Department of Surgery, Duke University, Durham, North Carolina; and 7 Southern Research Institute, Frederick, Maryland

Transcription

1 MAJOR ARTICLE Highly Attenuated Rabies Virus Based Vaccine Vectors Expressing Simian-Human Immunodeficiency Virus 89.6P Env and Simian Immunodeficiency Virus mac239 Gag Are Safe in Rhesus Macaques and Protect from an AIDS-Like Disease Philip M. McKenna, 1 Martin L. Koser, 1 Kevin R. Carlson, 5 David C. Montefiori, 6 Norman L. Letvin, 5 Amy B. Papaneri, 1 Roger J. Pomerantz, 3 Bernhard Dietzschold, 1,4 Peter Silvera, 7 James P. McGettigan, 1 and Matthias J. Schnell 1,2,4 Departments of 1 Microbiology and Immunology, 2 Biochemistry and Molecular Biology, and 3 Medicine and 4 Center for Human Virology, Thomas Jefferson University, Philadelphia, Pennsylvania; 5 Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts; 6 Department of Surgery, Duke University, Durham, North Carolina; and 7 Southern Research Institute, Frederick, Maryland We analyzed the safety and immunogenicity of attenuated rabies virus vectors expressing simian-human immunodeficiency virus (SHIV) P Env or simian immunodeficiency virus (SIV) mac239 Gag in rhesus macaques. Four test macaques were immunized with both vaccine constructs, and 2 control macaques received an empty rabies vector. Seroconversion against rabies virus glycoprotein (G) and SHIV 89.6P Env was detected after the initial immunization, but no cellular responses against SHIV antigens were observed. HIV/SIVspecific immune responses were not enhanced by boosts with the same vectors. Therefore, we constructed vectors expressing SHIV 89.6P Env and SIV mac239 Gag in which the rabies G was replaced with the G protein of vesicular stomatitis virus (VSV). Two years after initial immunization, a boost with the rabies VSV G vectors resulted in SIV/HIV-specific immune responses. Upon challenge with SHIV 89.6P test macaques controlled the infection, whereas control macaques had high levels of viremia and a profound loss of CD4 + T cells, with 1 control macaque dying of an AIDS-like disease. The search for an effective HIV-1 vaccine has been frustratingly long, with a limited number of promising approaches in the pipeline. The use of live attenuated Received 9 September 2006; accepted 10 November 2006; electronically published 20 February Potential conflicts of interest: none reported. Financial support: National Institutes of Health (NIH; grants AI49153 and AI44340 to M.J.S., training grant 5T32AI07523 from the NIH Training Program in AIDS Research to P.M.M. and Thomas Jefferson University). The contract resources that were used for the study included Simian Vaccine Evaluation Unit (N01 AI15429, awarded to the Southern Research Institute), cellular immunology laboratory (N01 AI30033, awarded to Beth Israel Deaconess), and the neutralization laboratory (N01 AI30034, awarded to Duke University) Reprints or correspondence: Dr. Matthias J. Schnell, 233 S. 10th St., Ste. 531 BLSB, Philadelphia, PA (matthias.schnell@jefferson.edu). The Journal of Infectious Diseases 2007; 195: by the Infectious Diseases Society of America. All rights reserved /2007/ $15.00 DOI: / simian immunodeficiency virus (SIV), which provided complete protection from challenge with homologous wild-type strains, indicates the potential feasibility of live viral vectors as HIV-1 vaccines (for a review, see [1]). As replicating immunogens, such vectors are capable of eliciting immune responses and of providing protection from infection, which is not seen with other approaches based on DNA or nonreplicating vectors. For HIV-1 and SIV, there is evidence that virus-specific CD8 + cytotoxic T lymphocytes (CTLs) play an important role in controlling viral replication and disease progression [2]. In nonhuman primates, studies based on DNA vaccines and recombinant viral vectors have shown the capacity to elicit virus-specific CTLs. Upon pathogenic simian-human immunodeficiency virus (SHIV) or SIV challenge, vaccinated macaques ex- 980 JID 2007:195 (1 April) McKenna et al.

2 perienced reduced peak viral loads and a delayed course of disease progression [3 5]. However, other studies of CTL-based vaccines reported only transient control of viral replication after pathogenic SIV mac239 challenge [6] and viral escape of immune control through mutations in CTL epitopes (reviewed in [7]). The role of antibodies in conferring protection to HIV-1 infection has been examined in adoptive transfer experiments in which the passive immunization of HIV-1 neutralizing antibodies was able to prevent either infection or disease in both chimpanzee and rhesus macaque challenge models [8]. Also in macaques, topical application of a broadly neutralizing antibody was shown to block the vaginal transmission of a SHIV virus [9]. Thus, it seems likely that a successful HIV-1 vaccine should be able to activate both arms of the immune system. Rabies virus is a member of the family Rhabdoviridae of nonsegmented, single-stranded, negative-sense RNA viruses. Expression of HIV-1 and other viral antigens from attenuated, recombinant rabies virus has been shown to be safe and to induce both humoral and cellular immune response in smallanimal models [10 13]. Although immune responses generated in rodent models are not always predictive of a potential HIV- 1 vaccine, the basic parameters of the anti HIV-1 response can be studied in such systems. We performed a proof-of-principle experiment in nonhuman primates with rabies virus vectors expressing SHIV 89.6P Env and SIV mac239 Gag. The vectors are based on a replication-competent rabies virus vaccine strain, SAD B19, that has been used for the oral immunization of wild animals in Europe for 120 years [14]. However, because of residual vector pathogenicity after nonperipheral inoculation [15], we introduced a previously described mutation in rabies virus glycoprotein G (rabies virus G) to abolish its neurotropic character [16, 17]. In mice, rabies viruses carrying this mutation are apathogenic even after direct intracranial inoculation [17]. MATERIALS AND METHODS Recombinant Vaccine Vectors The recombinant rabies virus cdna pspbn-333 has been described elsewhere [17]. pspbn-ig was constructed by polymerase chain reaction amplification of the chimeric vesicular stomatitis virus (VSV)/rabies virus G containing the ectodomain and transmembrane domain from VSV strain Indiana, and the cytoplasmic domain (CD) of rabies virus G from psn VSV-G [18], which was used to replace the rabies virus G in pspbn-333, resulting in pspbn-ig. SIV mac239 Gag or SHIV 89.6P Env containing the rabies virus CD were amplified from p239spsp5 [19] or pdg-89.6p- RVG [20], respectively, and cloned into SPBN-333 or SPBN-IG. The plasmids were designated pspbn P, pspbn-333 SIV Gag, pspbn-ig 89.6P, and pspbn-ig SIV Gag. Sequences were confirmed by DNA sequencing, and infectious viruses SPBN-333, SPBN-IG, SPBN P, SPBN-333 SIV Gag, SPBN-IG 89.6P, and SPBN-IG SIV Gag were recovered using standard methods [12]. Macaque Studies Six adult male rhesus macaques (Macaca mulatta) were used in the study. They were screened and confirmed to be free of antibodies to SIV, simian retrovirus, simian T cell leukemia virus type 1, isolatable SIV before their inclusion in the study. The macaques were housed at the Southern Research Institute (Frederick, MD) in accordance with American Association for Accreditation of Laboratory Animal Care standards. The experimental design met full approval of the institute s Animal Care and Use Committee. Vaccinations On study day 0, 4 test macaques ( ) were immunized 7 by intramuscular (im) inoculation with foci-forming units (ffu) of rabies virus vector SPBN-333 SIV Gag and ffu of SPBN P Env. The 2 control monkeys ( ) were immunized im with ffu of SPBN-333. At study week 45, all macaques received intranasal (inl) boosts with the same vectors and doses as those on day 0. At week 55, all macaques received an im inoculation as on day 0. At week 120, 4 test macaques received an im inoculation with ffu of SPBN-IG SIV Gag and 2 10 ffu SPBN-IG 89.6P Env. Control monkeys received ffu im of SPBN- IG with no transgene. Immunostaining Expression of rabies virus G or VSV-G, 89.6P Env, and SIV Gag was analyzed in BSR cells infected at an MOI of 0.01 for 48 h. Cells were fixed with 3% paraformaldehyde and washed in 10 mmol/l glycine in PBS. For SIV Gag staining, cells were incubated in 1% Triton X-100 for 10 min after fixation. Cells were immunostained with mouse anti rabies virus G polyclonal antibody, anti VSV G mouse monoclonal antibodies I1 and I14, a mouse monoclonal antibody directed against SIV Gag 2F12 (AIDS Research and Reference Reagent Program [ARRRP], Division of AIDS, National Institute of Allergy and Infectious Disease, National Institutes of Health 1610), or the human anti HIV-1 gp41 monoclonal antibody 2F5 (ARRRP 1475), followed by the respective fluorescein isothiocyanate labeled donkey anti mouse, anti rabbit or anti human IgG (Jackson ImmunoResearch). SHIV 89.6P Challenge Four months after the boost with vectors containing VSV-G, all macaques were challenged intravenously (iv) with 50 monkey infectious doses (MID) of pathogenic SHIV 89.6P. The SHIV 89.6P challenge stock has been described elsewhere [21 23] and Rabies Virus Based HIV-1 Vaccine in Macaques JID 2007:195 (1 April) 981

3 Figure 1. Schedule and viral vectors. A, Time line of vaccinations. The priming dose was administered intramuscularly (im, week 0), the boost at week 45 intranasally (inl), the boost at week 55 im, and the boost at week 120 im. All macaques were challenged intravenously at week 136 with 50 monkey infectious dose of simian-human immunodeficiency virus (SHIV) 89.6P. B, Schematic of the recombinant rabies viruses (RVs) encoding RV-G or vesicular stomatitis virus (VSV) G (SPBN-333 and SPBN-IG, respectively) or rabies virus vectors encoding RV-G or VSV-G in addition to SHIV 89.6P Env or SIV mac239 (SPBN P, SPBN-IG 89.6P, SPBN-333 SIV Gag, or SPBN-IG SIVGag). C, Immunostaining of BSR cells (a BHK clone) infected with SPBN P (column A), SPBN-IG 89.6P (column B), SPBN-333 SIV Gag (column C), or SPBN-IG SIV Gag (column D), with antibodies directed against VSV-G (row 1, anti VSV-G), RV-G (row 2, anti RV-G), SIV mac239 Gag (row 3, anti SIV-Gag), or HIV Env (row 4, anti-hiv-env). L, polymerase; M, matrix protein; N, nucleoprotein; P, phosphoprotein. was provided by Dr. N. Letvin (Harvard University, Cambridge, MA). Immune Assays Oligomeric gp140 ELISA. The oligomeric gp140 ELISA was performed essentially as described elsewhere [24], except that plates were coated with recombinant vaccinia virus derived oligomeric gp140 of strain 89.6P (J.P.M., unpublished data). Background was defined as the average signal at each dilution for control macaques and a positive signal as at least twice background levels. Virus neutralization assays. The rabies virus neutralizing antibody titers were determined using the CVS-11 reference strain, as described elsewhere [25]. The virus neutralizing antibody titers were transformed into international units using the World Health Organization s anti rabies virus antibody standard. To determine levels of VSV neutralizing antibodies, we modified the rabies virus neutralization assay by replacing the CVS-11 strain with the SPBN-IG virus. Neutralization titers were determined as the serum dilution that achieved 50% reduction of foci-forming units of input virus. HIV-1/SIV/SHIV neutralization assay were performed as described elsewhere [26]. Neutralization was measured as a function of reductions in luciferase reporter gene expression after a single round of infection in TZM-bl cells (ARRRP) [26]. Assay stocks of SHIV-89.6 and SHIV-89.6P were prepared in human peripheral blood mononuclear cells (PBMCs). Cellular responses. Enzyme-linked immunospot (ELISpot) assays for specific cellular immune responses to SIV mac239 Gag and SHIV 89.6P Env were performed as described elsewhere [27]. Background levels were determined by counts in medium-only control, and the criteria for a positive signal were levels at least 2 times background and at least 55 sfu/10 6 PBMCs. Viral load determinations. SIV viral RNA was quantitated as described elsewhere [28, 29], with the following modifications: Gag primers were SIV forward 5 -AGTATGGGCAGCAA- ATGAAT-3 and SIV reverse 5 -TTCTCTTCTGCGTGAATGC- 3, and the probe was SIV 6FAM-AGAT-TTGGATTAGCAGAA- AGCCTGTTGGA-TAMRA. The assay has a threshold sensitivity of 200 RNA copies/ml of plasma, with interassay variations averaging 0.5 log 10. RESULTS Vaccination schedule. Six Mamu-A*01 major histocompatibility complex (MHC) class I allele negative rhesus macaques (Macaca mulatta) were inoculated with the constructs shown in figure 1B. Four macaques (3413, 3414, 3415, and 3416) were 982 JID 2007:195 (1 April) McKenna et al.

4 Figure 2. Humoral immune response to vaccine antigens. A, Serum from designated time points tested in an oligomeric gp140 ELISA for detection of 89.6P Env. Black bars, test macaques; white bars, control macaques. The initial bar for each macaque represents a serum dilution of 1:960, the second bar 1:2880, and the third bar 1:8640. Error bars represent values of duplicate samples. B, Rabies virus neutralization titers. Bars indicate the level in international units of rabies virus neutralization achieved for each macaque at given time points. C, Rabies virus vesicular stomatitis virus (VSV) G neutralization titers. Serum from indicated time points were tested for the presence of VSV neutralizing antibodies. The neutralization assay was a modified rabies virus neutralization assay using SPBN-IG instead of rabies virus. Bars represent the serum dilution that achieved a 50% reduction of foci-forming units of input virus.

5 Figure 3. Interferon (IFN) g levels in peripheral blood mononuclear cells (PBMCs) as measured by enzyme-linked immunospot assay after boost with vesicular stomatitis virus G protein encoding vectors. Freshly isolated PBMCs were analyzed for IFN-g secretion after restimulation with peptide pools of 15-mer peptides overlapping by 11 aa spanning the entire simian immunodeficiency virus (SIV) mac239 (SIV Gag; white bars) or the simian-human immunodeficiency virus (SHIV) 89.6P N-terminal half (89.6P Env 1, gray bars) or C-terminal half (89.6P Env 1, black bars). Reported values are levels seen after subtraction of assay control (restimulation with media only), background levels were sfc/10 6 PBMCs (week 0), sfc/10 6 PBMCs (week 2), sfc/10 6 PBMCs (week 4), sfc/10 6 PBMCs (week 6), sfc/10 6 PBMCs (week 8), and sfc/10 6 PBMCs (week 12). Error bars represent mean values of triplicate samples. The time indicated is in weeks after week the 120 boost with SPBN-IG vectors. vaccinated with vectors expressing 89.6P Env and SIV 239 Gag according to the schedule described in figure 1A and Materials and Methods. Two macaques (3417 and 3418) were vaccinated with vectors not expressing transgenes (SPBN-333 and SPBN- IG; figure 1B). Expression of 89.6P Env, SIV Gag, rabies virus G, and VSV-G proteins from respective viral vectors was confirmed by indirect immune fluorescence (figure 1C). Humoral immune response to vaccine antigens. After each vaccination with vectors containing rabies virus G or VSV-G, we analyzed the humoral immune response to vaccine antigens. Four and 8 weeks after the initial inoculation, 4 weeks after the inl boost at week 45, and 4 weeks after the im boost at week 55, serum from all macaques was analyzed in an oligomeric gp140 ELISA assay [20] for seroconversion to 89.6P Env. As seen in figure 2A, 2 of 4 test macaques (3413 and 3416) seroconverted against 89.6P Env by 4 weeks after the initial inoculation. By week 8, macaque 3414 was also positive, and the response in macaque 3415 was just below the positive response cutoff of twice the average background signal seen in the controls. The week 45 inl boost appeared to have no effect, as is seen in the week 49 results. However, the im boost at week 55 did elicit a response, in that all 4 test macaques were positive at week 59, with the highest increase in signal seen in macaques 3415 and This modest boost is likely attributable to the presence of 89.6P Env on the rabies virus particle. Inclusion of the rabies virus G cytoplasmic domain on the 89.6P Env allows efficient incorporation of the chimeric protein into the rabies particle [30]. Serum from the designated time points was analyzed for HIV-1 neutralizing antibodies; however, no neutralizing antibodies were detected (data not shown). We also assessed the anti rabies virus G response in all macaques. Figure 2B shows that the initial inoculation elicited high-titer rabies virus neutralizing antibodies, in the range of IU, which persisted throughout the course of the study. For reference, 0.5 IU is considered to be protective against classic rabies infection in both humans and animals. The inl boost at week 45 had only a small effect on the anti rabies virus response, whereas the im boost at week 55 raised the already high titers of anti rabies virus antibodies (figure 2B). The inl boost with rabies was based on the finding for vaccinia virus that overcoming preexisting immunity can be achieved by mucosal immunization [31]. However, this seems not to be the case for rabies virus, and effective boosting of the anti rabies virus G and anti-env response was not observed. The im boost was administered to determine whether the Env and Gag antigens contained within the rabies vector could enhance the SHIV-specific response. The increases in anti-env and anti-g antibodies seen after the im inoculation were most likely due to Env and rabies virus G proteins contained in virions. The lack of detectable cellular responses indicates that no viral replication occurred. The above findings led to the development of the SPBN-IG (VSV-G expressing) vectors (figure 1B). With these new vectors, we sought to augment the anti-env and anti-gag response by infecting with a vector expressing a different membrane glycoprotein. Previous work showed that rhabdoviral glycoprotein 984 JID 2007:195 (1 April) McKenna et al.

6 Figure 4. Immunological and clinical results after intravenous challenge with simian-human immunodeficiency virus (SHIV) 89.6P. A, SHIV 89.6P loads after challenge. Assay cutoff was 200 copies RNA/mL of plasma. B, CD4 + T cell counts. HIV-1 neutralizing antibodies directed against SHIV 89.6 (C) and SHIV 89.6P (D) were determined over time as indicated. E, Postchallenge interferon (IFN) g secreting cells measured by IFN-g enzyme-linked immunospot using pools of 15-mer peptides overlapping by 11 aa spanning the entire simian immunodeficiency virus (SIV) mac239 (SIV Gag; white bars) or the SHIV 89.6P N-terminal half (89.6P Env 1; gray bars) or C-terminal half (89.6P Env 1; black bars). Reported values are levels seen after subtraction of assay control (restimulation with medium only); background levels were sfc/10 6 PBMCs (week 0), sfc/10 6 PBMCs (week 2), sfc/10 6 PBMCs (week 4), sfc/10 6 PBMCs (week 8), sfc/10 6 PBMCs (week 12), and sfc/10 6 PBMCs (week 20). Error bars represent mean values of triplicate samples. The time indicated is in weeks after week 120 boost with the SPBN-IG vectors. The cross indicates the time point at which macaque 3418 was euthanized. G can efficiently be substituted by other viral and cellular glycoproteins [18, 32, 33]. Foley et al. [18] showed that a rabies virus containing the VSV G protein grew similarly to rabies but failed to protect mice from challenge with a pathogenic rabies virus, which indicates that a rabies virus vaccine vector with a VSV G protein should be able to infect in the presence of high-titer anti rabies virus antibodies. On week 120, all macaques were boosted im with SPBN-IG 89.6P Env and SPBN- IG SIV Gag (test macaques) or SPBN-IG (control macaques). As displayed in figure 2C, anti VSV-G antibodies were absent in all macaques at the time of boosting. However, after the boost, all macaques seroconverted against VSV-G and raised Rabies Virus Based HIV-1 Vaccine in Macaques JID 2007:195 (1 April) 985

7 high titers of VSV neutralizing antibodies (figure 2C). The strong immune response to the SPBN-IG vectors shows that we were able to successfully infect all macaques and thereby circumvent the rabies virus neutralizing antibodies. We tested post SPBN-IG boost serum for HIV-1 neutralizing antibodies, but only very low level or undetectable responses were seen (data not shown). Cellular immune response. We assessed the anti HIV/SIV cellular immune response by interferon (IFN) g ELISpot assay. PBMCs from all macaques were tested against overlapping peptide pools spanning the entirety of 89.6P Env and SIV Gag for the presence of IFN-g secreting cells. At no point before the SPBN-IG boost did we detect any Env- or Gag-specific cellular immune responses (data not shown). This lack of cellular responses is in line with our observation with rabies virus vectors in mice, where cellular responses were very low after vaccination but were robust after challenge with recombinant vaccinia virus expressing the same HIV-1 or SIV antigen [11 13, 34]. However, after the SPBN-IG boost, we were able to detect 89.6P Env and SIV Gag specific IFN-g secreting cells (figure 3). Responses in 3 of 4 test macaques peaked 4 weeks after boost, with macaque 3413 responding to epitopes in SIV Gag and both Env pools tested and macaques 3414 and 3415 primarily responding to epitopes in the N-terminal half of 89.6P Env (89.6P Env 1; figure 3). The results of the ELISPOT assay for macaque 3416 are not shown, because this macaque had high numbers of IFN-g secreting cells throughout the study ( cells/10 6 PBMCs), even after restimulation with medium only. Because the observed responses were at least as potent as those seen in other vaccine approaches using this rhesus macaque model system [35, 36], all macaques were challenged iv 4 months after the SPBN-IG boost with 50 MID of pathogenic SHIV 89.6P. Post SHIV 89.6P challenge analysis. After challenge, we continued to monitor the HIV/SIV-specific immune responses, and we also determined SHIV 89.6P loads, CD4 + T cell counts, and clinical events in all macaques. The immunogenicity results presented in figure 4 indicate that rabies virus based vectors were able to efficiently prime both arms of the immune system. In the case of SHIV antigen-specific T cell responses, all vaccinated macaques showed a high level of SIV Gag and 89.6P Env specific T cells in the IFN-g ELISpot assay, with the largest expansion of SHIV-specific T cells at week 4 after infection (figure 4E). One control macaque (3417) did not mount any Gag- or Env-specific cellular immune response, and the other macaque (3418) had only transient responses at 8 weeks after challenge. Postchallenge serum was assayed for virus neutralizing antibodies against SHIV IIIB,SHIV 89.6 (figure 4C), and SHIV 89.6P (figure 4D). All 4 vaccinated macaques developed virus neutralizing antibodies against SHIV 89.6P, with high titers of antibodies seen in macaques 3414 and More modest responses were generated against strain By week 4 after challenge, macaques 3415 and 3416 had anti-89.6 titers 11:100, and, by week 20, macaque 3414 had an anti-89.6 titer of 1: 287. In the control macaques, we detected levels of virus neutralizing antibodies against SHIV 89.6 or 89.6P only close to baseline levels (figure 4C and 4D). Not surprisingly, no neutralizing antibodies were detected against SHIV-IIIB in any of the 6 macaques [37]. Protection by these observed cellular and humoral immune responses against an AIDS-like disease in the vaccinated macaques was indicated by all 4 macaques controlling the infection well during the 20 weeks of follow-up. All macaques had similar peak viral loads of 10 7 copies/ml 2 weeks after challenge. However, 3 of 4 test macaques had undetectable viral loads by week 12 after challenge, with the fourth macaque (3414) having only 270 copies/ml. By week 20 after challenge, all vaccinees had undetectable viral loads (figure 4A). By contrast, both control macaques maintained viral loads copies/ml throughout follow-up (figure 4A). In addition, the 2 control macaques showed a complete loss of their peripheral CD4 + T cells by week 3 after challenge, whereas 3 of the vaccinated macaques maintained their CD4 + T cell counts well. Vaccinated macaque 3414 had a significant decrease in CD4 + T cell counts but no clinical signs. Of note, this macaque had the most robust immune response to the challenge (figure 4C 4E) and was able to control the infection 4 weeks later than the other vaccinees (figure 4A), despite the low CD4 + T cell counts detected throughout follow-up. One control macaque (3418) had a poor clinical course, progressed to an AIDS-like disease, and was euthanized at week 18 after challenge. The experiment was terminated at week 22 after challenge. DISCUSSION We report here on a long-term vaccination study in which 4 test macaques were boosted with rabies virus VSV-G vectors expressing either SIV mac 239 Gag or SHIV 89.6P Env 2 years after initial priming with rabies virus vectors expressing the same 2 vaccine antigens. Two control macaques were similarly vaccinated with rabies virus and rabies virus VSV-G vectors expressing no transgene. Four months after the rabies virus VSV-G boost, macaques were challenged iv with pathogenic SHIV 89.6P. Vaccinated macaques controlled the infection, as seen by reduced viral loads, compared with those in control macaques, by mounting cellular immune responses to both SIV Gag and SHIV 89.6P Env and generating SHIV 89.6P neutralizing antibodies. Three vaccinees showed good, and 1 sufficient, retention of CD4 + T lymphocytes and showed no evidence of clinical disease from the rabies virus vaccination or SHIV challenge. By contrast, the 2 control macaques had high acute postchallenge viral loads and a near complete loss of CD4 + T cells. All 4 test macaques displayed a combination of cellular and 986 JID 2007:195 (1 April) McKenna et al.

8 humoral responses that resulted in undetectable levels of plasma viremia by 12 weeks after challenge in 3 of 4 macaques and only 270 copies/ml in the fourth (figure 4C). Of particular interest were the cellular responses to Gag and multiple epitopes in Env. Previous work in rhesus macaques has demonstrated the need for broad cellular responses to multiple epitopes to limit the selection of CTL escape mutants directed against single immunodominant epitopes [38]. The elicitation of cellular responses to multiple epitopes seen in the present study suggests that the cellular response to the challenge may be more robust and durable than those seen in Gag-only approaches. Although statistical resolution cannot be obtained by this small sample size, the proof of concept is shown in the protection from disease in the test macaques, and the results indicate the need for future experiments with larger groups of animals. This initial report demonstrates the potential of rabies virus vectors; however, it also raises several questions that need to be addressed as the technology moves forward. The fundamental question with this approach is whether a replicating rabies virus based vector as an HIV-1 vaccine in humans is feasible. A number of efforts have focused on developing vaccines against HIV-1 with replication-deficient viral vectors, with those based on adenovirus being the most notable. However, it is unclear whether these vectors will perform in humans as well as in monkeys because of preexisting vector immunity (for a review, see [39]). Other vectors are in early stages of development, and their outcome is uncertain (for a review, see [40]). For example, vectors based on alphaviruses and picornaviruses may be limited by their available coding capacity for foreign proteins. Rabies virus, however, is able to express large and multiple vaccine antigens [11]. The current HIV-1 pandemic does not allow any further delays in the development of novel vaccine approaches. The vector most closely related to rabies virus is another Rhabdovirus, VSV, which has also been tested in the rhesus macaque system [35, 36]. One study indicated that a good level of protection requires 3 inoculations with VSV containing 3 different VSV G proteins or 1 inoculation with the VSV vector followed by a boost with recombinant modified vaccinia virus Ankara [35, 36]. The use of rabies virus, compared with VSV, may seem more risky, but there are several reasons not to discount this strategy: (1) the pathogenesis of rabies virus is better understood and more clearly defined than that of most other viral vectors; (2) recent research allows us to use vectors that are not neurotropic; (3) new mutations can be introduced into the rabies virus genome that greatly restrict the infectivity of rabies virus on neuronal cells but not on nonneural cells (G. S. Tan and M.J.S., unpublished data); and (4) attenuated rabies virus vectors administered im are not less immunogenic, as has been seen for VSV-based vectors [17, 36, 41, 42]. The ability to create rhabdoviral glycoprotein exchange vectors, such as the SPBN-IG constructs described here, provides a flexible and potentially powerful new platform for prime-boost vaccination strategies. Additional questions with this and other emerging vaccine platforms involve study design. Can a prime-boost strategy using 2 live-recombinant viral vectors be optimized and applied to other, perhaps more relevant infection models? The low or undetectable cellular immune responses that we observed are typical for most experimental HIV-1 vaccines after initial immunization, except for macaques positive for the Mamu-A*01 MHC class I allele [4, 6, 32]. Recent work has indicated that Mamu-A*01 monkeys are more readily protected by a SIV Gag based vaccine than macaques negative for this MHC class I allele; therefore, the use of such Mamu-A*01 monkeys for vaccine testing is now questionable [6]. None of our monkeys were Mamu-A*01 positive, yet we achieved protection comparable to that in previous reports. It remains to be seen whether the protection from disease seen in this SHIV 89.6P challenge model can be extended to other infection models in rhesus macaques. We aim to address many of these issues in future primate experiments using the SPBN-333 and SPBN-IG vectors described here. References 1. Koff WC, Johnson PR, Watkins DI, et al. HIV vaccine design: insights from live attenuated SIV vaccines. Nat Immunol 2006; 7: Migueles SA, Tilton JC, Connors M. Advances in understanding immunologic control of HIV infection. Curr HIV/AIDS Rep 2004; 1: Barouch DH, Santra S, Schmitz JE, et al. Control of viremia and prevention of clinical AIDS in rhesus monkeys by cytokine-augmented DNA vaccination. Science 2000; 290: Shiver JW, Fu TM, Chen L, et al. Replication-incompetent adenoviral vaccine vector elicits effective anti-immunodeficiency-virus immunity. Nature 2002; 415: Amara RR, Villinger F, Altman JD, et al. Control of a mucosal challenge and prevention of AIDS by a multiprotein DNA/MVA vaccine. Science 2001; 292: Casimiro DR, Wang F, Schleif WA, et al. Attenuation of simian immunodeficiency virus SIVmac239 infection by prophylactic immunization with DNA and recombinant adenoviral vaccine vectors expressing Gag. J Virol 2005; 79: Peyerl FW, Barouch DH, Letvin NL. Structural constraints on viral escape from HIV- and SIV-specific cytotoxic T-lymphocytes. Viral Immunol 2004; 17: Moore JP, Parren PW, Burton DR. Genetic subtypes, humoral immunity, and human immunodeficiency virus type 1 vaccine development. J Virol 2001; 75: Shattock RJ, Moore JP. Inhibiting sexual transmission of HIV-1 infection. Nat Rev Microbiol 2003; 1: McGettigan JP, Foley HD, Belyakov IM, Berzofsky JA, Pomerantz RJ, Schnell MJ. Rabies virus-based vectors expressing human immunodeficiency virus type 1 (HIV-1) envelope protein induce a strong, crossreactive cytotoxic T-lymphocyte response against envelope proteins from different HIV-1 isolates. J Virol 2001; 75: McGettigan JP, Naper K, Orenstein J, Koser M, McKenna PM, Schnell MJ. Functional human immunodeficiency virus type 1 (HIV-1) Gag- Pol or HIV-1 Gag-Pol and env expressed from a single rhabdovirusbased vaccine vector genome. J Virol 2003; 77: McGettigan JP, Pomerantz RJ, Siler CA, et al. Second-generation rabies virus-based vaccine vectors expressing human immunodeficiency virus Rabies Virus Based HIV-1 Vaccine in Macaques JID 2007:195 (1 April) 987

9 type 1 gag have greatly reduced pathogenicity but are highly immunogenic. J Virol 2003; 77: McGettigan JP, Sarma S, Orenstein JM, Pomerantz RJ, Schnell MJ. Expression and immunogenicity of human immunodeficiency virus type 1 Gag expressed by a replication-competent rhabdovirus-based vaccine vector. J Virol 2001; 75: Conzelmann KK, Cox JH, Schneider LG, Thiel HJ. Molecular cloning and complete nucleotide sequence of the attenuated rabies virus SAD B19. Virology 1990; 175: Morimoto K, Hooper DC, Spitsin S, Koprowski H, Dietzschold B. Pathogenicity of different rabies virus variants inversely correlates with apoptosis and rabies virus glycoprotein expression in infected primary neuron cultures. J Virol 1999; 73: Dietzschold B, Wunner WH, Wiktor TJ, et al. Characterization of an antigenic determinant of the glycoprotein that correlates with pathogenicity of rabies virus. Proc Natl Acad Sci USA 1983; 80: McGettigan JP, Pomerantz RJ, Siler C, et al. Second generation rabiesbased vaccine vectors expressing HIV-1 Gag have greatly reduced pathogenicity but are highly immunogenic. J Virol 2003; 77: Foley HD, McGettigan JP, Siler CA, Dietzschold B, Schnell MJ. A recombinant rabies virus expressing vesicular stomatitis virus glycoprotein fails to protect against rabies virus infection. Proc Natl Acad Sci USA 2000; 97: Kestler H, Kodama T, Ringler D, et al. Induction of AIDS in rhesus monkeys by molecularly cloned simian immunodeficiency virus. Science 1990; 248: McKenna PM, Aye PP, Dietzschold B, et al. Immunogenicity study of glycoprotein-deficient rabies virus expressing simian/human immunodeficiency virus SHIV89.6P envelope in a rhesus macaque. J Virol 2004; 78: Barouch DH, Fu TM, Montefiori DC, Lewis MG, Shiver JW, Letvin NL. Vaccine-elicited immune responses prevent clinical AIDS in SHIV(89.6P)-infected rhesus monkeys. Immunol Lett 2001; 79: Reimann KA, Li JT, Voss G, et al. An env gene derived from a primary human immunodeficiency virus type 1 isolate confers high in vivo replicative capacity to a chimeric simian/human immunodeficiency virus in rhesus monkeys. J Virol 1996; 70: Reimann KA, Li JT, Veazey R, 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 1996; 70: Richardson TM Jr, Stryjewski BL, Broder CC, et al. Humoral response to oligomeric human immunodeficiency virus type 1 envelope protein. JVirol1996; 70: Dietzschold B, Schnell MJ. New approaches to the development of live attenuated rabies vaccines. Hybrid Hybridomics 2002; 21: Li M, Gao F, Mascola JR, et al. Human immunodeficiency virus type 1 env clones from acute and early subtype B infections for standardized assessments of vaccine-elicited neutralizing antibodies. J Virol 2005; 79: Santra S, Barouch DH, Korioth-Schmitz B, et al. Recombinant poxvirus boosting of DNA-primed rhesus monkeys augments peak but not memory T lymphocyte responses. Proc Natl Acad Sci USA 2004; 101: Rossio JL, Esser MT, Suryanarayana K, et al. Inactivation of human immunodeficiency virus type 1 infectivity with preservation of conformational and functional integrity of virion surface proteins. J Virol 1998; 72: Suryanarayana K, Wiltrout TA, Vasquez GM, Hirsch VM, Lifson JD. Plasma SIV RNA viral load determination by real-time quantification of product generation in reverse transcriptase-polymerase chain reaction. AIDS Res Hum Retroviruses 1998; 14: Foley HD, Otero M, Orenstein JM, Pomerantz RJ, Schnell MJ. Rhabdovirus-based vectors with human immunodeficiency virus type 1 (HIV-1) envelopes display HIV-1-like tropism and target human dendritic cells. J Virol 2002; 76: Belyakov IM, Moss B, Strober W, Berzofsky JA. Mucosal vaccination overcomes the barrier to recombinant vaccinia immunization caused by preexisting poxvirus immunity. Proc Natl Acad Sci USA 1999; 96: Schnell MJ. Viral vectors as potential HIV-1 vaccines. FEMS Microbiol Lett 2001; 200: Schnell MJ, Johnson JE, Buonocore L, Rose JK. Construction of a novel virus that targets HIV-1-infected cells and controls HIV-1 infection. Cell 1997; 90: Faber M, Pulmanausahakul R, Hodawadekar SS, et al. Overexpression of the rabies virus glycoprotein results in enhancement of apoptosis and antiviral immune response. J Virol 2002; 76: Ramsburg E, Rose NF, Marx PA, et al. Highly effective control of an AIDS virus challenge in macaques by using vesicular stomatitis virus and modified vaccinia virus Ankara vaccine vectors in a single-boost protocol. J Virol 2004; 78: Rose NF, Marx PA, Luckay A, et al. An effective AIDS vaccine based on live attenuated vesicular stomatitis virus recombinants. Cell 2001; 106: Crawford JM, Earl PL, Moss B, et al. Characterization of primary isolate-like variants of simian-human immunodeficiency virus. J Virol 1999; 73: Barouch DH, Kunstman J, Kuroda MJ, et al. Eventual AIDS vaccine failure in a rhesus monkey by viral escape from cytotoxic T lymphocytes. Nature 2002; 415: Vanniasinkam T, Ertl HC. Adenoviral gene delivery for HIV-1 vaccination. Curr Gene Ther 2005; 5: Girard MP, Osmanov SK, Kieny MP. A review of vaccine research and development: the human immunodeficiency virus (HIV). Vaccine 2006; 24: Roberts A, Kretzschmar E, Perkins AS, et al. Vaccination with a recombinant vesicular stomatitis virus expressing an influenza virus hemagglutinin provides complete protection from influenza virus challenge. JVirol1998; 72: Egan MA, Chong SY, Rose NF, et al. Immunogenicity of attenuated vesicular stomatitis virus vectors expressing HIV type 1 Env and SIV Gag proteins: comparison of intranasal and intramuscular vaccination routes. AIDS Res Hum Retroviruses 2004; 20: JID 2007:195 (1 April) McKenna et al.