114 BIOCHEMISTRY: BENZER AND FREESE Proc. N. A. S. 



of the mixture were rapidly distributed — one drop into each of 200 tubes. After 

 incubation at 37° C. (for 30 minutes) to allow the infected cells to burst, the content 

 of each tube was plated on B. As is typical in this sulfanilamide medium, the 

 average yield per cell was very small, of the order of one viable progeny particle per 

 infected cell. Each plate contained (after incubation) from 30 to 100 plaques, in 

 most cases including one or more r-type plaques. The over-all proportion of r-type 

 particles among the progeny was about 2 per cent. To assure the independent 

 origin of each mutant, no more than one r plaque was picked from a plate. Each 

 such mutant was purified by replating, and a stock prepared on S in broth. 



Genetic Mapping of the Mutow^s.— Different r mutants of T4, although producing 

 similar plaques on B, fall into groups distinguished by their behavior on a second 

 host, K. Those mutants with which we are here concerned, of the rll group, do not 

 produce plaques on K. This property is the key to the high resolution with which 

 they can be mapped genetically. When two rll mutants are crossed, the appearance 

 of any standard-type recombinants among the progeny is readily detected by plating 

 on K. If standard-type progeny are produced in a cross (above the background 

 rate due to spontaneous reversion of the mutants), it is concluded that the two 

 mutants contain alterations at different locations in their genetic structures. 



Our objective is to compare these locations for spontaneously arising and for 5- 

 bromouracil-induced mutants. The task of crossing a large number of mutants, 

 two by two, to see which pairs yield recombinants is enormous. However, this 

 process can be shortened by making use of a set of mutants having large alterations, 

 as shown in Figure 1. Each new isolated rll mutant is crossed with each mutant of 

 this set (by means of simple spot tests). By noting with which mutants of the set 

 it does or does not produce standard-type recombinants, the mutation can be as- 

 signed to a particular segment of the map. Thereafter, only mutants belonging to 

 the same segment need be crossed with each other. Thus the number of crosses 

 reciuired for analyzing a batch of mutants is greatly reduced. 



The genetic procedure has therefore been to (1) isolate many independently 

 arising r-type mutants; (2) choose those of the rll group; (3) test each rll mutant 

 against the mutants of Figure 1, thereby locating its mutational alteration in a 

 particular segment of the map; and (4) cross the mutants belonging to the same 

 segment with each other to determine which mutations share common locations. 

 For the present purposes, no attempt was made to determine the order of these loca- 

 tions within a map segment. 



Reversion Rates of Mutants.— ^The different rll are characterized not only by the 

 positions of their mutational alterations in the map but also by differences in their 

 fre(}uency of reversion to particles that resemble the standard type in their plating 

 behavior on K and B. These revertants arise spontaneously during the growth of a 

 given r-type phage, and their typical frequency, in a stock grown up from an 

 inoculum of 100 r phages, is called the "reversion index." 



RESULTS 



Over-all Mutation Frequencies.— The ratio of the induced to the spontaneous rate 

 of mutation cannot be given accurately, since different procedures of isolation were 

 used. A rough estimate may be made as follows: In a broth lysate of T4B grown 

 (on S) from an inoculum of a few particles up to a population of lO^", the proportion 



222 



