49 Am. J. Trop. Med. Hyg., 64(1, 2), 2001, pp. 49–55 Copyright  2001 by The American Society of Tropical Medicine and Hygiene COMPARATIVE NEUROVIRULENCE OF ATTENUATED AND NON-ATTENUATED STRAINS OF VENEZUELAN EQUINE ENCEPHALITIS VIRUS IN MICE GEORGE V. LUDWIG, MICHAEL J. TURELL, PETER VOGEL, JOHN P. KONDIG, WAYNE K. KELL, JONATHAN F. SMITH, AND WILLIAM D. PRATT Diagnostic Systems Division, Virology Division, and Pathology Division, United States Army Medical Research Institute of Infectious Diseases, Fort Detrick, Frederick, Maryland Abstract. A candidate live-attenuated virus vaccine for protection against Venezuelan equine encephalitis (VEE) (designated V3526) was tested in mice to measure the magnitude, duration, and kinetics of virus replication in the blood and the central nervous system and its phenotypic stability after multiple passages in mice and cell culture. All results were compared to parallel experiments with parental virus and the existing VEE virus vaccine, TC-83. Maximum virus titers in the brains of V3526-inoculated mice were between 10- and 100-fold less than those observed in brains of mice inoculated intracranially (ic) with either the parental virus or TC-83. Neither V3526 nor TC-83 was lethal in BALB/c mice inoculated ic. However, mice inoculated with TC-83 developed acute symptoms lasting at least 14 days. In contrast, ic inoculation of TC-83 was uniformly lethal for C3H/HeN mice. V3526 was avirulent in both BALB/c and C3H/HeN mice after ic inoculation. The virulence characteristics of V3526 remained unchanged after five serial ic passages in mouse brains or after five cell culture passages. Finally, pathologic changes induced after ic inoculation of V3526 were consistently less severe and of shorter duration than those observed in TC-83- inoculated mice. Based on these results, V3526 is stable and appears to be significantly less neurovirulent in mice than TC-83. INTRODUCTION The Venezuelan equine encephalitis (VEE) virus complex is composed of serologically related mosquito-borne viruses in the genus Alphavirus, family Togaviridae. Viruses from the VEE-complex have been responsible for extensive epi- demics in North, Central, and South America and are highly pathogenic for both humans and members of the family Eq- uidae. 1 VEE complex viruses are grouped, based on their epidemiology and ecology, into epizootic and enzootic va- rieties. All of the epizootic strains are exotic to the United States 2 and, until 1995, had not surfaced during natural ep- idemics since 1973. Between April and October of 1995, an outbreak of fatal encephalitis caused by epizootic VEE virus occurred in Venezuela and Colombia. The outbreak involved over 75,000 human cases and resulted in more than 20 deaths. 3,4 Like the other alphaviruses, VEE virus is an enveloped virus, consisting of three structural proteins: capsid (C), which encapsidates the viral genome, and two envelope gly- coproteins (E1 and E2). The glycoproteins project from the virus envelope as spikes, each formed from three copies of an E1/E2 heterodimer. These structural proteins are ex- pressed as one large polyprotein from a 26S, subgenomic RNA and the order of translation is C, PE2, 6K, and E1. PE2 is a direct precursor of E2 and includes at its N-terminus 59 amino acid residues (E3) which are removed late in mat- uration by a Golgi-resident, furin-like enzyme. 5 The U.S. Army developed a live-attenuated vaccine (TC- 83) for use in humans in the early 1960s. 6 Subsequent to its development, the vaccine was also approved for, and distrib- uted for use in horses. The vaccine was developed by clas- sical cell-culture passage techniques, and is currently avail- able for use in humans only under an investigational new drug (IND) protocol. Because the vaccine is not licensed, its use is limited to at-risk laboratory workers, field biologists, and certain other groups of individuals deemed to be at risk for infection. While TC-83 is normally quite effective, it also has sev- eral important disadvantages that ultimately prevent it from being licensed for general use. 7 First, approximately 20% of vaccinees will develop some degree of illness after vacci- nation. 8 Five percent of these individuals will develop a rel- atively-severe febrile disease that resembles naturally ac- quired VEE. In addition to its high reactogenicity rate, TC- 83 also has a high vaccine failure rate, as another 20% of vaccine recipients fail to mount any detectable immune re- sponse. 8 Rodent-virulent virus can be isolated from throat swabs of some vaccine recipients, indicating that reversion, or at least selection of virulent virus subpopulations occurs after vaccination (Jahrling PB, unpublished data). The vac- cine also causes fetal infection and wastage in rodents. 9 Ad- ditionally, viremias in horses, mules, and donkeys may be sufficient to infect mosquitoes that could potentially transmit virus to other susceptible hosts. 10 Finally, there is some ev- idence from accidental laboratory exposures that the vaccine does not adequately protect against infection from heterol- ogous VEE virus subtypes (Jahrling PB, unpublished data). 11 A new-generation, live-attenuated VEE vaccine, currently under development, is being prepared to overcome the prob- lems and limitations associated with TC-83. The new vac- cine candidate, V3526 was derived from a full-length cDNA clone of the Trinidad donkey strain (TrD) of VEE virus (V3000) modified to contain two independently-attenuating mutations at defined loci. 12 These mutations include a dele- tion of the four amino acid furin recognition site (RKRR) between structural proteins E3 and E2 and a Phe to Ser change at E1 amino acid residue 253. In vitro transcription of V3526 cDNA results in RNA that is fully infectious for cultured cells and produces an attenuated virus possessing PE2 in the mature virion. Rodent and primate testing of this vaccine indicates that it is safe and effective at protecting animals from a wild-type virus challenge by both parenteral and aerosol routes (Pratt WD, unpublished data). 13 This paper presents a comparative analysis of the relative neurovirulence of TC-83, V3526, and wild-type TrD virus