An Attenuated CMV Vaccine with a Deletion in Tegument Protein GP83 (pp65 Homolog) Protects against Placental Infection and Improves Pregnancy Outcome in a Guinea Pig Challenge Model

Abstract

Aims: Congenital human cytomegalovirus (HCMV) infection can lead to long-term neurodevelopmental sequelae, including mental retardation and sensorineural hearing loss. Preconception vaccine strategies relevant to prevention of HCMV-mediated injury to the newborn can be studied in the guinea pig cytomegalovirus (GPCMV) model. The objectives of this study were: 1) to assess in guinea pigs the protective efficacy against congenital infection and disease of a recombinant live, attenuated vaccine with a targeted deletion of the GPCMV homolog of the HCMV pUL83 tegument protein, GP83; and, 2) to compare the extent of placental infection in vaccine and control groups, using an in situ hybridization (ISH) assay.

Materials and methods: Outbred Hartley guinea pigs were vaccinated prior to pregnancy with a two-dose series of 5×10⁴ pfu of vAM409, a GP83 deletion virus. Deletion of the GP83 gene resulted in an attenuated virus, and vAM409 vaccinated animals did not demonstrate evidence of DNAemia following vaccination, although ELISA antibody responses were comparable to those observed in natural infection. After mating, pregnant animals were challenged with salivary gland-adapted (SG) GPCMV (1×10⁶ pfu) in the second trimester, and pregnancy outcomes were compared to controls.

Results: Compared to placebo-immunized controls, vaccination resulted in significantly reduced maternal DNAemia following SG challenge, and there was significantly decreased pup mortality in litters born to vaccinated dams (3/29; 10%), compared to control (35/50; 70%; p<0.001). By in situ hybridization study, recovered placentas in the vAM409 vaccine group demonstrated reduced infection and fewer infectious foci compared to the control group.

Conclusions: In summary, preconception immunization with a GP83 deletion vaccine reduced maternal DNAemia and results in protection against congenital GPCMV-associated pup mortality compared to unvaccinated controls. Vaccination resulted in reduced placental infection, probably related to the reduction in maternal DNAemia. Although the pp65 homolog in GPCMV, GP83, is a known target of protective T cell immune responses, it is nevertheless dispensable for effective vaccination against maternal and fetal CMV disease in this model.

Keywords: CMV UL83; CMV pp65; Cytomegalovirus; Cytomegalovirus vaccine; GPCMV GP83; Guinea pig challenge model; Guinea pig cytomegalovirus; Live; Vaccine efficacy; attenuated CMV vaccine; cytomegalovirus immune evasion.

Figure 1

Schematic overview and time course of vaccine study. GPCMV seronegative outbred Hartley guinea pigs were immunized twice (interval of 3 weeks between doses), with vAM409 vaccine, or placebo. Small-volume bleeds were performed by toenail clip were obtained (arrowheads) on all guinea pigs following vaccination at weeks 2, 3, 5, 7, 9 and 11, for both quantitative PCR analysis of viral load and for analysis of antibody response. Challenge with SG-passaged GPCMV was administered on days 89–91 of the experiment, approximately 7–8 weeks after initiation of breeding.

Figure 2

ELISA response following vAM409 vaccination. A) Antibody titers in response to vaccine (V) or placebo (P) were compared at multiple time points post-immunization. Antibodies titers (mean ± SE) were measured by ELISA using sonicated, clarified guinea pig lung fibroblast lysate from GPCMV-infected cells as previously described [13]. A control ELISA titer obtained using a polyclonal anti-GPCMV antisera from an animal inoculated with wild-type GPCMV is shown for comparison. B) Western blot demonstrating evolution of antibody response following vAM409 immunization. Immunized animals evolve antibody responses that include gB (GP58) as evidenced by comparison with a monospecific anti-gB polyclonal antibody (arrowhead).

Figure 2

ELISA response following vAM409 vaccination. A) Antibody titers in response to vaccine (V) or placebo (P) were compared at multiple time points post-immunization. Antibodies titers (mean ± SE) were measured by ELISA using sonicated, clarified guinea pig lung fibroblast lysate from GPCMV-infected cells as previously described [13]. A control ELISA titer obtained using a polyclonal anti-GPCMV antisera from an animal inoculated with wild-type GPCMV is shown for comparison. B) Western blot demonstrating evolution of antibody response following vAM409 immunization. Immunized animals evolve antibody responses that include gB (GP58) as evidenced by comparison with a monospecific anti-gB polyclonal antibody (arrowhead).

Figure 3

Reduced placental infection in vAM409-immunized dams compared to controls. (A) Distribution of in situ positive foci in vAM409-vaccinated (left panel) and control (right panel) animals. An average of 72.5 (range, 2–574) positive foci/cm² were noted in control placentas. In contrast, in placentas of vAM409-vaccinated animals, an average of 5.9 positive foci (range, 0.8–8.8)/cm² were noted (P < 0.05). (B) Demonstration of positive signal in control placenta. Left panel, 20× magnification of in situ labeled control dam demonstrating multiple positive foci (arrows) of GPCMV infection. Right panel, 80× magnification demonstrating positive cells (arrows) in trophoblasts of infected placenta.

Figure 3

Reduced placental infection in vAM409-immunized dams compared to controls. (A) Distribution of in situ positive foci in vAM409-vaccinated (left panel) and control (right panel) animals. An average of 72.5 (range, 2–574) positive foci/cm² were noted in control placentas. In contrast, in placentas of vAM409-vaccinated animals, an average of 5.9 positive foci (range, 0.8–8.8)/cm² were noted (P < 0.05). (B) Demonstration of positive signal in control placenta. Left panel, 20× magnification of in situ labeled control dam demonstrating multiple positive foci (arrows) of GPCMV infection. Right panel, 80× magnification demonstrating positive cells (arrows) in trophoblasts of infected placenta.