DNA vaccination affords significant protection against feline immunodeficiency virus infection without inducing detectable antiviral antibodies

Abstract

To test the potential of a multigene DNA vaccine against lentivirus infection, we generated a defective mutant provirus of feline immunodeficiency virus (FIV) with an in-frame deletion in pol (FIVDeltaRT). In a first experiment, FIVDeltaRT DNA was administered intramuscularly to 10 animals, half of which also received feline gamma interferon (IFN-gamma) DNA. The DNA was administered in four 100-microg doses at 0, 10, and 23 weeks. Immunization with FIVDeltaRT elicited cytotoxic T-cell (CTL) responses to FIV Gag and Env in the absence of a serological response. After challenge with homologous virus at week 26, all 10 of the control animals became seropositive and viremic but 4 of the 10 vaccinates remained seronegative and virus free. Furthermore, quantitative virus isolation and quantitative PCR analysis of viral DNA in peripheral blood mononuclear cells revealed significantly lower virus loads in the FIVDeltaRT vaccinates than in the controls. Immunization with FIVDeltaRT in conjunction with IFN-gamma gave the highest proportion of protected cats, with only two of five vaccinates showing evidence of infection following challenge. In a second experiment involving two groups (FIVDeltaRT plus IFN-gamma and IFN-gamma alone), the immunization schedule was reduced to 0, 4, and 8 weeks. Once again, CTL responses were seen prior to challenge in the absence of detectable antibodies. Two of five cats receiving the proviral DNA vaccine were protected against infection, with an overall reduction in virus load compared to the five infected controls. These findings demonstrate that DNA vaccination can elicit protection against lentivirus infection in the absence of a serological response and suggest the need to reconsider efficacy criteria for lentivirus vaccines.

FIG. 1

Partial restriction map of the FIV-PET F-14 molecular clone which was received as a subclone in plasmid pGEM7Zf+. The location of the major open reading frames (stippled boxes) and long terminal repeats (solid boxes) is shown underneath. The unique PacI site (P) used in mutagenesis and an adjacent XbaI site (X) are boxed in the diagram and underlined in the primary sequence below. The extent of the Bal 31-generated deletion in FIVΔRT is indicated by dots above the sequence. Numbering is relative to the published sequence (GenBank accession no. M25381 [28]). Other restriction sites: S, SacI; R, EcoRI; N, NcoI; K, KpnI; B, BglII.

FIG. 2

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of sedimented virus particles released from FL4 cells persistently infected with FIV-PET (lane1), mock-transfected CrFK cells (lane 2), CrFK cells transfected with the F-14 molecular clone (lane 3), and CrFK cells transfected with the FIVΔRT construct (lane 4).

FIG. 3

Design of the two trials. Timings of immunizations and challenges (in weeks) are represented by small and large arrows, respectively. CTL and serological assays were performed at the intervals denoted by the triangles and symbols representing antibodies, respectively.

FIG. 4

FIV Gag- and Env-specific effector CTL responses were measured directly on the fresh PBMC collected from vaccinated (FIVΔRT [a]; FIVΔRT plus IFN-γ [b]) and control (IFN-γ [c]; PBS alone [d]) cats in trial 1, 3 weeks after the first immunization. Autologous or allogeneic skin fibroblasts infected with recombinant vaccinia viruses expressing either FIV Gag (10) (■ and ░⃞) or FIV Env (41) (▨ and ▨) or wild-type vaccinia virus (□) were labelled with ⁵¹Cr and used as targets in the assay. The release of ⁵¹Cr into the culture supernatant was detected after 4 h of incubation at 37°C. The results shown represent the mean values for triplicate cultures at an E/T ratio of 50:1.

FIG. 5

Detection of FIV Gag- and Env-specific CTL responses directly in fresh PBMC collected from vaccinated (FIVΔRT [a]; FIVΔRT plus IFN-γ [b]) and control (IFN-γ [c]; PBS alone [d]) cats in trial 1, 6 weeks after the first immunization. Autologous or allogeneic target cell express either FIV Gag (■ and ░⃞) or FIV Env (▨ and ▨) or no FIV proteins (infected with wild-type vaccinia virus [□]). The results shown represent the mean ⁵¹Cr release after a 4-h incubation at 37°C for triplicate cultures at an E/T ratio of 50:1.

FIG. 6

Detection of FIV Gag- and Env-specific CTL responses directly in fresh PBMC collected from vaccinated (FIVΔRT [a]; FIVΔRT plus IFN-γ [b]) and control (IFN-γ [c]; PBS alone [d]) cats in trial 1, 1 week after the second immunization. Autologous or allogeneic target cells express either FIV Gag (■ and ) or FIV Env (▨ and ▨) or no FIV proteins (infected with wild-type vaccinia virus [□]). The results shown represent the mean ⁵¹Cr release after a 4-h incubation at 37°C for triplicate cultures at an E/T ratio of 50:1. ND, not done.

FIG. 7

Immunoblot analysis performed as described previously (19) on plasma samples taken from cats 12 weeks after challenge. (a) Trial 1 (numbered lanes) and plasma from an FIV-infected and an unvaccinated, uninfected cat as controls (lanes + and −, respectively). One cat in the group immunized with ΔRT (lane 4) and three cats in the group immunized with ΔRT and IFN-γ (lanes 1, 3, and 5) remained antibody negative after challenge. (b) Trial 2 (numbered lanes) and control lanes as in panel a. Two cats (lanes 2 and 4) immunized with ΔRT and IFN-γ remained antibody negative after challenge.

FIG. 8

Viral loads in trial 1 at 7 weeks postchallenge (a) and in trial 2 at 6 weeks postchallenge (b), expressed as the mean ± 2 standard errors of the mean of the log-transformed maximum-likelihood estimates of the initial number of infected cells present in 2 × 10⁶ PBMC.