INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Human immunodeficiency virus type 1 (HIV-1) is primarily a sexually transmitted virus which causes persistent systemic infection and loss of CD4⁺ T cells which culminates in loss of immunocompetence and development of AIDS. Although several effective anti-HIV drugs, notably viral protease inhibitors, are now available, their widespread use is hampered by their high cost and the need for multiple treatments daily for an indefinite period. Their use is thus not feasible in less developed countries, where the bulk of HIV infections occur. Under these conditions, development of a safe and effective vaccine is a priority. Moreover, such a vaccine should protect against sexual transmission of the virus. The inability of HIV-1 to infect animals other than chimpanzees has meant that macaque models using nonhuman primate lentiviruses have become the only practical alternative for testing proof-of-concept approaches to vaccines against HIV-1. The simian immunodeficiency virus SIV_mac, which is closely related to HIV-2 and SIV_sm, causes AIDS in macaques and has been extensively used in vaccine studies. Macaques vaccinated with attenuated strains of SIV_mac239, produced by deleting auxiliary genes, including nef, resisted infection after challenge with virulent strains of SIV_mac (1, 5, 21, 23, 24, 30). A major limitation of the SIV_mac239 model of AIDS is that this virus is only distantly related to HIV-1 genetically, and although SIV_mac causes AIDS in macaques, the biological properties of SIV_mac vary greatly from those of HIV-1, especially with reference to neutralization and replication in macrophages (31). Since some of these differences are attributable to the envelopes of the viruses, chimeric simian-human immunodeficiency virus (SHIV), which contains the core of SIV_mac and the envelope of HIV-1, was created in hopes of simulating the effects of HIV-1 in macaques more accurately than does SIV_mac. Newly constructed SHIVs were infectious and induced neutralizing antibodies to HIV-1 but were not pathogenic, and thus vaccines could not be evaluated for efficacy in preventing disease. Our recent derivation of virulent SHIV which causes almost total loss of CD4⁺ T cells and AIDS in macaques 6 months after inoculation established a new and reproducible model which accurately reflected HIV-1 disease, although compressed into a short time frame (9, 11, 12). Moreover, the new virus is virulent after intravenous, oral, or intravaginal infusion (7, 9, 13). This model thus proved ideal for evaluating efficacy of vaccines against sexually transmitted HIV-1.

Our choice of live-virus vaccines was based on observations of genetic changes occurring in SHIV during the animal passages that yielded virulent SHIV_KU-1. The original molecularly cloned SHIV-4 is avirulent and has a stop codon in vpu (16). SHIV_KU-1 is extremely virulent and has an open vpu in addition to numerous mutations in the env and nef genes (25, 27). However, most of the changes in the latter two genes developed after vpu became functional (17). The genetic changes correlated with a newly acquired ability of the virus to replicate efficiently in macrophage cultures and at extremely high titers in the animals, with resultant subtotal loss of CD4⁺ T cells and development of AIDS. Assuming that the changes in the auxiliary genes vpu and nef were essential for development for virulence of the virus, we deleted portions of both genes from original SHIV-4 and used this virus as a vaccine (group 1). Oral inoculation of this virus into two newborn pig-tailed macaques failed to result in infection in the animals (19a). Therefore, in the vaccine trial described in this report, the animals were inoculated subcutaneously with this virus. Chimeric virus SHIV_PPc was created by replacing the env and nef genes of SHIV-4 with the corresponding regions of virus isolated from macaque PPc, which died with AIDS (26). SHIV_PPc is macrophagetropic (27). To test the attenuating potential of a vpu deletion alone, we created virus vpuSHIV_PPc and used this virus as a vaccine. This virus caused a transiently productive infection in two newborn macaques after oral inoculation. Therefore, in the vaccine experiment (group 2), this virus was inoculated orally. Four unvaccinated controls and the 12 vaccinated animals were challenged 6 to 7 months later by intravaginal infusion of 1 ml of undiluted SHIV_KU-1 stock.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Viruses. We obtained a SHIV-4 DNA encoding the env, tat, rev, and vpu genes of HIV-1 HXB2c on a background of SIV_mac239 (16) from Joseph Sodroski, Harvard University. Viral DNA was transfected into CEMx174 cells to produce a virus that was used to initiate passage in macaques. Virus from the fourth passage of SHIV-4, which was associated with AIDS and death of macaque PNb at 6 months, was amplified in culture of peripheral blood mononuclear cells (PBMC) from a healthy macaque (12). Supernatant fluid from this culture (SHIV_KU-1) had a titer of 10⁴ 50% tissue culture infectious doses (TCID₅₀)/ml in macaque PBMC and in C8166 cells. Aliquots of this SHIV_KU-1 stock were stored in liquid nitrogen.

Construction of vpunef SHIV-4. The original SHIV-4 DNA consisted of two plasmids with the 5' and 3' regions, respectively. All manipulations were performed with the p3'SHIV-4. In the first step, as shown in Fig. 1, plasmid pUC19 was digested with SspI and NdeI, blunt ended with the Klenow fragment of DNA polymerase, and ligated to generate pDS. The SphI-KpnI fragment of p3'SHIV-4 was subcloned into pDS to generate pDSvpu. Plasmid pDSvpu was digested with SspI and BbsI, blunt ended, and religated to delete a 60-bp SspI-BbsI fragment and generate pDSDvpu. Figure 2 shows the construction of vpunef SHIV-4. First, the SphI-KpnI fragment of 3'SHIV-4 was replaced with the corresponding fragment of pDSDvpu to yield p3'vpuSHIV-4. Next, p3'vpuSHIV-4 was digested with RsrII and NcoI, blunt ended, and ligated, resulting in the deletion of an RsrII-NcoI fragment (including 205 bp of the nef gene, encoding the first 69 amino acids of Nef). This plasmid was designated pvpunefSHIV-4. Plasmids pvpunefSHIV-4 and p5'SHIV-4 were digested with SphI and ligated with T4 DNA ligase, and the ligated DNA was used to transfect C8166 cells as described previously (27). Virus stocks were prepared, and aliquots were stored at 80°C. The virus stock had a titer of 10⁴ TCID₅₀/ml in C8166 cells.

View larger version (15K): [in this window] [in a new window]   FIG. 1.   Construction of recombinant pDSDvpu. Open bars, pUC19 sequences; shaded bars, HIV-1 sequences.

View larger version (17K): [in this window] [in a new window]   FIG. 2.   Genomic structures of 3'SHIV-4, 3'vpuSHIV-4 and 3'vpunefSHIV-4. Striped bars, HIV-1 sequences; solid bars, SIVmac239 sequences. LTR, long terminal repeat.

Construction of vpuSHIV_PPc. The construction of SHIV_PPc has been described previously (26). Briefly, we used plasmid pSHIV_PPc-3, which is a derivative of plasmid p3'SHIV-4 containing all of the changes found in the env and nef genes of SHIV_PPc. The SphI/KpnI fragment from pDSDvpu, containing the 60-bp deletion in vpu, was subcloned into the SphI and KpnI sites of pSHIV_PPc-3. The resulting plasmid was designated pvpuSHIV_PPc (Fig. 3). Plasmids p5'SHIV-4 and pvpuSHIV_PPc were digested with SphI and ligated with T4 DNA ligase, and the ligated DNA was used to transfect C8166 cells as described above.

View larger version (27K): [in this window] [in a new window]   FIG. 3.   Genomic structure of vpuSHIVPPc. , derived from PPc spleen; , derived from PPc lymph node. LTR, long terminal repeat.

Vaccination of macaques and challenge with SHIV_KU-1. Sixteen sexually mature (3- to 15-year-old) female pig-tailed macaques (Macaca nemestrina) were obtained from the Yerkes Primate Center, Atlanta, Ga. Six were inoculated subcutaneously, close to the inguinal and axillary lymph nodes, with 1.0 ml of vpunefSHIV-4 stock (vaccine 1), while two control animals (PFy and PLy) were mock inoculated with saline. Seven months (29 weeks) after vaccination, the six vaccinated and two control macaques were challenged with 1.0 ml of undiluted SHIV_KU-1 stock which was inoculated intravaginally in nontraumatic fashion as described previously (9). This inoculum represents 30 50% animal infectious doses determined by titration of the virus in animals via the intravaginal route of inoculation (9). A second intravaginal challenge dose of SHIV_KU-1 was given 5 weeks after the first challenge to ensure exposure to this virus. Six other macaques were vaccinated with 1.0 ml of vpuSHIV_PPc stock (vaccine 2) by the oral route, while two control animals (4267 and PDb) were mock inoculated with culture medium. Six months later, all eight macaques were challenged by intravaginal inoculation twice, 1 day apart, with 1.0 ml of undiluted SHIV_KU-1 stock.

Cell cultures. The human T-cell line C8166 was used as the indicator line to measure virus infectivity. Cells were cultured at a concentration of 10⁶/ml in RPMI medium (RPMI 1640 supplemented with 10 mM HEPES buffer [pH 7.3], 50 µg of gentamicin per ml, 50 µM 2-mercaptoethanol, and 2 mM glutamine) with 10% fetal bovine serum (FBS).

Processing of samples. Heparinized blood obtained from the femoral vein was centrifuged to separate plasma and buffy coats. Plasma was assayed for p27 by using a capture enzyme-linked immunosorbent assay kit (Coulter Laboratories, Hialeah, Fla.) and for infectivity in C8166 cells. PBMC were separated from buffy coats by centrifugation through a Ficoll-Paque (Pharmacia Biotech, Piscataway, N.J.) density gradient. Infectious cell frequency was measured by inoculation of serial 10-fold dilutions of PBMC, starting with 10⁶ cells, into 24-well tissue culture plates containing 10⁵ indicator C8166 T cells, which were observed for development of syncytial cytopathic effects during a 7-day period; then cells and supernatant fluid in 100 µl from each well were transferred to another plate, fresh indicator cells were added, and observation continued for a further 7 days (14). Results were expressed as the number of infectious cells per 10⁶ PBMC. Mesenteric lymph nodes were obtained from vaccine group 1 at 19 weeks postchallenge, and axillary nodes were obtained from vaccine group 2 at 18 weeks postchallenge. Single-cell suspensions were prepared, and infectious cell frequency was assessed as described for PBMC. Other portions of the biopsy material were used for analysis of DNA. Inguinal lymph nodes were obtained later from animals in both groups (between weeks 54 and 60 for group 1 and weeks 31 and 37 for group 2) and snap-frozen in liquid nitrogen immediately upon removal. They were stored at 80°C until processed for RNA analysis.

Fluorescence-activated cell sorting analysis. PBMC or lymph node cells were reacted with monoclonal antibody to CD4⁺ T or CD8⁺ T cells (Dako, Carpinteria, Calif.). After washing, the cells were stained with fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse immunoglobulin G (Dako), fixed in 1% buffered formalin, and analyzed on a fluorescence-activated cell counter (12, 14).

Neutralizing antibody assays. We performed the test as described previously (10). Briefly, serial doubling dilutions of plasma in RPMI medium were prepared in quadruplicate in 96-well plates, 10 to 20 TCID₅₀ of the virus was added to each well, plates were incubated 1 h at 37°C, and 10⁴ indicator C8166 cells were added to each well. Plates were observed for cytopathic effect 7 days later, wells were scored individually, and the 50% neutralization endpoint was calculated by the Kärber method (15).

Detection of viral DNA in tissues. Tissues from inoculated animals were first screened for the presence of the SIV region of SHIV by PCR using oligonucleotide primers specific for both SIV and SHIV. Total cellular genomic DNA was extracted from PBMC and/or lymph nodes of the animals and was used as a template in nested PCR to amplify SIV gag sequences which were common to both viruses. The gag-specific oligonucleotides and the conditions for amplification were exactly as described previously (14).

Detection of viral RNA in lymph nodes. Snap-frozen lymph nodes were homogenized in Trizol reagent (Gibco-BRL, Gaithersburg, Md.), by using an Omni-mixer homogenizer (Omni International, Waterbury, Conn.). Total RNA was isolated as described by the manufacturer, and the final RNA was dissolved in 50 µl of distilled H₂O per 100 mg of original tissue. The quality of the RNAs was assessed by reverse transcriptase (RT)-mediated PCR (RT-PCR) for the cellular gene GAPDH mRNA as described previously (22), using the Titan One-Tube RT-PCR system (Boehringer Mannheim, Indianapolis, Ind.). The possibility of contaminating DNA in the RNA preparations was assessed by parallel reactions in which the RT activity in the reaction was first inactivated by 2 min at 99°C followed by 3 min at 95°C. If needed, RNase-free DNase (Gibco-BRL) was used to remove residual DNA from the samples, followed by extraction and precipitation. Sample RNAs were amplified by RT-PCR using primers for the SHIV pol gene (corresponding to the parent SIV sequence): SIVpolA (5'GAAAAGATGGAAAAGGATGG3') and SIVpolB (5'TGGCTTCTAATGGCTTGC3'). One microgram of total RNA was used in the one-step reaction containing the manufacturer's buffer and enzyme mix, appropriate primers, and 1.6 U of Prime RNase inhibitor (5 Prime3 Prime, Inc., Boulder, Colo.). The reactions were performed with a Perkin-Elmer DNA Thermal Cycler 480 with the following thermal profile: 42°C for 30 min, 1 cycle; 94°C for 5 min, 1 cycle; 94°C for 30 s, 55°C for 30 s, and 68°C for 45 s, 10 cycles; 94°C for 30 s, 55°C for 30 s, 68°C for 45 s, with an additional 5-s extension/cycle, 25 cycles; 68°C for 6 min. Nine microliters of the product was loaded onto an agarose-Tris-borate-EDTA gel containing ethidium bromide, DNA was separated by electrophoresis, and the gel was photographed. When no visible product was detected, 1 µl of the initial reaction mixture was added to a nested PCR mixture containing primers SIVpolC (5'ACCAATCCATACAACACC3') and SIVpolD (5'CTGCCCAATTTAATACTCC3'), 3 mM MgCl₂, and 1.25 U of Taq enzyme (Sigma, St. Louis, Mo.), and a further 35 cycles were performed with the following thermal profile: 97°C for 1 min 45 s, 55°C for 2 min, and 72°C for 5 min, 1 cycle; 94°C for 30 s, 55°C for 30 s, 72°C for 45 s, with an additional 1-s extension/cycle, 33 cycles; 94°C for 30 s, 55°C for 30 s, and 72°C for 6 min, 1 cycle. The amplified SIVpolAB fragment is 869 bp, whereas the SIVpolCD nested fragment is 645 bp. For the detection of multiply spliced mRNAs, primers spanning the intron of the tat and rev genes were used (8, 20) without the additional restriction sites; the second-round primers are heminested, containing one of the outer primers plus a nested primer. The PCR products of the Msp1 AB primer pair range from 131 to 159 nt, and those for Msp1 CB primer pair range from 110 to 138 nt. Second-round reactions included 1 M betaine (Sigma).

Quantitation of plasma RNA viral load. Plasma samples collected in acid citrate dextrose from animals in group 1 during weeks 52 and 58 and from group 2 during weeks 27 and 35 were analyzed by a real-time RT-PCR assay as previously described (28).

Lymphocyte proliferation assay. PBMC from different macaques, collected 49 weeks after virus challenge, were cultured in triplicate at 10⁵ cells/well in 96-well tissue culture plates in 200 µl of RPMI 1640 containing 10% FBS. Stock SHIV_KU-1 (10^4.2 TCID₅₀/ml) was UV irradiated for 30 minutes and heat treated at 56°C for 60 min before use as antigen; 20 µl of this material was added to each well. Three wells each containing unstimulated PBMC from different pig-tailed macaques served as negative controls. Cells were cultured for 6 days before the addition to each well of 1 µCi of [³H]thymidine (specific activity, 247.9 GBq/mmol; NEN, Boston, Mass.). Cultures were harvested onto glass fiber filter mats by using an automated plate harvester (Skatron, Sterling, Va.) 24 h after addition of [³H]thymidine, and [³H]thymidine incorporation was determined with a Microbeta liquid scintillation counter (Packard). Stimulation index (SI) was calculated as mean counts per minute in stimulated wells/mean counts per minute in control well, and an SI of greater than 1.5 was considered significant (3).

Development of CD4⁺ T-cell clones. CD4⁺ cells from group 1 macaques PDj and PNa (week 50 postchallenge) were negatively selected from PBMC first by incubating PBMC with mouse anti-human CD8 monoclonal antibody and then with anti-mouse immunoglobulin G magnetic beads (Dynal, Lake Success, N.Y.). Cells enriched for CD4⁺ T cells were then inoculated with herpesvirus saimiri (American Type Culture Collection, Manassas, Va.) to generate immortalized CD4⁺ T-cell lines (2). CD4⁺ T-cell clones were derived by the limiting-dilution method, and the phenotype was confirmed by staining with mouse anti-human CD4⁺ antibody. These CD4⁺ T cell clones were inoculated with SHIV_KU-1 at a multiplicity of infection of about 0.1 and used 7 days later as stimulators or targets.

Generation of bulk cytotoxic T-lymphocyte (CTL) population and chromium release assay. PDj and PNa PBMC were collected 50 weeks after virus challenge and cocultured with UV-irradiated SHIV_KU-1-infected autologous CD4⁺ T cells (effector/stimulator ratio of 10:1) in 24-well tissue culture plates in RPMI 1640 containing 10% FBS. They were restimulated on day 7 and used as effectors on day 14 in a chromium release assay.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Uncoagulated blood samples were collected weekly for the first month, at 2-week intervals for the next month, and monthly thereafter from all inoculated animals. Fresh plasma was tested for infectivity, content of p27, viral RNA, and antiviral antibodies. PBMC were examined for virus content (infectious center assays) and for CD4⁺ and CD8⁺ T-cell markers and were used for cellular immunity studies.

Vaccine group 1. Six macaques were inoculated subcutaneously with 1 ml of tissue culture fluid containing 10⁴ TCID₅₀ of vpunefSHIV-4, and heparinized blood was collected from all six according to the schedule outlined above to assess virus replication and host responses. Cocultivation of 10⁶ PBMC of each animal with C8166 cells failed to yield infectious virus from any of the vaccinated animals at any time point (Table 1), even after depletion of CD8⁺ T cells. However, PCR analysis of PBMC from all six revealed SIV_mac gag DNA sequences, showing that all six had become infected with the vaccine virus. Proof of infection was substantiated by immunoprecipitation analysis of plasma obtained 16 weeks after vaccination. This assay showed that all six had developed binding antibodies to SIV_mac Gag and HIV-1 Env proteins (data not shown). By 19 weeks, all six had developed neutralizing antibodies at titers of 1:10 and 1:20 to SHIV_KU-1. CD4⁺ T-cell counts remained normal in all of the six vaccinated macaques. Thus, although the vaccine virus caused a persistent infection, there was little evidence of productive replication of the virus.

View larger version (45K): [in this window] [in a new window]   FIG. 4.   PCR detection of viral sequences from lymph node biopsies 19 weeks after challenge of vaccine group 1 with SHIVKU-1. Lymph node biopsies were performed; DNA was extracted and used in nested PCR with oligonucleotides that amplified the vpu gene of SHIVKU-1 as described in the text. Aliquots of the nested PCR were run on a 1.5% agarose gel and stained with ethidium bromide. Lanes 1 to 7, amplification of vpu sequences from lymph node DNA of macaques 42105, 42107, PLk, PPm, PDj, PNa, and PFy, respectively; lane 8, amplification of vpu sequences from plasmid p3'SHIV (positive control for full length vpu gene); lane 9, amplification of vpu sequences from plasmid pDSDvpu (positive control for truncated vpu gene); lane 10, amplification of vpu sequences from plasmids p3'SHIV and pDSDvpu; lane 11, amplification of vpu from uninfected PBMC DNA (negative control).

Vaccine group 2. Since the death of macaque 42107 and the low-grade persistent infection in macaque 42105 may have been the result of inadequate immunity resulting from poor replication of vaccine virus 1, we sought to use another vaccine virus that had the potential for more robust replication and also one that could be administered orally. As shown in previous reports, SHIV-4 had become virulent after sequential passage in macaques (12). Macaque PPc was the first animal to develop AIDS, and the disease course correlated with a number of factors, including high replication efficiency of the virus in CD4⁺ T cells and macrophages, infectious viremia, infection in high numbers of circulating PBMC in peripheral blood, and loss of CD4⁺ T cells (12). New genetic changes in the virus included an open, functional vpu gene and numerous consensus amino acid changes in the variable and constant regions of the env and nef genes (17, 25). Since vpu appeared to be important for pathogenicity, we deleted this gene but used the env and nef genes of the PPc virus in a new construct, termed vpuSHIV_PPc (Fig. 3). In contrast to SHIV-4, vpuSHIV_PPc is macrophagetropic as a consequence of mutations in the env genes of SHIV_PPc. This property is probably important for the successful systemic infection that occurs following inoculation of this virus on mucosal surfaces.

vpunef

View larger version (63K): [in this window] [in a new window]   FIG. 5.   RT-PCR detection of full-length viral RNA and spliced mRNA in lymph nodes. Total RNA was extracted from snap-frozen inguinal lymph nodes from control and vaccinated animals. RT-PCRs were performed with 1 µg of total RNA, except as noted; nested or heminested PCRs were performed with 1 µl from the RT-PCR. (A to C) Nested PCR with SIVpolCD primers. (A) Macaque 8124, lanes 1 and 2; 42106, lanes 3 and 4; PNa, lanes 5 and 6; PPm, lanes 7 and 8; heated controls in the RT-PCR step, lanes 1, 3, 5, and 7. (B) PWv, lanes 1 and 2; 7024, lanes 3 and 4; PEy, lanes 5 and 6; PWl, lanes 7 and 8; PDb, lanes 9 and 10; heated controls, lanes 1, 3, 5, 7, and 9. (C) PLk, lanes 1 and 2; 42105 (5 µg), lanes 3 and 4; PDj, lanes 5 and 6; PFy, lanes 7 and 8; heated controls, lanes 1, 3, 5, and 7. (D to F) Multiply spliced primers. (D) RT-PCR with MsplAB primers. PLk, lane 1; 42015, lane 2; PDj, lane 3; PFy, lane 4; PWv, lane 5; 7024, lane 6; PEy, lane 7; PWl, lane 8; PDb, lane 9. (E) Heminested reactions with MsplBC primers. Lanes 1 to 9 are as described for panel D. (F) Heminested reactions with MsplBC primers. Macaque 8124, lanes 1 to 3 (lanes 1 and 2, 1 µg; lane 3, 5 µg); 42106, lanes 4 and 5; PNa, lanes 6 and 7; PPm, lanes 8 and 9; heated controls, lanes 1, 4, 6, and 8; reactions with no RNA added to original RT-PCR step, . Lanes c in all panels represent control reactions with related KU2 RNA.

View larger version (11K): [in this window] [in a new window]   FIG. 6.   SHIVKU-1-specific CTL activity in macaques immunized with vaccine 1 and challenged with SHIVKU-1 at various E/T ratios. PBMC from macaques PDj (A) and PNa (B) collected 50 weeks postchallenge were stimulated in vitro with SHIVKU-1-infected, UV-irradiated autologous CD4+ T cells and used as effectors in a chromium release assay on day 14. Target cells used in the experiment were autologous CD4+ T cells either infected with SHIVKU-1 () or sham infected () and labeled with 51Cr.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Six sexually mature macaques inoculated orally on a single occasion with vpuSHIV_PPc (group 2) developed an uneventful, transiently productive infection that conferred protection against AIDS caused by intravaginally inoculated, highly virulent SHIV_KU-1 (² test, P = 0.005). All four of the unvaccinated control macaques inoculated intravaginally with this virus developed highly productive infection, subtotal loss of CD4⁺ T cells, and AIDS, confirming our identical, previously reported finding (9) on development of disease in six of six animals that became infected after nontraumatic intravaginal infusion of this virus. The first vaccine virus, vpunefSHIV-4, was incapable of causing infection after mucosal inoculation (and therefore was injected subcutaneously) and replicated poorly in all six macaques, but nevertheless induced binding and neutralizing antibodies to the virus, proving infection. Four of these six vaccinated macaques resisted productive infection and disease following intravaginal challenge with SHIV_KU-1 (² test, P = 0.102). Of the two animals that developed productive infection, one developed AIDS and another remains at risk for developing AIDS. Given this prognosis, only 2 of the 12 vaccinated macaques, compared to four of four of the control animals, developed disease (² test, P = 0.003). Mechanisms of vaccine-induced protection among the 10 resistant vaccinated macaques are still being evaluated, but the purpose of this report is to illustrate proof of concept that the sexually transmitted disease caused by HIV-1 in human beings could probably be prevented by prophylactic immunization.

Both vaccine viruses caused persistent infection irrespective of whether they replicated productively during the first few weeks following inoculation. Six months following immunization, even though infectious virus could no longer be isolated from PBMC from any of the 12 vaccinated macaques, vaccine virus DNA was still detectable. All 12 animals had developed binding and very low neutralizing antibody titers to SHIV_KU-1 in plasma by the time of virus challenge (approximately 6 months following vaccination). As shown, all 12 became infected with the challenge virus, since its DNA was clearly detectable in PBMC or mesenteric lymph node tissue a few months following challenge. Two of the six vaccinated macaques in group 1 developed a persistent productive infection in PBMC and one, 42107, succumbed to AIDS approximately 6 months following challenge. The other macaque, 42105, had a phase of low-grade productive infection for 30 weeks followed by a 20-week period when it had no circulating infectious PBMC in blood. However, from week 50 onward, infectious PBMC, accompanied by relatively low concentrations of viral RNA (5,000 to 70,000, versus 4 × 10⁵ to 3 × 10⁶ in controls) have reappeared in blood. At 60 weeks postchallenge, this animal still has large numbers of uninfected CD4⁺ T cells (which are highly susceptible to infection), in contrast to control animals, nearly all of whose CD4⁺ T cells became infected and eliminated during the first 3 weeks of infection. Nevertheless, the resurgence of virus in macaque 42105 suggests that this animal may be at risk for developing late-onset AIDS.

The other 10 vaccinated macaques, 4 in group 1 and all six in orally immunized group 2, developed a unique type of control over replication of SHIV_KU-1. Rarely and transiently, small numbers of infectious cells have appeared in blood, but at nearly all time points, replication-competent virus could not be isolated from PBMC or lymph nodes (sampled two to three times during this period of observations) by any procedure. Attempts have included three successive cocultivations of CD8-depleted PBMC with mitogen-activated normal PBMC or T-cell lines. Nevertheless, both viral DNA and viral RNA have persisted in lymph node tissue. Further, the detection of spliced viral RNA in lymph nodes of a few animals in both groups suggested that an extremely low level of replication of SHIV_KU-1 has probably been in progress continuously in all of the vaccinated macaques. This type of replication of this virus had never been observed before. In our earlier study on pathogenesis of infection following intravaginal inoculation of SHIV_KU-1 (9), we had shown that viral DNA became detectable by day 2 after inoculation in mesenteric lymph nodes and that explosive replication of the agent occurred during the following week. In this study, although the intravaginal route of inoculation was shown to be about 30-fold less sensitive than the intravenous route, all animals that became infected after intravaginal inoculation nevertheless developed disease. Extrapolating these data to our vaccinated macaques, since all had developed infection in lymph nodes, we surmised that systemic rather than mucosal factors were responsible for curtailment of virus replication.

It is doubtful that neutralizing antibodies had a significant role in curtailment of virus replication in the vaccinated macaques. In an earlier report (7), we showed that passively administered immune serum was effective only when given before parenteral inoculation of the virus. This serum had no therapeutic effect when given 2 h following virus inoculation. Worse still, unlike the success at preventing infection after parenteral inoculation of the virus, we were unable to protect macaques from the dire effects of infection by infusing the serum 24 h before mucosal inoculation of the agent (13). Immune serum therefore had no apparent effect on an already established infection, a well-known phenomenon in lentivirus infections, nor apparently could the previously administered serum prevent infection across mucosal surfaces. Restriction of virus replication in our 10 vaccinated macaques must therefore have been mediated by factors other than neutralizing antibodies.

Studies on CMI responses of the vaccinated macaques have only recently begun. That this type of immunity is important came from an earlier finding that macaques that had "recovered" from infection with avirulent SIV_mac were protected from effects of infection with SHIV_KU-1 (24). Since these two viruses have distinctly different envelopes, the protection was probably mediated by CMI responses to viral core proteins shared by the two viruses. In the present study, our findings that CD4⁺ T cells in the vaccinated macaques at week 50 in group 1 and in week 30 in group 2 proliferated when exposed to antigens of SHIV_KU-1 showed that the animals had developed anti-SHIV CMI responses that have persisted along with the infection. The further preliminary finding that 2 of the 10 vaccinated macaques which have controlled virus replication have developed CTL to the virus has strengthened the premise that CMI responses are responsible for control and curtailment of replication of the virulent virus. Possibly, a dynamic interaction between virus replication and CTL curtailment responses is continuously at play. A recent study of CTL responses in HIV-infected humans showed an inverse correlation between the levels of CTL and plasma viral RNA loads (19), implicating a role for CTL control of viral replication and delay of disease progression. We hope that more complete studies on CMI responses among our vaccinated macaques, currently in progress, will shed new light on the mechanisms of curtailment of virus replication in immune animals. However, although mechanisms of protection against SHIV_KU-1 are still not understood fully, the present study clearly lends support to the concept that sexually transmitted HIV-1 disease can be averted by immunization resulting from a single exposure to a live attenuated vaccine.