Potential sources of the 1995 Venezuelan equine encephalitis subtype IC epidemic

Abstract

Venezuelan equine encephalitis viruses (VEEV) belonging to subtype IC have caused three (1962-1964, 1992-1993 and 1995) major equine epizootics and epidemics. Previous sequence analyses of a portion of the envelope glycoprotein gene demonstrated a high degree of conservation among isolates from the 1962-1964 and the 1995 outbreaks, as well as a 1983 interepizootic mosquito isolate from Panaquire, Venezuela. However, unlike subtype IAB VEEV that were used to prepare inactivated vaccines that probably initiated several outbreaks, subtype IC viruses have not been used for vaccine production and their conservation cannot be explained in this way. To characterize further subtype IC VEEV conservation and to evaluate potential sources of the 1995 outbreak, we sequenced the complete genomes of three isolates from the 1962-1964 outbreak, the 1983 Panaquire interepizootic isolate, and two isolates from 1995. The sequence of the Panaquire isolate, and that of virus isolated from a mouse brain antigen prepared from subtype IC strain P676 and used in the same laboratory, suggested that the Panaquire isolate represents a laboratory contaminant. Some authentic epizootic IC strains isolated 32 years apart showed a greater degree of sequence identity than did isolates from the same (1962-1964 or 1995) outbreak. If these viruses were circulating and replicating between 1964 and 1995, their rate of sequence evolution was at least 10-fold lower than that estimated during outbreaks or that of closely related enzootic VEEV strains that circulate continuously. Current understanding of alphavirus evolution is inconsistent with this conservation. This subtype IC VEEV conservation, combined with phylogenetic relationships, suggests the possibility that the 1995 outbreak was initiated by a laboratory strain.

FIG. 1

Map of the regions of Venezuela and Colombia affected by the 1962–1964 and 1995 outbreaks. Sites of isolation of the VEEV subtype IC isolates that we studied are identified by boxed strain names. The following bold numbers show the states involved in both the 1962–1964 and 1995 outbreaks: 1, Guajira (Colombia); 2, Zulia; 3, Trujillo; 4, Falcon; 5, Lara; 6, Portuguesa; 7, Yaracuy; 8, Carabobo; 9, Cojedes; and 10, Guarico. The following numbers show states involved only in the 1962–1964 outbreak: 11, Aragua; 12, Miranda; 13, Anzoategui; 14, Sucre; 15, Monagas; and 16, Delta Amacuro.

FIG. 2

Maximum parsimony phylogenetic analysis of VEEV subtype IC isolates from the 1962–1964 and 1995 outbreaks. An outgroup comprised of subtype IE strains MenaII (14) and 68U201 (18) was used to root the tree, and additional subtype I viruses described previously (–15, 33) were included in the analysis. All virus strains are denoted by VEEV subtype, followed by country abbreviation and year of isolation, followed by strain name. Numbers at the nodes represent bootstrap confidence indices based on 1,000 replicates. The bar represents the number of nucleotide changes in each horizontal branch.

FIG. 3

Regression analyses of sequence evolution rates of enzootic subtype ID (dashed line) and epizootic subtype IC (solid line) VEEV. Using maximum parsimony analyses, nucleotide changes from a node representing the common ancestor of a clade were plotted versus the year of isolation. The linear regression represents the rate of nucleotide substitution over time.