120 CELL HEREDITY 



FIGURE 5.3. The lack of correlation be- 

 tween the numbers of reciprocal recombi- 

 nants produced by individual bacteria, all 

 of which were mixedly infected with two 

 types of phages, h* r* and hr. The h mutant 

 will infect a greater range of hosts than will 

 the h strain (from Hershey and Rotman, 

 10 20 30 1949, Gene//cs, 34:44). 



Per Cent /j+r 



parents when three types of phages infect the same cell. We can also 

 understand how, when one parent forms a very small minority, the fre- 

 quency of its genotype among the progeny may be smaller than the 

 frequency of recombinants between it and the majority parent. The 

 number of rounds of mating per cycle of multiplication in a bacterium 

 can be calculated, and varies from 0.5 to 5 according to the phage. 



But what is a "mating between molecules of DNA? Is there a cross- 

 ing o\er between the strands of duplicating DNA, or does the exchange 

 occur in some other way? In the classic sense crossing over is a behavior 

 of chromosomes, and it occurs in the four-strand stage, yielding recipro- 

 cal recombinants. When the progeny from large numbers of mixedly 

 infected bacteria are collected as one assemblage, equal frequencies of 

 reciprocal recombinants are found (Table 5.2). But when recombinants 

 are recovered from singly infected bacteria, equal numbers of reciprocal 

 recombinant types are not found (Figure 5.3). Sometimes, in fact, a cell 

 will yield only one or the other of the two types of reciprocal recombi- 

 nants. The equality observed when the products of many cells are 

 pooled is due to randomness. 



This lack of correlation can be explained by assuming that phage 

 maturation consists of taking a small sample from a large pool of DNA 

 particles for inclusion in protein coats. Thus, even though reciprocal 

 recombinants may be formed by the recombination event, random 

 sampling would lead to the formation of mature phages in which the 

 frequencies of the reciprocal recombinants are scattered around a 1:1 

 ratio. Sometimes the population of phages formed in a single cell may 

 even have included one but not the other reciprocal recombinant. But 

 if this mechanism were operating, there should also be a low correlation 

 between recombinants of the same type because the progeny of one re- 

 combinant will, on the average, be no more numerous than the progeny 

 of its reciprocal. The facts contradict this line of thinking. In single 

 bacterial bursts, if a recombinant of a particular type appears, there is a 



