Massive anamnestic CD8⁺ lymphocyte responses were measured by tetramer staining after challenge with SIVmac239. Mamu A*01 Gag CM9_181-189 (□) and Mamu A*01 Tat SL8_28-35 (○) tetramers were used to stain PBMC. Tetramer-positive populations are expressed as percentages of CD3⁺ CD8⁺ lymphocytes. Plasma viral concentrations are shown as solid black lines. The figure also includes a geometric mean analysis of each cohort. Clearly, the vaccinees had stronger and earlier tetramer-positive populations to both Gag CM9_181-189 and Tat SL8_28-35, as seen in the comparison figure. Summary figures are in blue for vaccinees and red for controls.

Flow analysis of CD4⁺ populations. PBMC were stained with antibodies to CD3, CD4, CD28, CCR5, and CD95 and read on an LSR II flow cytometer. (a) Cells were first gated through a lymphocyte gate and then gated through CD3⁺ and CD4⁺ gates. The resultant population was then gated for CD28 versus CD95 or CCR5 versus CD95, and populations were labeled as indicated. Cells are gated through a lymphocyte gate by forward and side scatter and then gated subsequently for CD3⁺ and CD4⁺ populations. (b) Effector memory is defined as the population of cells in the CD95 versus CD28 dot plot as being CD95⁺ CD28⁻. Central memory is CD95⁺ CD28⁺, and naive CD4⁺ T cells are CD95⁻ CD28⁺. (c) CCR5⁺ memory cells are CCR5⁺ and CD95⁺. FITC, fluorescein isothiocyanate.

Six of seven vaccinees did not develop neutralizing antibodies by days 170 to 210 postinfection. Neutralizing antibodies were analyzed using pseudovirions with the SIVmac239 envelope, which is a more sensitive test than by infection of PBMC with wild-type virus (68). The pseudovirus contains luciferase, and results indicate how much the luciferase signal is decreased by neutralizing antibodies present in the plasma. Animal numbers and viral loads (VL) at the time of the assay are indicated. A comparison of individual viral loads and the geometric mean for each group are shown in Fig. 3. Results are displayed as percentages of neutralization. Prevaccination assays are indicated by tan triangles, prechallenge assays are denoted by light blue circles, and postinfection results are displayed as a blue diamonds.

Five of six control animals developed neutralizing antibodies by day 170 to 210 postinfection. Neutralizing antibodies were analyzed using pseudovirions with the SIVmac239 envelope, which is a more sensitive test than by infection of PBMC with wild-type virus (68). The pseudovirus contains luciferase, and results indicate how much the luciferase signal is decreased by neutralizing antibodies present in the plasma. Animal numbers and viral loads (VL) at the time of the assay are indicated. A comparison of individual viral loads and the geometric mean for each group are shown in Fig. 3. Results are displayed as percentages of neutralization. Prevaccination assays are indicated by tan triangles, prechallenge assays are denoted by light red circles, and postinfection results are displayed as red diamonds.

Experimental design. Eight Mamu-A*01⁺ Indian rhesus macaques (blue outline) were given DNA plasmid encoding Gag, Tat, Rev, and Nef added to the CRL1005/benzalkonium chloride adjuvant on weeks 0, 4, and 8. Each construct was given intramuscularly into a separate limb to minimize immunodominant suppression of a broad immune response. The same limb was used for the same encoded protein on each occasion. At week 24, Ad5 vectors, encoding the same four proteins, were given intramuscularly into the same limb as that used for the DNA primes. At week 50, the 8 vaccinees and 8 naive controls (red outline) were challenged with a low dose (300 TCID₅₀) of SIVmac239 (arrows) until a day 7 blood draw indicated that the animal had a positive plasma viremia by quantitative PCR. LTR, long terminal repeat.

Plasma viral concentrations are lower in vaccinees. Plasma viral concentration determinations were carried out by quantitative RT-PCR of plasma samples. (a) The majority of the vaccinees (in blue) had plasma viral concentrations of less than 10⁵ copies/ml, including several animals whose plasma viral concentrations were less than 10⁴ copies/ml, and at times, approached the limits of detection (open symbols). (b) However, control animals (in red) had chronic-phase plasma viral concentrations greater than 10⁵ copies/ml. (c) Vaccinees had a significant reduction in peak viremia and chronic-phase set point. Taking a geometric mean for each group, we observed a significant (P = 0.007) reduction in peak viremia, from 5 × 10⁷ on day 14 for control animals to 4 × 10⁶ on day 11 for vaccinees. Chronic-phase viremia was also significantly decreased (P = 0.0192), from 1.49 × 10⁵ in control animals, to 5,300 copies/ml in vaccinees. This is a reduction of plasma viral concentration of nearly 1.5 logs. (d) Compared to plasma viral concentrations for the Gag-only vaccine study (11), by 80 days postinfection, viral loads in the Gag-only study are increasing, whereas in the current study, control of plasma viremia is maintained well into the late chronic phase.

Loss of CD4⁺ memory populations is reduced in vaccinees at 140 to 180 days postinfection. PBMC were stained with antibodies to CD3, CD4, CD28, CCR5, and CD95 and read on an LSR II flow cytometer (see Fig. 5 for gating analysis). (a) The effector memory subset is significantly (P = 0.0448) preserved in vaccinees (blue) compared to controls (red). (b) The CCR5⁺ population is also preserved in vaccinees (P = 0.0195). (c) Overall CD95⁺ CD4⁺ memory is preserved in vaccinees compared to controls (P = 0.0145). (d) Overall, the CD4⁺ compartment is more robust in vaccinees than controls (P = 0.0025). (e) Analysis of longitudinal samples from frozen PBMC reveals that there is a drop in memory populations in control animals (red) during the acute phase, whereas some subsets actually increase in vaccinees (blue) during immediate acute infection. This is seen in the effector subset as well as the CCR5⁺ subset. At week 2, rather than a depletion of this subset, there is actually an increase in these two populations in the vaccinated animals, while in the control animals, there is a decrease in these populations at this time point. The CD95⁺ population is slightly higher in vaccinees at week 1 postinfection, likely because memory cells were engendered by the vaccination and the anamnestic response is already beginning to expand this population. By 25 weeks postinfection, the CCR5⁺ subset is recovered to prechallenge levels in both control and vaccinated animals. However, the effector memory populations are still diverging, and the vaccinated animals have a net gain in this population. Normalized CD4⁺ absolute counts are protected throughout the infection, both during the acute and early chronic phase (day 0 to 196; P = 0.0092) as well as during late chronic phase (day 203 to 336; P = 0.0062), during which time vaccinees are recovering their CD4⁺ absolute counts, moving toward prechallenge values.