HEMAGGLUTINATION BY ANIMAL VIRUSES 23 



smaller particle is produced in the course of infection with NDV (Burnet et 

 al., 1945; Granoff et al., 1950; Granoff and Henle, 1954). The latter authors 

 found that the small component was in higher concentration in the cells of 

 the chorioallantoic membrane than in the allantoic fluid. It eluted from red 

 cells more rapidly than standard virus and at lower temperatures. It had 

 none of the special "binding" capacities of normal Newcastle disease virus, 

 which are described later (Section I, C, 2, d). They considered that it might 

 represent a precursor of the virus. What may be a similar small particle 

 hemagglutinin can be obtained by disruption of at least some of the myxovirus 

 group by prolonged ether treatment (Hoyle, 1952; Schafer and Zillig, 1954). 

 In the case of fowl plague, the soluble hemagglutinin is a particle 10-15 nut 

 in diameter, composed of protein and carbohydrate but without nucleic acid 

 (Schafer and Zillig, 1954). As far as they have been studied, the properties of 

 the soluble hemagglutinin are similar to those of the virus particle hemag- 

 glutinin from which it was derived; further discussion in this chapter will be 

 limited to the latter. 



The identity of infective virus particle with hemagglutinin was established 

 by demonstrating the correlation of infectivity and hemagglutinin titers 

 under various conditions — in fresh preparations of complete virus (Hirst, 

 1942b; Liu and Henle, 1951), in the various stages of adsorption and elution 

 with red cells (Hirst, 1942b; Taylor et al., 1943), and on high-speed 

 centrifugation (Lauffer and Miller, 1944; Sharp et al., 1944; Taylor et al., 

 1943). 



With fully active virus, the amount of infective virus giving partial ag- 

 glutination in a pattern test can be determined accurately. Fazekas de St. 

 Groth and Cairns (1952) carefully defined titration methods for both hemag- 

 glutinin and infectivity, and gave a figure of about 10 6 ID 50 for one nonag- 

 glutinating unit. 



There is still controversy as to the number of particles, visible by electron 

 microscopy, in the units of infectivity and hemagglutinin. Donald and Isaacs 

 (1954) find that, in standard egg fluids with about 10 10 ID 50 per milliliter, 

 there are approximately 10 particles per ID 50 . Horsfall (1954), on the other 

 hand, holds that with virus protected from the possibility of thermal degra- 

 dation the infective unit corresponds to one visible, hemagglutinating 

 particle. Levine et al. (1953) have claimed that a single particle is capable of 

 holding two red cells in a dinier sufficiently stable to show a sedimentation 

 boundary in the presence of unaggregated cells. This should provide an 

 absolute method of measuring the number of virus particles in a given pre- 

 paration. Tyrrell and Valentine (1957), on the other hand, find that one 

 dimer occurs in the presence of ten particles of type A influenza, and Magill 

 (1951) believes that there is always a proportion of free virus particles in the 

 system, irrespective of elution through receptor destruction. 



