During surveillance of the swine population in Israel a haemagglutinating agent was isolated and identified as a paramyxovirus (PMV). Serological studies based on haemagglutination and neuraminidase inhibition (HI and NI) tests displayed a close antigenic relationship between the isolate and the prototype strain of a variety of the avian PMV serotype 3 (PMV-3). However, comprehensive HI and NI cross-reaction tests between the isolate and PMV-3 reference strains, on the one hand, and the other

more »

... ght avian PMV serotypes, on the other hand, showed that the viruses compared differed in terms of their antigenic interrelationships with the other avian PMVs and also in the quantitative values of the cross-reactivity. This is interpreted to mean that the swine isolate and the avian PMV-3 prototype strain are different viruses, although very closely related, but are not different isolations of the same virus from avian and mammalian hosts. 0000-6785 © 1986 SGM 428 M. LIPKIND, D. SHOHAM AND E. SHIHMANTER METHODS Virus isolation. Nasal swabs and blood samples were taken from apparently healthy pigs just after slaughter. The swabs were put into phosphate-buffered saline (PBS) glycerol medium with antibiotics (Hinshaw et al., 1978) . After transfer to the laboratory the swabs were either immediately, or after a week-long storage at -70 °C, inoculated into 10-day-old chick embryonated eggs which were incubated at 35 °C until embryo death or until 6 days post-inoculation. Treatment of the swabs, examination of the allantoic fluids of the infected eggs and passaging of the isolated HA agents before their detailed identification have been described elsewhere (Lipkind et al., 1980) . Viruses. The following avian PMV reference strains were used in the studies (designated according to the nomenclature suggested by Tumova et al., 1979b5:PMV-Goat monospecific antisera against all the haemagglutinin (HA) and neuraminidase (NA) antigenic subtypes of influenza A virus were kindly supplied by Dr R. G. Webster (St. Jude Children's Research Hospital, Memphis, Tenn., U.S.A.). Anti-NDV serum was from turkeys which had survived Newcastle disease in a local farm and were then boosted by NDV. Sera against the other viruses were either kindly supplied by Dr D. J. Alexander or prepared in chickens by intravenous inoculation of allantoic virus (Palmer et al., 1975) . Antiserum against the isolate was prepared by intramuscular inoculation of rabbits with concentrated purified virus prepared by differential centrifugation followed by two consecutive centrifugations through 30 to 709o sucrose gradients. Before use the antisera were treated with receptor-destroying enzyme (Palmer et al., 19755. Haemagglutination inhibition (HI) test (Hirst, 1942) . This was carried out in a micro-test modification using Uwell plastic microplates and red blood cells (RBC) in 0.5 °/o suspension (Palmer et al., 1975) . After 1 h incubation of the virus and antiserum mixture at 20 °C, the RBC suspension was added and the plates were incubated at 4 °C for 30 min before reading. A fractional dilution method was employed, using several initial dilutions of the material to be titrated before further twofold dilution. The titration endpoint (100% hacmagglutination) patterns in all the dilution series were used for calculation of the HI titre (Lipkind et al., 19735. This method appeared to be much more precise than that conventionally used. Therefore, in some cases a system of crosses was used for estimation of the partial haemagglutination pattern for calculation of HA and HI titres (Horsfall & Tamm, 1953) . Neuraminidase inhibition (NI) test. This was performed using a thiobarbituric acid assay (Warren, 1959) with the modifications of Aminoff (1961) which has been applied to influenza virus diagnosis (Palmer et al., 1975) . Fetuin (Sigma) was used as a substrate. Statistical treatment of HI and NI values. It was found that although there were considerable fluctuations in the values of both HI and NI titres in various experiments, the fluctuations in the corresponding values for crossreactivity (differences between the values of homologous and heterologous HI and NI titres) were substantially less. Therefore, the value of the cross-reactivity, designated as a difference d between homologous and heterologous titres expressed in logz, was used for statistical treatment based on the t-test (Spiegel, 1961) . The Student's probability value P for the differences d for a given number of experiments n was calculated. The reliability of the d values (their difference from zero) was accepted as significant if it was in agreement with Fisher's second criterion of probability (95 ~o probability, P = 0.95). When there was no cross-reactivity between certain viruses d was expressed as infinity (o05. Degrees of relatedness are referred to as identity (d 0 to 1 log2), high (d 1 to 2 logz), moderate (d 2 to 4 log2) and low (d 4 to 7 log2). The same principles of statistical treatment (Spiegel, 196l) were applied to two-sided asymmetric crossreactivity (cases displaying one-sided asymmetry evidently did not need any statistical treatment). In this case the cross-reactivity between two viruses X and Y was considered to be asymmetric if the difference d r (when anti-X serum was used against both viruses) was statistically different from d2 (when anti-Y serum was used against both viruses). Electron microscopy. Investigations were carried out using negative contrast with phosphotungstic acid staining.