290 F. M. BURNET 



1. Transfer of N euwpathogenicity 



What can now almost be called the classic example is neuropathogenicity 

 in influenza viruses. By using one or other of the available neuro-strains of 

 WS (NWS— Stuart-Harris, 1939; WSN— Francis and Moore, 1940), it is 

 possible to transfer neuropathogenicity to the influenza strains MEL (Burnet 

 and Lind, 1951b; Burnet and Edney, 1951; Hirst and Gotlieb, 1953; Edney 

 and Lim, 1954), Swine 15, WSM (Burnet and Lind, 1949), Oc. I. (Burnet and 

 Lind, 1951b), KUNZ (Appleby, 1952), WSE (Lind and Burnet, 1954). By 

 appropriate choice of recombinant strains, examples of all of these have been 

 obtained with pathogenicity on intracerebral inoculation in mice approaching 

 that of the original neuro-WS strain. 



When, however, the whole yield of recombinants is considered, all workers 

 find a high proportion of strains with low or equivocal neuropathogenicity. 



Hirst and Gotlieb (1955) for instance, using W -f- and M — parents, found 

 it easy to obtain W — recombinants but only 2/58 strains serologically M had 

 potential neuropathogenicity. These were detected only by back-crossing with 

 W — and obtaining neuropathogenic W -f- recombinants. Subsequent use of 

 M — inactivated by ultraviolet light gave M + recombinants with full 

 neuropathogenicity. 



Fraser (1955) inoculated mixtures into the brains of infant mice and 

 obtained a high yield both of neuro-MEL strains and the reciprocal non-neuro- 

 pathogenic WS — . Like other workers he found that the neuro-MEL strains 

 showed a wide range of expression in virulence by intracerebral 

 inoculation. 



In Melbourne, all our studies have shown a wide range of expression of 

 neuropathogenicity in recombinants from NWS (Burnet and Edney, 1951). 

 Whenever neuropathogenicity was evident in derivative strains serologically 

 MEL, the two other characters of conversion to indicator and pathogenicity 

 for chick embryos, later called c and e, were present. If, however, one took 

 all strains with the complex of characters A-ce and tested their pathogenicity 

 by intracerebral injection in mice, a complete range from no pathogenicity 

 to a virulence corresponding to NWS was observed. 



When recombination took place in the allantoic cavity between NWS and 

 MEL, it was easy to obtain MEL strains A-ce but none of these was virulent 

 for mice of standard size. Later work along somewhat different lines makes it 

 likely that these strains would have been pathogenic for infant mice. Fully 

 virulent neuro-MEL strains were only obtained when either the mouse brain 

 or the chorioallantois was used as the route of inoculation. Fraser (1958) has 

 succeeded in producing neuropathogenic recombinants in the allantoic 

 cavity by repeated crossing of progeny showing A-c character on isolation, 

 with the NWS parent. Ultimately, strains of overt but not maximal neuro- 

 pathogenicity were obtained. 



