AAV-Delivered Antibody Mediates Significant Protective Effects against SIVmac239 Challenge in the Absence of Neutralizing Activity

Abstract

Long-term delivery of potent broadly-neutralizing antibodies is a promising approach for the prevention of HIV-1 infection. We used AAV vector intramuscularly to deliver anti-SIV monoclonal antibodies (mAbs) in IgG1 form to rhesus monkeys. Persisting levels of delivered mAb as high as 270 μg/ml were achieved. However, host antibody responses to the delivered antibody were observed in 9 of the 12 monkeys and these appeared to limit the concentration of delivered antibody that could be achieved. This is reflected in the wide range of delivered mAb concentrations that were achieved: 1-270 μg/ml. Following repeated, marginal dose, intravenous challenge with the difficult-to-neutralize SIVmac239, the six monkeys in the AAV-5L7 IgG1 mAb group showed clear protective effects despite the absence of detectable neutralizing activity against the challenge virus. The protective effects included: lowering of viral load at peak height; lowering of viral load at set point; delay in the time to peak viral load from the time of the infectious virus exposure. All of these effects were statistically significant. In addition, the monkey with the highest level of delivered 5L7 mAb completely resisted six successive SIVmac239 i.v. challenges, including a final challenge with a dose of 10 i.v. infectious units. Our results demonstrate the continued promise of this approach for the prevention of HIV-1 infection in people. However, the problem of anti-antibody responses will need to be understood and overcome for the promise of this approach to be effectively realized.

Fig 1. Serum concentration of the IgG1 mAbs 4L6 and 5L7, and emerging anti-mAb responses following recombinant AAV administration.

Levels of produced antibodies were measured in sera from immunized animals over time by a gp140 capture ELISA. SIV challenge is indicated by arrows. (A) Animals that received rAAV-5L7 IgG1. 84–05, 153–10 and 157–10 were given one ssAAV vector expressing both heavy and light chain of the antibody 5L7 IgG1. 111–09, 120–09 and 154–10 were given an equal mixture of two scAAV vectors expressing either heavy chain or light chain of the antibody 5L7 IgG1. (B) Animals that received rAAV-4L6 IgG1. All six animals were given one ssAAV vector expressing both heavy and light chain of the antibody 4L6 IgG1. (C and D) Antibody responses to 5L7 IgG1 and 4L6 IgG1, respectively, following recombinant AAV administration. Humoral immune responses were measured in sera from immunized animals over time by ELISA. The reactivity of serum was tested against homologous purified protein using a conjugated anti-lambda secondary antibody for detection since both 5L7 IgG1 and 4L6 IgG1 bear a kappa light chain. (E) Reactivity of sera from 5L7 IgG1 recipients to purified mAbs and immunoadhesins at week 10 after AAV administration. (F) Reactivity of sera from 4L6 IgG1 animals to purified mAbs and immunoadhesins at week 6 after AAV administration.

Fig 2. Repeated low-dose SIVmac239 challenge of AAV-immunized and control animals.

(A) Kaplan-Meier analysis of the three test groups. The percentage of animals remaining uninfected is plotted against the number of SIV exposures. SIV virus challenges were conducted in 3-week intervals. Four low-dose challenges with one half-maximal infectious dose (1x ID50) were followed by a two ID50 (2x) and a subsequent ten ID50 (10x) challenge. The survival curves are not significantly different (Mantel-Cox test). (B) Peak viral load comparison. The geometric mean (red line) is significantly different in the 5L7 IgG1 group compared to the control group (two-tailed, unpaired t test). (C) Set-point viral load analysis. All values within one group between weeks 8 and 25 were averaged, the resulting values were logarithmized (log10) and compared to each other. The geometric mean of the 5L7 IgG1 group differs significantly from the other two groups (two-tailed, unpaired t test). (D) Time of peak viremia. Most of the immunized animals show peak viremia at week 3, while 5 of 6 controls peak at week 2. (E and F) Viral loads in vivo after infectious exposure with SIVmac239. Viral loads in plasma of AAV animals (blue) and controls (red) measured as SIV RNA genome equivalent per ml are shown as a function of time since the infectious exposure. Control animal 139–08 was a rapid progressor and had to be sacrificed after 9 weeks due to AIDS-related symptoms. Test animal 84–05 remained uninfected after 6 SIV challenges.

Fig 3. Neutralization of SIV and binding of purified proteins in vitro.

The dashed line indicates 50% RLU (relative light units) representing 50% neutralization activity against the tested SIV strain. Lowest RLU indicates highest neutralization. Binding of purified proteins was tested by a SIVmac239 gp140 ELISA, high absorbance indicates high binding. SPF serum (specific pathogen free) from naïve animals was used as negative control, a pool of antisera from SIV-infected animals served as positive control. (A) None of the antibodies or immunoadhesins showed neutralization against the resistant virus strain SIVmac239 tested at 50 μg/ml. Similar results were obtained in both the SEAP and TZMbl assays. (B) Neutralization curve of SIVmac239 with 5L7 IgG1 starting at 1mg/ml. (C) All of the tested antibodies and immunoadhesins showed equivalent neutralization activity against SIVmac316 with an average IC50 (half-maximal inhibitory concentration) of 0.002 μg/ml. (D) Animal sera from week 10 after rAAV administration (see Fig 1A) were diluted to 4 μg/ml based on previous ELISA quantitations, and tested for neutralizing activity against SIVmac316. The average IC50 measured corresponds to the IC50 of the purified proteins. (E) Animal sera of 84–05 from weeks 1 and 12 after AAV administration lacked detectable neutralizing activity against SIVmac239. (F) Antibodies and immunoadhesins were tested for their ability to bind SIVmac239 gp140. All of the antibody constructs showed equivalent binding activity as determined by absorbance at 450 nm.

Fig 4. ADCC activity of purified proteins and sera against SIV-infected target cells in vitro.

The dashed line indicates 50% RLU (relative light units) or 50% ADCC activity against SIV-infected target cells. ADCC was measured by the luciferase activity in SIV-infected cells after an 8 to 10 h incubation in the presence of a macaque CD16⁺ NK cell line and a serial dilution of antibodies or animal sera. The loss of RLU indicates the loss of virus-infected cells during the 8 to 10 h incubation period and represents a high ADCC activity. Purified 5L7 IgG1 was diluted to match the equivalent 5L7 IgG1 concentration of 5L7 IgG1-containing animal sera from week 21 post AAV administration, and compared for mediating ADCC towards (A) SIVmac239-infected cells and (B) SIVmac316-infected cells. The ADCC activities of sera from 153–10 and 157–10 overlapped with the purified protein at their corresponding concentration. 5L7 IgG1-containing animal serum from 84–05 mediated superior ADCC activity compared to purified 5L7 IgG1 at equivalent concentration of this antibody.