500 A. TSUCMTA AND II. I- KAKX KF.!.-('< )\HAT 



Tilt' iiu'tlit)(l> ciiiplox'cil l)y us wcic hiiclly as follows. The IvXA was 

 sul)jt.'C'tc'(l to one of a \ai-i('t\- of cliciiiical agents, then freed of tlie 

 reagent, and i-cconstiluled with untrt'atcd pi-otein. Only rarely was the 

 conipU'tt' \-iiiis subjected to the cheinical treatment. As pre\-i()usly 

 stated, only methylation and hroniination yielded significant numbers of 

 nuitants while dcamination mutated almost all survivors. The extent of 

 chemical modification was gaged from the inactivation, 37, 14, 5, and 

 2% residual infectivity being regarded as indicating 1, 2, 3, and 4 

 inactivating hits per average molecule (e~^ = 0.37). Previous studies 

 had demonstrated that each chemical modification, as ascertained ana- 

 lytically, had at least a 50% chance of being an inactivating event. 



Two techniques were employed for the detection of mutants. Dif- 

 ferential host mutants, as defined by Siegel (I960), are those which 

 are detected as local lesions on .V. sijlvestns, a systemic host for the 

 wild type virus. Such local lesions were individually excised and re- 

 peatedly transferred to the same host until they produced local lesions 

 unaccompanied by systemic symptoms and thus appeared to have been 

 freed from contaminating wild type virus particles. Symptom mutants 

 are detected by inoculating the reconstituted RNA onto a local lesion 

 host, Xanthi, nc, and then transferring homogenates of randomly selected 

 (or all) individual lesions to N. tabacum, N. sylvestris, and again to 

 Xanthi. IMutants are distinguished by the appearance of unusual num- 

 bers or shapes of lesions on Xanthi, nc, unusually mild or severe symp- 

 toms on the systemic hosts, or other odd symptoms, singly or combined. 

 To verify the genetically stable nature of any such symptoms, the virus 

 was usually passed at least once through a local lesion and again onto 

 that host which showed a distinctive reaction. Stable mutants were then 

 transferred to 24-72 plants of .V. tabacum and virus harvested from these 

 after 21 days. The symptomatology of th(> nuitant progeny was again 

 verified by plant tests. 



The mutant virus was split with acetic acid, and its protein was 

 isolated. Samples of 4—5 mg of protein were hydrolyzed with 6 N HCl 

 at 108=!=1°C for 24 and 72 hours. It is important that the acid be 

 evaporated in a few minutes to avoid losses in serine and glutamic acid. 

 One milliliter of the residue, redissolved to about 1 mg 'ml, was ai)plied 

 to the amino acid analyzer (Tsugita and Fraenkel-Conrat, 1962). 



The f|uestion arises whether direct analysis of hydrolyzates of a 

 protein of molecular w'eight 18,000 can yield amino acid composition 

 data sufficiently accurate to detect a single amino acid change. Since the 

 most frequently occurring amino acid in TAIV is aspartic acid with 18 

 residues per mole, an analytical error of ±2.59^ would repi-esent the 

 limit for the accurate detei-mination of this integral number. An accuracv 



