Controlled viral glycoprotein expression as a safety feature in a bivalent rabies-ebola vaccine

Abstract

Using a recombinant rabies (RABV) vaccine platform, we have developed several safe and effective vaccines. Most recently, we have developed a RABV-based ebolavirus (EBOV) vaccine that is efficacious in nonhuman primates. One safety feature of this vaccine is the utilization of a live but replication-deficient RABV construct. In this construct, the RABV glycoprotein (G) has been deleted from the genome, requiring G trans complementation in order for new infectious viruses to be released from the initial infected cell. Here we analyze this safety feature of the bivalent RABV-based EBOV vaccine comprised of the G-deleted RABV backbone expressing EBOV glycoprotein (GP). We found that, while the level of RABV genome in infected cells is equivalent regardless of G supplementation, the production of infectious virus is indeed restricted by the lack of G, and most importantly, that the presence of EBOV GP does not substitute for G. These findings further support the safety profile of this replication-deficient RABV-EBOV bivalent vaccine.

Keywords: Biodefense; Ebola; Filovirus; RNA viruses; Rabies virus; Vaccine.

Figure 1

Quantitative PCR to detect RABV nucleoprotein over time in a multi-step growth curve (moi 0.01). Data points are log mean genomic equivalents per microgram of total RNA (n=2) with SEM error bars. RVΔGGP, □;RVGP, ●. Viral genome replication was not statistically different on A) BSR-G and B) VeroE6 cells for either virus (n=2; Student’s t test, p=0.35 and 0.24 respectively).

Figure 2

Quantitative PCR to detect RABV nucleoprotein upon serial passage of RVΔGGP on VeroE6 cells. Viral supernatant was passaged at 48h post-infection. Bars are log mean genomic equivalents per microgram of total RNA (n=2) with SEM error bars. Positive control is RNA from the RVΔGGP stock used to infect for passage 1. Negative control is an irrelevant RNA.

Figure 3

Confocal microscopy at 126X magnification using fluorescently labeled antibodies shows that RVG is not synthesized in RVΔGGP-infected VeroE6 cells. BSR-G and VeroE6 cells were infected with RVGP or RVΔGGP at an MOI of 0.01, and then fix and stained at 48hpi according to staining method in Table 1. Antibody set N/G is for RABV N and RABV G detection; N is red and G is green. Antibody set N/GP is for RABV N and EBOV GP detection; N is red and GP is green.

Figure 4

Electron microscopy of VeroE6 and BSR-G cells infected with RVΔGGP and RVGP followed by immunogold labeling for RABV G and EBOV GP at 96hrs post infection. Cells were infected and prepared as described in the text. Gold-tagged antibodies confirm glycoprotein expression on virion surface. A) RVGP in BSR-G cells; B) RVΔGGP in BSR-G cells; C) RVGP in VeroE6 cells; D) RVΔGGP in VeroE6 cells (67900X). 96hpi @ 108000X unless noted. RVG labeled with 5nm particles. GP labeled with 15nM particles. Immunogold labeling of D was not performed due to lack of extracellular virions at sufficient quantities to justify staining.