HIV/AIDS Vaccine Candidates Based on Replication-Competent Recombinant Poxvirus NYVAC-C-KC Expressing Trimeric gp140 and Gag-Derived Virus-Like Particles or Lacking the Viral Molecule B19 That Inhibits Type I Interferon Activate Relevant HIV-1-Specific B and T Cell Immune Functions in Nonhuman Primates
Product# 1103 HIV-1 p24 Monoclonal Antibody
According to UNAIDS, about 36.7 million adults and children were living with human immunodeficiency virus (HIV) worldwide at the end of 2015; however, the number of people newly infected continues to fall, being 38% lower in 2015 than in 2001 (http://www.unaids.org/). These data are the result of the global implementation of preventive and therapeutic strategies. Nevertheless, the development of a vaccine remains among the best hopes for controlling the HIV/AIDS pandemic.
To date, the phase III clinical trial RV144 is the only HIV-1 vaccine efficacy trial that has demonstrated a modest level of protection (31.2%) against HIV-1 infection in humans (1). The RV144 study combined a recombinant canarypox virus vector (ALVAC) expressing HIV-1 antigens from clades B and E as a prime with recombinant HIV-1 gp120 proteins from clades B and CRF01_AE as a boost. Further studies evaluating potential immune correlates of protection have shown that CD4+ T cells, IgG antibodies to the V1/V2 and V3 loops of HIV-1 gp120, and IgG3 antibodies to gp120, together with antibody-dependent cellular cytotoxicity (ADCC) responses, correlated with a decreased risk of HIV-1 infection, whereas IgA antibodies to the envelope protein correlated with decreased vaccine efficacy in the vaccine group (2,–7). These clinical findings for the first time provided evidence that an HIV/AIDS vaccine can prevent HIV-1 infection and highlighted that poxvirus vectors should be considered one of the future HIV/AIDS vaccine candidate vectors.
Among the poxviruses, the highly attenuated vaccinia virus (VACV) strain NYVAC has emerged as a potential HIV/AIDS vaccine candidate (8, 9). Importantly, several NYVAC-based recombinant viruses expressing HIV-1 antigens from clade B or C have been evaluated as HIV/AIDS vaccine candidates in preclinical (10,–28) and clinical (20, 29,–34) studies, with encouraging results for HIV-1-specific T cell and humoral immune responses. However, new strategies have been implemented to improve poxvirus vector immunogenicity (35). Among these strategies, we previously reported that the deletion of immunomodulatory VACV genes such as B19R and/or B8R, which block interferon (IFN) type I and type II pathways (17), or the insertion of the host range VACV C7L gene into NYVAC recombinant vectors expressing clade C(CN54) HIV-1 Env(gp120) and Gag-Pol-Nef antigens as a polyprotein (NYVAC-C) significantly improved the magnitude and quality of HIV-1-specific immune responses in mice (23). Furthermore, the deletion of the B19R and/or B8R gene in NYVAC-C triggered an upregulation of innate immune pathways in infected human monocytes, with robust expression of type I IFNs and IFN-stimulated genes (ISGs), strong activation of the inflammasome, and an upregulation of the expression of interleukin-1β (IL-1β) and proinflammatory cytokines (12). Moreover, the restoration of replication competence of NYVAC-C in human cells by the reincorporation of the K1L and C7L VACV host range genes (NYVAC-C-KC) with or without the removal of the immunomodulatory viral molecule B19 enhanced the cross-presentation and proliferation of HIV-1-specific memory CD8+ T cells in vitro (26). These recombinant vectors selectively activated IFN-induced genes and genes involved in antigen processing and presentation, as determined by microarray analysis of infected human dendritic cells (DCs) (19, 26). At the same time, these constructs maintained limited virus spread in tissues and an attenuated phenotype (26). Additionally, further improved NYVAC recombinant vectors expressing HIV-1 immunogens, such as HIV-1 clade C(ZM96) trimeric soluble gp140 or Gag(ZM96)-Pol-Nef(CN54) as Gag-derived virus-like particles (VLPs), have been shown to have an enhanced HIV-1-specific immunogenicity profile in mice (24) and nonhuman primates (NHPs) (10, 13).
Clinical trials with homologous NYVAC vectors expressing HIV-1 antigens (gp120 Env and the polyprotein Gag-Pol-Nef) have shown a limited immunogenicity profile with a preference for CD4+ T cell activation, which was markedly enhanced when priming was performed with a DNA vector expressing the same HIV-1 antigens (29,–33). Thus, in order to optimize the immunization protocol with NYVAC vectors expressing HIV-1 antigens, various approaches in NHPs have been evaluated, either comparing NYVAC to ALVAC (13) or combining NYVAC with DNA vectors (10), peptides (22), and dendritic cell targets (28), which have all demonstrated promising results.
Here, as part of the Poxvirus T Cell Vaccine Discovery Consortium (PTVDC) from the Collaboration for AIDS Vaccine Discovery (CAVD) of the Bill and Melinda Gates Foundation, we extended our previous studies with NYVAC recombinant vectors (19, 26) and evaluated novel NYVAC recombinant vectors in NHPs. Hence, using a single recombinant NYVAC vector, we combined a set of strategies: restoration of replication competence, expression of novel HIV-1 immunogens (trimeric gp140 and Gag-Pol-Nef as Gag-derived VLPs), and deletion of the B19R immunomodulatory gene (NYVAC recombinant vectors termed NYVAC-C-KC and NYVAC-C-KC-ΔB19R). Thus, NYVAC-C-KC and NYVAC-C-KC-ΔB19R were compared in immunized NHPs to evaluate the HIV-1-specific immunogenicity profile induced by these novel NYVAC recombinant vectors when applied in a prime-boost approach according to a protocol for the delivery of the immunogens, poxvirus, and protein similar to the protocol used in the RV144 phase III clinical trial. The aim was to define the type of HIV-1-specific T cell and humoral immune responses induced by these vectors as a function of immunological markers that have been correlated with HIV-1 immune efficacy. The results showed that replicating NYVAC-C-KC vectors together with a booster of the purified gp120 protein component induced an enhanced activation of HIV-1-specific CD4+ and CD8+ T cell immune responses, together with a strong induction of HIV-1-specific humoral immune responses. These results demonstrate that replicating NYVAC-C-KC vectors triggered relevant HIV-1-specific immunological properties as potential correlates of protection, with the VACV B19 protein exerting some control of immune functions and supporting the use of these novel NYVAC-C-KC recombinant vectors as HIV/AIDS vaccine candidates.
We previously described the generation and characterization of nonreplicating NYVAC vectors expressing clade C HIV-1 trimeric gp140 or Gag-Pol-Nef as a polyprotein processed into Gag-derived VLPs and their immune behavior in mice (24) and in NHPs (13). Since these vectors do not replicate in human cells, it was important to define whether novel replication-competent NYVAC-KC vectors could be more immunogenic as a function of higher levels of antigen expression during infection. Analysis of the expression of HIV-1 gp140 in human HeLa cells by Western blotting is shown in Fig. 1A. Clearly, higher levels of expression at late times postinfection were observed in the NYVAC-C-KC vectors than those of the parental NYVAC-C vector. These differences were also noticeable after analysis of the virus plaque size phenotype in cultured BSC-40 cells. The NYVAC-C-KC vectors have a larger plaque size, with or without B19R, than did the parental wild-type NYVAC (NYVAC-WT) or NYVAC-gp140 vector, consistent with a higher replication capacity of NYVAC-KC vectors than of parental NYVAC (Fig. 1B). Analysis of the innate immune response elicited in human macrophages (THP-1) infected with the NYVAC-C-KC vectors by real-time PCR (RT-PCR) showed that compared to NYVAC-KC-gp140, NYVAC-KC-gp140-ΔB19R triggered a significant upregulation of the mRNA levels of type I IFN (IFN-β), macrophage inflammatory protein 1α (MIP-1α), IL-8, and IL-1β (Fig. 1C), indicating some differential innate immune responses between these vaccines.
To define the HIV-1-specific immune responses induced by these two novel optimized replication-competent NYVAC vectors expressing HIV-1 clade C immunogens and containing a deletion in the VACV B19R gene, and to further evaluate whether these responses could be relevant for the control of HIV-1 replication, we analyzed a number of immune parameters that were previously described to be potential indicators of correlates of protection (2,–7). The protocol for immunization of NHPs is depicted in Fig. 2 and follows an immunization scheme similar to the one described for the RV144 efficacy clinical trial (1), allowing a head-to-head comparison of the immunogenicities induced by both NYVAC-C-KC vectors. Figure 2A summarizes the 2 immunization groups included in the study (see Materials and Methods for details). Rhesus macaques were divided into two groups of eight animals each: group 1 received an intramuscular (i.m.) administration of 2 × 108 PFU of NYVAC-C-KC in the upper right arm (1:1 mixture of 1 × 108 PFU of NYVAC-KC-gp140 and 1 × 108 PFU of NYVAC-KC-Gag-Pol-Nef), and group 2 received a similar administration but with NYVAC-C-KC-ΔB19R (1:1 mixture of 1 × 108 PFU of NYVAC-KC-gp140-ΔB19R and 1 × 108 PFU of NYVAC-KC-Gag-Pol-Nef-ΔB19R). The first two administrations with the NYVAC-C-KC vectors were given at weeks 0 and 4; thereafter, at weeks 12 and 24, animals received i.m. booster doses of the corresponding NYVAC-C-KC vectors in the upper right arm, while in the left arm, animals received booster immunization with a bivalent clade C gp120 protein (1:1 mixture of 50 μg of 1086 gp120 and 50 μg of TV1 gp120 proteins adjuvanted with MF59). Immunological monitoring of HIV-1-specific T and B cell immune responses induced by both groups was performed by using peripheral blood mononuclear cells (PBMCs), serum samples, or rectal mucosal samples isolated at weeks 0, 6, 14, 26, and 36 (at the beginning of the study; 2 weeks after the second, third, and fourth immunizations; and at the end of the study, respectively), according to analytical approaches similar to those described previously for NHPs (13) (Fig. 2B).
The total magnitude of HIV-1-specific T cell immune responses induced by the NYVAC-C-KC and NYVAC-C-KC-ΔB19R vectors was measured at weeks 0, 6, 14, 26, and 36 by an IFN-γ enzyme-linked immunosorbent spot (ELISpot) assay. The results showed that the mean spot-forming unit (SFU) values induced in both immunization groups were low until week 14 and clearly peaked at week 26 (2 weeks after the completion of the prime-boost immunization protocol) (Fig. 3A). As expected, the SFU values declined at week 36. There were no significant statistical differences in the levels of IFN-γ-positive T cells between the two groups at any time point. However, at week 26 of immunization, NYVAC-C-KC elicited a trend toward higher numbers of SFUs, while at week 36 of immunization, NYVAC-C-KC-ΔB19R induced a trend toward higher responses.
Furthermore, we also analyzed the magnitude of HIV-1-specific T cell immune responses induced by the NYVAC-C-KC and NYVAC-C-KC-ΔB19R vectors by an intracellular cytokine staining (ICS) assay. Thus, at weeks 6, 14, 26, and 36, we measured the percentages of HIV-1-specific CD4+ and CD8+ T cells secreting IFN-γ, IL-2, and/or tumor necrosis factor alpha (TNF-α) after stimulation of PBMCs obtained from each immunized rhesus macaque with peptide pools that spanned the HIV-1 Env, Gag, Pol, and Nef antigens present in the inserts. As shown in Fig. 3B and andC,C, the total magnitude of HIV-1-specific CD4+ and CD8+ T cell responses peaked at week 26, with a decline at week 36, similar to the results obtained with the IFN-γ ELISpot assay. There were no statistically significant differences in the magnitudes of HIV-1-specific CD4+ T cell responses induced in both immunization groups at any time point (Fig. 3B). However, at weeks 14 and 36 of immunization, NYVAC-C-KC-ΔB19R induced a trend toward higher HIV-1-specific CD4+ T cell responses, while at week 26 of immunization, NYVAC-C-KC induced a trend toward higher responses. On the other hand, the analysis of the magnitude of the total HIV-1-specific CD8+ T cell responses (Fig. 3C) showed that immunization with NYVAC-C-KC-ΔB19R induced a trend toward increased levels over NYVAC-C-KC at all time points, but the differences were not significant.
Moreover, the analysis of the cytokine responses generated by both immunization groups revealed that in agreement with the total magnitude of HIV-1-specific CD4+ T cell responses, at week 36, immunization with NYVAC-C-KC-ΔB19R induced a trend toward higher magnitudes of responses of HIV-1-specific CD4+ T cells producing TNF-α (Fig. 4B) or IL-2 (Fig. 4C), but the differences were not significant. However, at week 26 of immunization with NYVAC-C-KC, a trend toward higher magnitudes of responses of HIV-1-specific CD4+ T cells producing IFN-γ (Fig. 4A), TNF-α (Fig. 4B), or IL-2 (Fig. 4C) was observed.
In the case of HIV-1-specific CD8+ T cells, and in agreement with the total values for CD8+ T cells shown in Fig. 3C, immunization with NYVAC-C-KC-ΔB19R induced a trend toward higher magnitudes of responses of HIV-1-specific CD8+ T cells producing IFN-γ (Fig. 4D), TNF-α (Fig. 4E), or IL-2 (Fig. 4F) at most of the time points, although there were no statistical differences between groups.
Since the RV144 phase III clinical trial showed that IgG antibodies against the V1/V2 and V3 regions of HIV-1 gp120 correlated with a decreased risk of HIV-1 infection (2,–7), we next analyzed the HIV-1-specific humoral immune responses elicited after immunization with the two NYVAC-C-KC vectors. Thus, we quantified the total binding of IgG antibody against clade C HIV-1 gp140 and gp120 and murine leukemia virus (MuLV) gp70-scaffolded V1/V2 proteins in individual serum samples obtained from each immunized rhesus macaque at weeks 6, 14, 26, and 36 (Fig. 5). The results showed that there was a strong induction of IgG binding antibodies to the different HIV-1 envelopes tested in all animals when the NYVAC-C-KC vectors were administered together with the gp120 protein component, although there were no statistically significant differences between the two groups. In these immunization groups, maximum IgG antibody levels were detected at week 26 (2 weeks after the second booster dose with NYVAC-C-KC/gp120 protein), and by week 36, IgG antibody levels declined but were still maintained at levels superior to those with two doses of NYVAC-C-KC vectors (week 6). Immunization with NYVAC-C-KC elicited a trend toward higher levels of binding IgG antibodies against the consensus HIV-1 gp140 protein of group M (Fig. 5A), gp120 from isolate TV1 (Fig. 5B), gp120 from isolate 1086 (Fig. 5C), and the MuLV gp70-scaffolded V1/V2 protein (Fig. 5D), although there were no statistical differences between groups.
Similarly to the results shown in Fig. 5, cross-clade binding IgG antibody levels induced in both immunization groups were low by week 6 but increased markedly by week 26 and waned with time at week 36 but remained at levels similar to or higher than those at week 6 (data not shown). Overall, both immunization groups induced similar high levels of cross-clade binding IgG antibodies against consensus HIV-1 gp140 proteins of clade A and clade B, clade A gp140(00MSA), and clade B gp140(JRFL), with no statistically significant differences between the groups. A trend toward higher levels of cross-clade binding IgG antibodies for NYVAC-C-KC was noted at week 14.
In addition, rectal IgG binding responses against the group M HIV-1 gp140 consensus and clade C gp120 proteins from isolates 1086 and TV1 were detected in both immunization groups, with a peak response at week 26 and a decline at week 36 (Fig. 5E), although differences were not significant. The rates of responders at week 26 were higher in the NYVAC-C-KC-ΔB19R immunization group, while at week 36, the rates were higher in the NYVAC-C-KC immunization group.
Because the RV144 phase III clinical trial showed that high levels of binding plasma IgA antibodies to HIV-1 Env correlated directly with an increased risk of infection (4, 5), we next analyzed the binding IgA antibodies induced after immunization with the two NYVAC-C-KC vectors in serum and in rectal mucosa (data not shown). The results for serum showed that very low levels (barely above the background level for most time points) of binding IgA antibodies against the consensus HIV-1 gp140 protein of clade C, gp120 from isolate TV1, and the MuLV gp70-scaffolded protein were induced in both immunization groups. Furthermore, similar negligible levels of binding IgA antibodies to consensus HIV-1 gp140 proteins of clades A and B and group M, clade A gp140(00MSA), and clade B gp140(JRFL) were observed for the two vaccinated groups. There were no rectal IgA binding antibodies (data not shown).
Since ADCC can mediate nonneutralizing antibody effector functions dependent on Fc receptor engagement and was suggested to be a correlate of protection in the RV144 phase III clinical trial (4, 36), we next analyzed the ability of NYVAC-C-KC and NYVAC-C-KC-ΔB19R to induce ADCC responses against HIV-1 gp120 proteins from isolates TV1 and 1086 at weeks 0, 26, and 36 in individual serum samples obtained from each immunized rhesus macaque (Fig. 6A). The results showed that both immunization groups elicited similar titers of ADCC against HIV-1 gp120 from isolates TV1 and 1086 at week 26, with a decline at week 36. The titers against target cells coated with the HIV-1 gp120 protein of the TV1 isolate were higher than those obtained with HIV-1 gp120 of the 1086 isolate. Although there were no statistical differences between plasma samples from animals immunized with the two NYVAC-C-KC vectors, immunization with NYVAC-C-KC-ΔB19R induced a trend toward higher ADCC titers against the HIV-1 gp120 protein of the TV1 isolate at week 36 than with NYVAC-C-KC. On the other hand, NYVAC-C-KC elicited higher-level ADCC responses against HIV-1 gp120 from isolate 1086.
The induction of HIV-1-neutralizing antibodies (NAbs) is of special importance for the control of HIV-1 (37). Thus, we next analyzed the capacity of NYVAC-C-KC and NYVAC-C-KC-ΔB19R to elicit NAbs against a panel of different tier 1 and 2 HIV-1 isolates at weeks 26 and 36 in individual serum samples obtained from each immunized rhesus macaque (Fig. 6B). The results showed that neutralization titers induced in both immunization groups were higher at week 26 than those at week 36, with overall no significant differences between both groups in neutralization against the different tier 1 and 2 HIV-1 isolates. However, some minor differences were observed. By using the more sensitive A3R5 assay, the results showed that NYVAC-C-KC-ΔB19R induced a trend toward higher titers of NAbs against HIV-1 gp120 of TV1 (tier 2, clade C) and Ce1086 (tier 2, clade C) at weeks 26 and 36, although there were no statistical differences compared to immunization with NYVAC-C-KC. The TZM-bl assay showed that similar levels of NAbs against a wide range of tier 1 HIV-1 isolates were induced in both immunization groups.
Next, to define if there is a correlation between the two vectors in the elicitation of CD4/CD8 T cell responses versus the induced antibody responses, we performed a statistical correlation analysis at week 26. Figure 7 shows correlation coefficients between T cell responses (ICS assays of both CD4/CD8 cells) and antibody responses (binding antibody multiplex assays [BAMAs], ADCC GranToxiLux [GTL] assays, and NAb assays). In general, CD4+ T cell responses showed a higher number of positive correlations with antibody responses than did CD8+ T cell responses. For CD4+ T cell responses, NYVAC-C-KC-ΔB19R showed a more favorable trend toward a correlation with antibody responses than did NYVAC-C-KC. Most of the correlations failed to be significant (P < 0.05) due to the small sample size, except for those between CD4+ cells and ADCC and NAbs in the NYVAC-C-KC-ΔB19R group.
After the modest efficacy obtained in the RV144 phase III clinical trial, the scientific community has focused on the generation and optimization of HIV/AIDS vaccine candidates to afford protection to a large percentage of vaccinees.
Poxviruses, and in particular the highly attenuated VACV strains MVA and NYVAC, have being widely used as HIV/AIDS vaccine candidates and are included as components of some of the clinical trials planned within the next 5 years on the basis of lessons learned from recent trials (8, 9, 38). However, despite the safety and immunogenicity profiles exhibited by these attenuated VACV strains, more efficient vectors that enhance the magnitude, breadth, polyfunctionality, and durability of the immune responses to exogenously expressed antigens are desirable. Different strategies have been used for poxviruses to achieve this purpose (35). One of them is the deletion of VACV immunomodulatory genes that are still present in the vector genome and whose gene products may be predicted to interfere with the optimal induction of cellular and humoral responses. In this sense, we previously reported enhanced immunogenicity in mice with MVA- and NYVAC-based recombinants by single or multiple deletions of VACV immunomodulatory genes such as B8R and/or B19R (17), A46R (25), C12L (39), C6L and/or K7R (40, 41), A41L and/or B16R (42), N2L (43), and F1L (44). Another strategy is the generation of new vectors with replication competence in human cells in order to increase the timing and levels of expression of the heterologous antigen in the host. To date, some replication-competent recombinant VACV-based vaccines have been used for various infectious diseases, demonstrating that they are able to elicit potent humoral and cell-mediated immune responses and to confer long-lasting protection while maintaining a safety phenotype (23, 45,–47).
In this study, we aimed to define best-in-class protocols of immunization following lessons learned from the RV144 efficacy clinical trial with poxvirus vectors. We have evaluated and compared the immunogenicities in NHPs of two replication-competent NYVAC-C recombinants generated by restoration of replication competence in human cells after the reincorporation of the K1L and C7L host range genes (NYVAC-C-KC) or by a combination of restoration of replication competence and the removal of the immunomodulatory viral molecule B19 (NYVAC-C-KC-ΔB19R) that blocks the type I IFN response. Through extensive analysis of HIV-1-specific immune responses, including measurement of the levels of several immune markers that have been associated with immune efficacy in the RV144 phase III clinical trial (2,–7), we observed that both the NYVAC-C-KC and NYVAC-C-KC-ΔB19R viruses were able to induce a broad spectrum of HIV-1-specific B and T cell responses against the HIV-1 antigens. Thus, the magnitudes of the vaccine-induced T cell immune responses were high in both immunization groups, and the deletion of the immunomodulatory gene B19R on the NYVAC-C-KC recombinant vector improved to some extent the magnitude of the HIV-1-specific CD4+ and CD8+ T cell immune responses. Moreover, a low level of IgA binding antibodies, a high level of binding IgG antibodies in serum and rectal mucosal samples, ADCC responses, and NAbs against different tier 1 and 2 HIV-1 isolates were elicited by NYVAC-C-KC and NYVAC-C-KC-ΔB19R, with some minor differences between the two groups. Interestingly, at the time of the peak response (week 26), the higher levels of HIV-1-specific CD4+ and CD8+ T cells elicited by NYVAC-C-KC and NYVAC-C-KC-ΔB19R coincided with similar high levels of antibodies to Env as well as with high HIV neutralization titers of ADCC at the same time, suggesting a correlation between enhanced CD4+/CD8+ T cell responses and B cell responses and pointing out that CD4+ and/or CD8+ T cells might influence the B cell response. In fact, a correlation analysis (Fig. 7) showed that the percentages of HIV-1-specific CD4+ T cell responses induced by the NYVAC-C-KC-ΔB19R vector have a more favorable trend toward a correlation with antibody responses than with NYVAC-C-KC. However, the correlation between CD8+ T cell and B cell responses was less pronounced than that for CD4+ T cells.
How does the B19R gene operate? It has been shown that B19 is a type I IFN binding protein that acts both in solution and when bound to glycosaminoglycans to inhibit the action of a wide range of IFNs from different species, sequestering and suppressing efficiently the function of type I IFNs (48, 49). In cells infected with a NYVAC vector lacking B19, we have shown enhanced IFN-α production, maturation, and expression of IFN-induced pathways, IFN-regulated transcription factors, as well as multiple inflammatory cytokines in human DCs (12). Immunization of mice with NYVAC lacking B19R and expressing Env, Gag, Pol, and Nef clade C (CN54) HIV-1 antigens improved the magnitude and quality of HIV-1-specific CD8+ T cell adaptive immune responses and impacted their memory phase, changing the contraction, the memory differentiation, the magnitude, and the functionality profile, while for B cell responses, B19 had no apparent effect on antibody levels (17). The immunomodulatory effects of B19 were more pronounced in mice than in NHPs. Here, in NHPs, we noted the influence of B19 by a trend toward slightly higher levels of T cells and of some cytokines triggered by NYVAC-C-KC-ΔB19R than those elicited by NYVAC-C-KC at some time points after boost immunization. The overall effect of B19 on Env-specific binding IgG antibodies, IgA, NAbs, and ADCC was not different between the NYVAC-C-KC vectors. Thus, it seems that the main effect of B19 is on T cell responses, and the role of B19R in blocking type I IFN could explain the small differences in immunogenicity, in particular in T cell responses, observed in macaques inoculated with NYVAC-C-KC versus NYVAC-C-KC-ΔB19R.
In our previous study in NHPs, we demonstrated that a nonreplicating NYVAC vector expressing the same HIV-1 antigens and the same protocol as those used in this study was superior to the ALVAC vector used in the RV144 phase III clinical trial (13). Comparison of the immunogenicity profiles triggered by the NYVAC-C-KC vectors (this study) with that of the nonreplicating NYVAC vector (13) using the same standardized immune assays revealed an overall higher magnitude of HIV-1-specific T cell and humoral immune responses induced by the NYVAC-C-KC vectors. This is likely due to more antigen expression from the NYVAC-C-KC vectors (Fig. 1), as the virus replicates more efficiently in NHPs than does NYVAC lacking the VACV host range genes K1L and C7L. These results correlated with our previously reported in vitro findings, which demonstrated that NYVAC-KC vectors acquired specific biological characteristics over nonreplicating NYVAC, giving replication competency in human keratinocytes and dermal fibroblasts, the activation of selective host cell signal transduction pathways, and higher-level virus spread in tissues while maintaining a highly attenuated phenotype (19, 26). While antigen exposure was defined recently as a determinant of HIV-1 broadly neutralizing antibody (bnAb) induction (50), the question remains as to how to best induce bnAbs by vaccination approaches. Using the protocols described here, we have been able to trigger a broad spectrum of T and B cell immune responses. A balance of T and B cell immune responses has been correlated with levels of protection by HIV vaccines in NHP models of HIV infection (51). A number of HIV/simian immunodeficiency virus (SIV) vaccine protection studies in NHPs have revealed a role for humoral responses as potential correlates of immunity (52,–54), broadly including the types of immune responses elicited by this vaccine regimen. However, it remains to be defined whether the responses elicited by NYVAC-C-KC vectors confer protection against a simian-human immunodeficiency virus (SHIV) challenge and/or are able to induce bnAbs or other potentially protective immune responses in NHPs in a prophylactic phase I clinical trial.
The immunological profiles induced by the NYVAC-C-KC vectors, including the induction of CD4+ and CD8+ T cell immune responses; the high levels of anti-Env binding antibodies to both gp140 and V1/V2 sites; and enhanced ADCC responses and neutralizing capacity, together with low levels of IgA antibodies, are all potential correlates of protection, suggesting that these viral vectors could be considered improved HIV/AIDS vaccine candidates.
The VACV K1L and C7L genes were inserted into poxvirus recombinant vectors NYVAC-gp140 and NYVAC-Gag-Pol-Nef, as previously described (24), to generate the replication-competent recombinants NYVAC-KC-gp140 and NYVAC-KC-Gag-Pol-Nef, respectively, according to the same methodology as the one described previously (19). Thus, the replication-competent NYVAC-C-KC vector consists of two NYVAC-KC viruses that express different clade C HIV-1 antigens under the control of the same synthetic early/late poxvirus promoter (55), one (NYVAC-KC-gp140) expressing Env gp140 from strain 96ZM651 and the other (NYVAC-KC-Gag-Pol-Nef) expressing Gag from strain 96ZM651 and Pol and Nef from strain CN54. A description of the HIV-1 gp140 and Gag-Pol-Nef sequences included in the NYVAC-KC recombinant vectors was previously reported (10, 24). Next, the VACV B19R gene (B18R in the WR strain) was deleted from the replication-competent recombinant vectors NYVAC-KC-gp140 and NYVAC-KC-Gag-Pol-Nef to generate the replication-competent recombinant NYVAC-KC-gp140-ΔB19R and NYVAC-KC-Gag-Pol-Nef-ΔB19R vectors, respectively, according to the same methodology as the one described previously (19). Once generated, the resulting replication-competent NYVAC-KC-gp140, NYVAC-KC-Gag-Pol-Nef, NYVAC-KC-gp140-ΔB19R, and NYVAC-KC-Gag-Pol-Nef-ΔB19R recombinant viruses were expanded in large cultures of primary chicken embryo fibroblast (CEF) cells followed by virus purification through two 36% (wt/vol) sucrose cushions. Titers were determined by plaque immunostaining in BSC-40 cells. The expression of HIV-1 antigens by Western blotting and analysis of plaque size were done as previously described (24). Infection of human macrophages (THP-1 cells) and RNA analysis of IFN-β, MIP-1α, IL-8, and IL-1β by quantitative real-time PCR were done as previously described (40, 41). For simplicity of terminology, we refer to the combined mixed inoculation of NYVAC-KC-gp140 plus NYVAC-KC-Gag-Pol-Nef as NYVAC-C-KC and that of NYVAC-KC-gp140-ΔB19R plus NYVAC-KC-Gag-Pol-Nef-ΔB19R as NYVAC-C-KC-ΔB19R.
For the immunizations performed in this study, we used a bivalent HIV-1 gp120 protein containing a mixture of TV1 gp120 and 1086 gp120, both from clade C. These proteins were expressed from stably transfected Chinese hamster ovary (CHO) cell lines, purified, and characterized as previously described (56).
Animals used in this study (designated AUP513) were outbred adult male Indian rhesus macaques (Macaca mulatta), which were housed, fed, given environmental enrichment, and handled at the animal facility of Advanced BioScience Laboratories (ABL), Inc. (Rockville, MD). The study protocol and animal care are in accordance with the standards of the Association for the Assessment and Accreditation of Laboratory Animal Care International (AAALAC International). The study was approved by the ABL, Inc., Institutional Animal Care and Use Committee in accordance with international guidelines. All procedures were carried out under anesthesia (ketamine administered at 10 mg/kg of body weight) by trained personnel under the supervision of veterinary staff, and all efforts were made to ameliorate animal welfare and to minimize animal suffering. During the study, macaques were observed for general behavior, clinical symptoms, and local reactions at the injection sites twice daily the week after immunizations. When animals were sedated for immunizations or sample collections, body weight and temperature were measured. The age of the animals ranged between 2.5 and 2.8 years, with a mean of 2.6 years, and the weight range was between 3.1 and 5.7 kg, with a mean of 3.8 kg. At selected time points, a physical examination was performed by a veterinarian, and clinical chemistry and hematology parameters were measured. All rhesus macaques were negative for tuberculosis, simian retrovirus (SRV), simian T cell leukemia virus 1 (STLV-1), herpesvirus B, SIV, measles virus, and poxvirus immunogens prior to the study and also had negative fecal cultures for salmonella, shigella, campylobacter, and yersinia. Furthermore, animals were immunologically naive for the vaccine components. A total of 16 animals were divided into two groups of 8 animals each.
Two immunization groups of eight rhesus macaques were included in the study protocol (termed AUP513), which was also reported previously (13). Group 1 received two immunizations with NYVAC-C-KC (containing a 1:1 mixture of NYVAC-KC-gp140 and NYVAC-KC-Gag-Pol-Nef) at weeks 0 and 4 and was boosted with two immunizations of NYVAC-C-KC plus bivalent HIV-1 gp120 proteins from clade C (containing a 1:1 mixture of TV1 gp120 and 1086 gp120) at weeks 12 and 24. Group 2 received two immunizations with NYVAC-C-KC-ΔB19R (containing a 1:1 mixture of NYVAC-KC-gp140-ΔB19R and NYVAC-KC-Gag-Pol-Nef-ΔB19R) at weeks 0 and 4 and was boosted with two immunizations of NYVAC-C-KC-ΔB19R plus bivalent HIV-1 gp120 proteins from clade C (TV1 gp120 and 1086 gp120) at weeks 12 and 24. All immunizations with the replication-competent NYVAC-C-KC vectors and proteins were given i.m. in the deltoid muscle of the upper right arm for the replication-competent NYVAC-C-KC vectors and in the opposite site, the upper left arm, for the gp120 proteins. A dose of 1 × 108 PFU of each replication-competent NYVAC-C-KC vector (NYVAC-KC-gp140 plus NYVAC-KC-Gag-Pol-Nef in the case of NYVAC-C-KC and NYVAC-KC-gp140-ΔB19R plus NYVAC-KC-Gag-Pol-Nef-ΔB19R for NYVAC-C-KC-ΔB19R; 2 × 108 PFU of total virus in 1.0 ml of Tris-buffered saline) and 50 μg each of the TV1 and 1086 clade C HIV-1 gp120 proteins (with adjuvant MF59; 100 μg of total protein in 1.0 ml) were used for each immunization. At weeks 0, 6, 14, 26, and 36 (at the beginning of the study; 2 weeks after the second, third, and fourth immunizations; and at the end of the study, respectively), PBMCs and serum samples were obtained from each immunized animal, and HIV-1-specific T cell and humoral immune responses were analyzed. Blood samples were processed according to current procedures (57).
Overlapping peptides (15-mers with 11 amino acids overlapping) spanning the Env, Gag, Pol, and Nef HIV-1 clade C regions were matched to the inserts expressed by NYVAC-C-KC and NYVAC-C-KC-ΔB19R. Peptides used in the IFN-y ELISpot and ICS assays for T cell stimulation were grouped into nine different peptide pools (Env-1, Env-2, Env-3, Pol-1, Pol-2, Gag-1, Gag/Pol, Gag-2/Pol, and Nef), with about 60 peptides per pool.
The overall strength of the HIV-1-specific T cell immune responses induced was analyzed by an IFN-γ ELISpot analysis using freshly isolated PBMCs obtained from each immunized rhesus macaque, as previously described (10). Briefly, PBMCs were stimulated in triplicate with HIV-1 Env, Gag, Pol, and Nef peptide pools at 1 μg/ml or with phytohemagglutinin (PHA) (2.5 μg/ml) as a positive control, while the addition of medium only served as a negative control, according to protocols described previously (10). SFUs were counted as a measure of the magnitude of HIV-1-specific T cell responses.
The HIV-1-specific CD4+ and CD8+ T cell immune responses induced were analyzed by polychromatic ICS using PBMCs obtained from each immunized rhesus macaque, as previously described (13, 57). In short, cryopreserved PBMCs were thawed and rested overnight in R10 medium (RPMI 1640 [BioWhittaker, Walkersville, MD], 10% fetal bovine serum [FBS], 2 mM l-glutamine, 100 U/ml penicillin G, 100 μg/ml streptomycin) with 50 U/ml Benzonase (Novagen, Madison, WI) in a 37°C incubator with 5% CO2. The next day, 1 × 106 to 3 × 106 cells were stimulated in 96-well plates with the corresponding HIV-1 Env, Gag, Pol, and Nef peptide pools (2 μg/ml) in the presence of brefeldin A (10 μg/ml; BD Biosciences, San Jose, CA) for 6 h. Negative controls received an equal concentration of dimethyl sulfoxide (DMSO) instead of peptides. For staining, cells were fixed, permeabilized, and stained with different antibodies. The following fluorescence-labeled antibodies were used: CD3-Cy7-allophycocyanin (APC) (clone SP34.2; BD Biosciences), CD4-BV421 (clone OKT4; BioLegend), CD8-BV570 (clone RPA-T8; BioLegend), IFN-γ–APC (clone B27; BD Biosciences), IL-2–phycoerythrin (PE) (clone MQ1-17H12; BD Biosciences), and TNF-α–fluorescein isothiocyanate (FITC) (clone Mab11; BD Biosciences). An Aqua Live/Dead kit (Invitrogen, Carlsbad, CA) was used to exclude dead cells. All antibodies were previously titrated to determine the optimal concentration. Samples were acquired on an LSR II flow cytometer and analyzed by using FlowJo version 9.8 (TreeStar, Inc., Ashland, OR).
HIV-1-specific binding IgG and IgA antibodies to HIV-1 gp120/gp140 proteins and MuLV gp70-scaffolded V1/V2 were measured in serum and rectal weck elutions from each immunized rhesus macaque by using a BAMA, as previously described (6, 7, 58). HIV-1 antigens used to analyze the levels of IgG or IgA binding antibodies included (i) consensus gp140 proteins of clade A (A1.con.env03 140 CF), clade B (B.con.env03 140 CF), clade C (C.con.env03 140 CF), and group M (Con S gp140 CF); (ii) primary Env variants, including 1086 gp120 (clade C), TV1 gp120 (clade C), JRFL gp140 (clade B), and MSA4076 gp140 (clade A1; also known as OOMSA); and (iii) MuLV gp70-scaffolded V1/V2 (from clade B; gp70_B.CaseA2 V1/V2). Different plasma serial dilutions were made, and the primary readout was the mean fluorescence intensity (MFI). For IgA, results are expressed as the MFI minus the background MFI at a dilution of 1/80. For IgG, MFI data were transformed to values defining the area under the curve (AUC), which resemble the integral of the curve for the MFI plotted against serial dilutions. Moreover, the specific activity of rectal mucosal IgG binding antibodies was calculated by dividing the antibody titer by the total IgG concentration. Values 3-fold above the values for the baseline visit were considered positive, and the cutoff was established by using negative serum samples. All assays were performed under good clinical laboratory practice (GCLP)-compliant conditions.
The ADCC activity against recombinant gp120 proteins from isolate 1086 or TV1 (clade C) in serum from each immunized rhesus macaque was analyzed according to the ADCC GranToxiLux (GTL) procedure, as previously described (59, 60). The results of the GTL assay were considered positive if the percentage of granzyme B activity after background subtraction was ≥8% for the infected target cells (61). The log10 titer of the ADCC antibodies present in plasma was calculated by interpolating the log reciprocal of the last plasma dilution that yielded positive granzyme B activity (≥8%). The GTL ADCC assay was performed under GCLP-compliant guidelines.
NAbs against HIV-1 in serum from each immunized rhesus macaque were analyzed by using the TZM-bl cell assay and the more sensitive A3R5 cell assay, as previously described (62,–65). For the TZM-bl assay, the following isolates carrying Envs were used: BaL.26 (tier 1B, clade B), Ce1086 (tier 2, clade C), Ce1176 (tier 2, clade C), Ce2010 (tier 2, clade C), Du151.2 (tier 2, clade C), MN.3 (tier 1A, clade B), MW965.26 (tier 1A, clade C), SF162.LS (tier 1A, clade B), TH023.6 (tier 1, clade AE), TV1.21 (tier 2, clade C), and MuLV (as a negative control). For the A3R5 cell assay, we used viral infectious molecular clones (IMCs), which encoded luciferase on the genome (expressed in the cells only upon infection) and carried the ectodomains of env genes of the following isolates: Ce1086 (tier 2, clade C), Ce1176 (tier 2, clade C), Ce2010 (tier 2, clade C), Du151.2 (tier 2, clade C), and TV1.21 (tier 2, clade C). For both assays, the neutralization titers are represented as the log10 serum dilution at which relative luminescence units (RLU) were reduced by 50% (50% infective dose [ID50]) compared to values for virus control wells after subtraction of background RLU in cell control wells. Both assays were done under GCLP-compliant conditions.
The Wilcoxon rank sum test was used to compare differences between groups. P values of <0.05 were considered significant. All values used for analyzing the proportionate representation of responses were background subtracted. Box plots were used to summarize the distribution of various immune responses, where the midline of the box indicates the median, the ends of the box denote the 25th and 75th percentiles, and whiskers extended to the extreme data points that are no more than 1.5 times the interquartile range (IQR) or, if no values meet this criterion, to the data extremes. Spearman's rank-based correlations were used to compare the association of CD4+ and CD8+ T cell responses with antibody responses from BAMAs, GTL ADCC assays, and NAb assays at week 26 between the NYVAC-C-KC and NYVAC-C-KC-ΔB19R vectors. Specifically, the DMSO-adjusted percent positivity for each cytokine and pooled antigen combination from the CD4+ and CD8+ T cell responses was assessed for correlation with (i) the AUC against the gp120.TV1 antigen for the IgG subclass in the BAMA assay, (ii) the titer against the gp120.TV1 antigen in the ADCC GTL assay, and (iii) the titer against the MW965.26 isolate in the TZM-bl NAb assay. The significance of the correlations was derived by using a one-sided P value with the alternative that a true correlation is greater than 0.
This investigation was supported by the PTVDC/CAVD Program with support from the Bill and Melinda Gates Foundation (BMGF). Humoral immune monitoring data were supported by BMGF CAVIMC grant 1032144 and the NIH/NIAID Duke Center for AIDS Research (CFAR; 5P30 AI064518). Novartis Vaccines received support for this work under contract number HHSN266200500007C from the DAIDS, NIAID, NIH.
We thank Marcella Sarzotti-Kelsoe for quality assurance oversight; William T. Williams, Robert Howington, R. Glenn Overman, Cristina Sánchez, and Victoria Jiménez for technical assistance; Sheetal Sawant for BAMA data management; and Hua-Xin Liao and Bart Haynes for envelope and V1/V2 protein reagents.