Recombinant mouse cytomegalovirus expressing a ligand for the NKG2D receptor is attenuated and has improved vaccine properties

Abstract

Human CMV (HCMV) is a major cause of morbidity and mortality in both congenitally infected and immunocompromised individuals. Development of an effective HCMV vaccine would help protect these vulnerable groups. NK group 2, member D (NKG2D) is a potent activating receptor expressed by cells of the innate and adaptive immune systems. Its importance in HCMV immune surveillance is indicated by the elaborative evasion mechanisms evolved by the virus to avoid NKG2D. In order to study this signaling pathway, we engineered a recombinant mouse CMV expressing the high-affinity NKG2D ligand RAE-1γ (RAE-1γMCMV). Expression of RAE-1γ by MCMV resulted in profound virus attenuation in vivo and lower latent viral DNA loads. RAE-1γMCMV infection was efficiently controlled by immunodeficient hosts, including mice lacking type I interferon receptors or immunosuppressed by sublethal γ-irradiation. Features of MCMV infection in neonates were also diminished. Despite tight innate immune control, RAE-1γMCMV infection elicited strong and long-lasting protective immunity. Maternal RAE-1γMCMV immunization protected neonatal mice from MCMV disease via placental transfer of antiviral Abs. Despite strong selective pressure, the RAE-1γ transgene did not exhibit sequence variation following infection. Together, our results indicate that use of a recombinant virus encoding the ligand for an activating NK cell receptor could be a powerful approach to developing a safe and immunogenic HCMV vaccine.

Figure 1. RAE-1γMCMV is attenuated in vivo in an NKG2D-dependent manner.

(A) The HindIII cleavage map of the MCMV genome is shown at the top, with the genomic region encoding the m152 ORF below. The m152 ORF was replaced by an expression cassette (bottom) comprising the HCMV major immediate early promoter (CMV-P), the RAE-1γ ORF, and the SV40 polyadenylation signal sequence. (B) SVEC4-10 cells, NIH 3T3 cells, and MEFs were infected with the indicated viruses and 12 hours later analyzed for the surface expression of RAE-1γ by staining with anti–RAE-1γ Ab, followed by PE-conjugated goat anti-rat IgG. Cells incubated with the secondary Ab in the absence of the primary Ab were used as negative control (thin line). Each histogram represents 10,000 gated propidium iodide–negative cells. (C) Untreated BALB/c mice or BALB/c mice treated with blocking anti-NKG2D mAb were i.v. injected with 10⁵ PFU of the indicated viruses. Viral titers were determined in lungs and spleen 3 days p.i. by plaque assay. Ø, untreated BALB/c mice. (D) BALB/c mice were f.p. injected with 2 × 10⁵ PFU of the indicated viruses. Viral titers were determined by plaque assay and (E) viral genome load by qPCR at different time p.i. Individual mice (circles) and median values (horizontal bars) are shown. DL, detection limit. Results from 1 of 2 similar experiments are shown. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 2. RAE-1γMCMV is attenuated in neonatal mice.

(A) Neonatal BALB/c mice were i.p. injected with 500 PFU of RAE-1γMCMV or WT MCMV 6 hours postpartum; viral titers were determined by plaque assay and (B) viral genome load by qPCR at the indicated times p.i. Individual mice (circles) and median values (horizontal bars) are shown. Data are representative of 2 experiments. *P < 0.05.

Figure 3. Comparable kinetics and phenotype of MCMV-specific memory CD8⁺ T cells in RAE-1γMCMV–, WT MCMV–, and Δm152 MCMV–infected mice.

(A) BALB/c mice were f.p. injected with 2 × 10⁵ PFU RAE-1γMCMV, WT MCMV, or Δm152 MCMV. Splenocytes were isolated at different times after infection and stained with IE1/m123 or m164 MHC class I tetramers and anti-CD8 Ab. The percentage of tetramer-specific CD8⁺ T cells for individual mice (circles) and median values (horizontal bars) are shown. (B) Splenocytes were isolated 9 months p.i. and stained with IE1/m123 MHC class I tetramer, anti-CD8 Ab, and Abs to indicated cell surface molecules. The percentage of IE1/m123-specific CD8⁺ T cells displaying effector memory (Tem) or central memory (Tcm) phenotype (left) and the percentage of Tem- and Tcm-expressing indicated cell surface molecules are shown (right). Error bars show mean ± SEM. (C) Representative histogram showing NKG2D staining on IE1/m123-specific CD8⁺ T cells in spleen 9 months after WT MCMV (filled histogram) or RAE-1γMCMV (dotted line) infection. Tetramer-negative CD8⁺ T cells are indicated by the dashed line. (D) Splenocytes were isolated 9 months p.i. and stained with the indicated tetramers or stimulated with the indicated peptides and stained for IFN-γ production or (E) costained for IFN-γ and TNF-α production. Splenocytes were stimulated as above in the presence of the anti-CD107a Ab and costained for IFN-γ production. (D and E) Representative dot plots gated on CD8⁺ T cells of 3 mice per group are shown. Numbers indicate means. Results from 1 of 2 similar experiments are shown.

Figure 4. RAE-1γMCMV infection induces protective immunity.

(A) Donors of memory CD8⁺ T cells were μMT/μMT B cell–deficient mice either naive or latently infected with RAE-1γMCMV or WT MCMV (>6 months p.i.). Splenocytes from 3 donors per group were pooled, and the number of MCMV-specific CD8⁺ T cells was assessed by combined staining with IE1/m123, m164, M83, M84, and m04 MHC class I tetramers. 10⁴ naive CD8⁺ T cells or graded numbers of MCMV-specific CD8⁺ T cells were i.v. transferred to recipient BALB/c mice immunocompromised by 6 Gy γ-irradiation. Recipients were f.p. injected with 10⁵ PFU WT MCMV 6 hours after the cell transfer. Viral titers in spleen were determined 12 days p.i. by plaque assay. Titers of individual mice (circles) and median values (horizontal bars) are shown. Ø, no transfer. (B) Mice infected as described in Figure 3 were i.p. challenged with 10⁵ PFU of salivary gland–derived MCMV (SGV) 6 months p.i. Lymphocytes were isolated from blood, spleen, and liver at different time points after the challenge and stained with IE1/m123 MHC class I tetramer and anti-CD8 Ab. The percentage of IE1/m123-specific CD8⁺ T cells for individual mice (circles) and median values (horizontal bars) are shown. (C) Naive mice and mice infected as described in Figure 3 were i.p. challenged with 2 × 10⁵ or 5 × 10⁵ PFU of SGV 6 months p.i. Survival rates were monitored daily. Results from 1 of 2 similar experiment are shown. *P < 0.05, **P < 0.01.

Figure 5. RAE-1γ remains intact during latent RAE-1γMCMV infection.

(A) μMT/μMT B cell–deficient mice latently infected with RAE-1γMCMV or WT MCMV were depleted of CD4⁺, CD8⁺, and NK cells by use of Abs. Viral titers were determined by plaque assay 13 days after immunodepletion. Titers of individual mice (circles) and median values (horizontal bars) are shown. Numbers indicate individual mice. (B) A total of 73 recurrent, plaque-purified viruses (termed RAE-1γMCMVr1 to RAE-1γMCMVr73) and recurrent WT MCMV (WTr MCMV) were isolated from organ homogenates of B cell–deficient μMT/μMT mice with recurrent RAE-1γMCMV infection. SVEC4-10 cells were infected with the indicated recurrent, plaque-purified RAE-1γMCMV viruses and analyzed for surface RAE-1γ expression by FACS as described in Figure 1B. (C) Untreated BALB/c mice or BALB/c mice treated with blocking anti-NKG2D Ab were i.v. injected with 10⁵ PFU WT MCMV or recurrent, plaque-purified RAE-1γMCMV (clone RAE-1γMCMVr5). Viral titers were determined in spleen 3 days p.i. by plaque assay. Titers of individual mice (circles) and median values (horizontal bars) from a representative of two independent experiments are shown. *P < 0.05, **P < 0.01.

Figure 6. RAE-1γMCMV is attenuated in immunocompromised host.

(A) IFN-α/βR^–/– mice were i.p. injected with 2 × 10⁵ PFU of RAE-1γMCMV, WT MCMV, or Δm152 MCMV, and survival rates were monitored daily. Combined results from 2 similar experiments are shown. (B) BALB/c mice were subjected to 6 Gy total-body γ-irradiation 6 hours prior to f.p. injection with 10⁵ PFU of RAE-1γMCMV or WT MCMV. Some groups of mice were depleted of NK cells by anti-asialoGM1 (αGM1) Ab. Viral titers were determined 7 days p.i. by plaque assay. Titers of individual mice (circles) and median values (horizontal bars) are shown. *P < 0.05.

Figure 7. Maternal immunization with RAE-1γMCMV protects offspring from MCMV disease.

(A) Female BALB/c mice were i.v. injected with 2 × 10⁵ PFU of RAE-1γMCMV or WT MCMV or mock injected 2 weeks before mating. Antiviral Ab titers in their serum and in serum of their neonates were determined by ELISA 6 hours postpartum. (B) A group of the neonates was i.p. injected with 500 PFU WT MCMV 6 hours postpartum, and viral titers were determined in various organs 9 days p.i. by plaque assay. Titers of individual mice (circles) and median values (horizontal bars) are shown.