Thomas W. Geisbert, Peter Pushko, Kevin Anderson, Jonathan Smith, Kelly J. Davis, and Peter B. Jahrling

Abstract

Ebola virus (EBOV) causes acute hemorrhagic fever that is fatal in up to 90% of cases in both humans and nonhuman primates. No vaccines or treatments are available for human use. We evaluated the effects in nonhuman primates of vaccine strategies that had protected mice or guinea pigs from lethal EBOV infection. The following immunogens were used: RNA replicon particles derived from an attenuated strain of Venezuelan equine encephalitis virus (VEEV) expressing EBOV glycoprotein and nucleoprotein; recombinant Vaccinia virus expressing EBOV glycoprotein; liposomes containing lipid A and inactivated EBOV; and a concentrated, inactivated whole-virion preparation. None of these strategies successfully protected nonhuman primates from robust challenge with EBOV. The disease observed in primates differed from that in rodents, suggesting that rodent models of EBOV may not predict the efficacy of candidate vaccines in primates and that protection of primates may require different mechanisms.

Introduction

Ebola virus (EBOV) and Marburg virus (MBGV), which make up the family Filoviridae, cause severe hemorrhagic disease in humans and nonhuman primates, killing up to 90% of those infected. EBOV was first recognized in the former Zaire in 1976. Subsequently, outbreaks have been documented in Sudan, Gabon, the former Zaire, Côte d'Ivoire, and Uganda^[1-3]. In addition to the African outbreaks, the species Reston Ebola virus, which may be less pathogenic for humans, was isolated from cynomolgus monkeys imported from the Philippines to the United States^[4]. Although outbreaks of EBOV have been self-limiting, the lack of an effective vaccine or therapy has raised public health concerns about these emerging pathogens.

In early attempts to develop a vaccine against EBOV, guinea pigs or nonhuman primates were vaccinated with formalin-fixed or heat-inactivated virion preparations. Results from these studies were inconsistent: Lupton et al.^[5] partially protected guinea pigs against EBOV, while Mikhailov et al.^[6] achieved complete protection of four of five hamadryad baboons by vaccinating them with an inactivated EBOV vaccine. However, other studies suggested that inactivated EBOV did not induce sufficient immunity to reliably protect hamadryl baboons against a lethal challenge^[7]. Conventional strategies of attenuating viruses for use as human vaccines have not been pursued for EBOV because of concerns about reversion to a wild-type form. However, the possibility of following this strategy by using newly developed infectious clones of EBOV may now be feasible^[8].

Recent efforts have focused on the use of recombinant DNA techniques to stimulate cytotoxic T-lymphocyte responses. Vaccinating guinea pigs with plasmids against EBOV nucleoprotein (NP), soluble glycoprotein, or glycoprotein (GP) elicited humoral and cellular immune responses against these gene products but only partially protected them against lethal challenge^[9]. However, results of this study were difficult to interpret because all the guinea pigs were killed 10 days after EBOV challenge, which is within the expected survival time for untreated animals (8-14 days)^[10]. In 2000, Sullivan et al.^[11] reported protection of cynomolgus monkeys from EBOV infection by injecting them with naked-DNA GP, followed by an adenovirus-expressing GP booster. Results of this study document the feasibility of vaccination against EBOV. However, these results require confirmation and further evaluation, as a low dose (6 PFU) was used for the challenge. Other studies reported a protective effect of EBOV vaccination with a low infective challenge dose (10 50% lethal doses [LD₅₀])^[7]; however, all vaccinated animals in these dosing studies died after receiving higher infective doses (100 and 1,000 LD₅₀), which may more accurately mimic natural or nosocomial exposures.

Our efforts to develop a vaccine against EBOV focused on several potential vaccine candidates. First, we used Venezuelan equine encephalitis virus (VEEV) replicon particles (VRP) expressing EBOV genes known to protect guinea pigs and mice from EBOV disease^[10]; VRP expressing MBGV genes also protected guinea pigs and cynomolgus monkeys against MBGV^[12]. Second, we used a recombinant Vaccinia virus (VACV) system expressing EBOV GP and demonstrated that this vector protected guinea pigs from EBOV hemorrhagic fever^[13]. A third strategy used encapsulated, gamma-irradiated EBOV particles in liposomes containing lipid A^[14]; and the fourth approach evaluated vaccination with a concentrated, gamma-irradiated whole-virion preparation. None of these approaches, which successfully protected rodents from lethal infection, were protective for cynomolgus or rhesus macaques challenged with EBOV.