The Hemagglutinin of West Nile Virus: Recovery, Properties, and Antigenic Relationships

Abstract

Summary A specific hemagglutinin for chick erythrocytes was demonstrated in both unbuffered saline and pH 9 buffered saline extracts of the brain of 3 to 4 week old mice infected with the original (Uganda) strain of West Nile fever virus. In the unbuffered saline extracts (pH 7.1) much of the hemagglutinin was in large aggregates which sedimented at 13,000 rpm, while no such aggregates were present in the alkaline extracts (pH 8.1). Freshly prepared, 13,000 rpm supernates of the alkaline extracts produced hemagglutination between pH 6.7 and 7.5 or 8.0, but maximum titers (as high as 20,480) without zones of absent or poor agglutination in the lower dilutions were obtained at pH 7.1–7.3. After such preparations were stored at 4 C for 4 to 50 days the pH range shifted to the acid side with maximum titers at pH 6.7–6.9. The hemagglutinin in alkaline extracts was not destroyed by repeated freezing and thawing and could be maintained without loss of potency at -70 C. To obtain the maximum titers it was necessary to mix the hemagglutinin with the red cells at 20 to 25 C and to permit sedimentation to occur at 4 C. With certain unbuffered saline extracts no reaction at all occurred when the reagents were at 4 C at the time of mixing, or if sedimentation occurred at 25 C. The reaction between the hemagglutinin and red cells is of the equilibrium type. Hemagglutination can be reversed by repeated resuspensions of the cells, because the uncombined hemagglutinin is unstable at the pH that is optimum for reaction. No receptor-destroying enzyme is involved because the addition of fresh hemagglutinin causes reagglutination of such cells. When 3 other strains of West Nile virus (Egypt 19, 21, and 101) were tested no zone phenomena were encountered and fresh preparations yielded maximum titers at pH 6.7–6.9. Sedimentation at 4 C was needed to obtain maximum titers with strains 19 and 21 but not with 101. All 4 strains yielded identical hemagglutination-inhibition titers with specific antisera. The normal inhibitor present in human and animal sera was best eliminated by precipitating the specific antibody with acetone. Cancer patients experimentally infected with West Nile virus developed high titers of hemagglutination-inhibition antibody for the West Nile hemagglutinins and somewhat lower titers for the St. Louis and Japanese B virus hemagglutinins. Human beings naturally infected with the viruses of St. Louis or Japanese B encephalitis also developed hemagglutination-inhibition antibodies for all 3 viruses but the homologous titers were invariably higher. Hyperimmune rabbit sera for the 3 viruses showed much more extensive crossing by the hemagglutination-inhibition than by the neutralization tests. Hemagglutination-inhibition tests with West Nile, St. Louis and Japanese B hemagglutinins on “normal” Egyptian human sera gave positive results indicative of unrecognized infection only with West Nile virus.