INACTIVATION OF VIRUSES 393 



after coverage of a major part of the amino groups. Attempts to remove the 

 amino nitrogen with nitrite led, however, to complete loss of activity, prob- 

 ably on account of micontrolled oxidation. 



Independently, Miller and Stanley (1941a,b, 1942) attacked the same pro- 

 blems. These authors found that about 70 % of the amino groups and 20 % 

 of the tjTOsine -j- tryptophan groups could be covered without loss of infec- 

 tivity. If the reaction was carried beyond these limits a gradual inactivation 

 occurred. Analyzed in the ultracentrifuge, the acetyl and phenylureido deri- 

 vatives appeared homogeneous; electrophoretically they were likewise homo- 

 geneous, with a mobility quite distmct from that of the native virus. As 

 further derivatives, carbobenzoxy, ^-clilorobenzoyl, and benzenesulfonyl 

 viruses were prepared. Again, about 70 % of the amino groups and 20 % of 

 the phenol + indole groups could be covered without loss of infectivity. 

 Finally, Anson and Stanley (1941) found that oxidation of suKhydryl groups 

 with iodine could be carried out without appreciable reduction of the infec- 

 tivity. The progeny of all of these derivatives were invariably fomid to cause 

 the normal disease and to have the chemical composition of normal virus. 

 Thus, none of the in vitro reactions produced any "mutants." 



Most of the substitutions now described undoubtedly involve only surface 

 groups; they cannot be considered to represent major chemical alterations 

 and they may be more or less readily reversed after mtroduction of the virus 

 into the cell, although Miller and Stanley were unable to demonstrate rever- 

 sion after treatment with homogenates of fresh tobacco leaves. The signifi- 

 cance of these early observations is, therefore, limited. 



Later, however, more drastic changes of the protein with no or little 

 effect upon the infectivity have been described. Harris and Knight (1952) 

 found that treatment with carboxypeptidase removed about 7 % of the 

 total threonine from the TMV protein without apparently affecting the 

 infectivity. 



Schramm (1947a, b), in confirmation of several earlier reports, had ob- 

 served that TMV in a moderately alkaline medium was gradually broken 

 down into smaller fragments. A systematic study of this phenomenon 

 (Schramm et al. 1955) revealed the appearance of several fractions distin- 

 guishable by their sedimentation and electrophoretic properties. Of particu- 

 lar interest in this connection are those fractions designated as I and II, 

 which had a greater electrophoretic mobihty and higher nucleic acid contents 

 than the native virus. In electron micrographs this material appeared as 

 mutilated virus rods, more or less deficient in protein, and with the nucleic 

 acid showing as a centra] strand. After careful preparation these fractions 

 retained practically full infectivity. 



Similar results were reported by Hart (1955a,b), who treated the virus 

 with sodium dodecyl sulfate (SDS) at pH 8 and a brief exi30sure to 



