune responses in addition to an Ad5hr-SIVenv/rev recombinant, we have extended earlier findings of partial protection against an SIVmac251 intravaginal challenge using an envelope-based vaccine regimen (6, 7) for improved viremia control following an intrarectal challenge with the highly pathogenic SIVmac251 strain. In the earlier study, reduction in viremia was observed during acute infection; however, viral loads in immunized macaques quickly rose by 8 weeks postchallenge above acute-phase levels. Here, in contrast, a subset of immunized macaques exhibited falling, rather than increasing, viral burdens at set point. In addition to significantly reduced acute-phase viral burdens, 8 of 12 non-Mamu-A01 macaques also exhibited statistically significant reductions in viral burden at set point. Further, we have shown that the SIV gag gene in the Ad recombinant, in addition to the env/rev genes, is immunogenic and elicits persistent cellular immune responses. Moreover, antibodies in both serum and secretory fluids were elicited by the prime/boost strategy.
Because the vaccine regimen was similar to the one used previously, with the exception of the added Ad5hr-SIVgag recombinant, we expected that the improved protective efficacy as shown by reductions in set point viremia might be strongly associated with cellular responses to the Gag immunogen. Others have shown that multicomponent vaccine strategies are more efficacious than those relying solely on envelope (10, 12). In fact, a highly significant correlation of postchallenge Gag response with reduced set point viremia was observed (Fig. (Fig.4A4A and Table Table3).3). It is not possible to determine with certainty, however, if this Gag response was solely responsible for the more prolonged viremia control, as the previous experiment used a vaginal exposure and the challenge here was intrarectal. Further experimentation with the same challenge stock administered by the same route will be necessary to resolve this point. The cellular response to Env rather than Gag was a better prechallenge correlate with set point viremia and in macaques with low- and intermediate-viral-burden outcomes was significantly elevated compared to that of macaques with high viral burdens. However, the postchallenge Env response, although statistically significant, was not as strong a correlate as the postchallenge Gag response (Fig. (Fig.4B4B and Table Table33).
The vaccine regimen elicited strong anti-gp120 binding antibodies in serum and mucosal secretions, and several showed good correlation with protective outcome. Neutralizing antibody titers to the T-cell-line-adapted SIVmac251 isolate were elevated in the macaques with low viral burden compared to those of macaques with high viremia, but statistical significance was not reached. Furthermore, the antibodies did not neutralize primary SIV isolates. However, SIV gp120-specific IgG binding antibodies in nasal and vaginal secretions were significantly elevated in the eight macaques with low- and intermediate-viral-burden outcomes, and paired SIV gp120-specific nasal IgG and IgA antibodies also were significantly higher in these eight macaques. Passive administration of neutralizing IgG antibodies has been shown to protect against vaginal and oral mucosal transmission of SHIV isolates (2, 24), although the mechanism of protection has not been elucidated. In these studies, vaginal fluids and saliva did not exhibit neutralizing activity against the SHIV isolates. While secretory IgA is thought to be a major defense mechanism at mucosal sites, IgG can clearly exert protective effects by several potential mechanisms (42). Other than direct neutralization, these include prevention of viral binding or attachment, trapping of virions in mucous, antibody-dependent cell cytotoxicity, and inhibition of transcytosis. One of the advantages of the Ad recombinant vaccine strategy is its ability to target mucosal sites. Here we demonstrate that vaccine-induced antibodies present in secretory fluids are correlated with protective outcome. Whether these antibodies arise from local immune responses or reflect transudation from plasma remains to be determined.
In addition to vaccine-induced immunity, innate immunity can play a role in protection from HIV and SIV infection (15-17, 48). Here we investigated whether CD8+ T-cell antiviral activity was correlated with the reduced viral burdens observed. Unfortunately, cells were not available for assay prior to challenge or in the initial weeks postchallenge. However, we examined this response 28 weeks postchallenge at a time when two of the macaques, 773 and 784, still exhibited strong viremia control. We found that both of these macaques exhibited potent CD8+ T-cell antiviral activity, particularly in the acute assay system, which may have contributed to their continued low-level viremia. It has previously been shown that the acute type of CD8+ T-cell suppressor activity develops after viral infection, whereas the endogenous type of activity occurs in both infected and uninfected individuals (3). We were not able to associate the CD8AA response of the macaques overall with viral burdens, however, and did not observe significant differences in CD8AA levels among the low-, intermediate-, and high-viremia groups of macaques. This is not inconsistent with the fact that at 28 weeks postchallenge, with the exception of macaques 773 and 784, the majority of macaques had begun to escape the level of viremia control observed at set point.
Overall, our study further develops the replication-competent Ad recombinant approach by demonstrating significant protective efficacy against SIVmac251 challenge, with reductions in viremia during both the acute phase and set point of infection. No single immune correlate of protection was discerned. Both cellular immunity to Gag and Env and antibodies in mucosal secretions were associated with development of low to intermediate viral burdens at set point. Furthermore, the contribution of CD8+ T-cell antiviral activity to the continued control of viremia in macaques 773 and 784 cannot be overlooked. We conclude that protection against a pathogenic virus challenge will require broad immunity, activating all components of the immune system.
We thank Ranajit Pal, Advanced BioScience Laboratories, Inc., for preparing the SIVmac251 challenge stock and Nancy Miller, NIAID, DAIDS, for making it available under NIAID contract N01-AI15430 with Advanced BioScience Laboratories, Inc.; Ruth Woodward and John Parrish and the ABL animal care staff for excellent animal care and performance of all animal procedures; and Sharon Orndorff for administrative assistance with animal scheduling. The SIV Gag peptides were provided by the NIAID AIDS Research and Reference Reagent Program.