Viruses: Recombination in Bacteriophage (I) 



387 



procedure being the equivalent of studying 

 the resuhs of "crossing" genetically different 

 virulent phages), it is possible to construct a 

 genetic map of phage in which the mutant 

 loci are arranged in linear order. 



Plaque mutant r is rapidly lysing and pro- 

 duces a larger plaque with sharper margins 

 than is produced by the wild-type allele /•+. 

 Mixed infections with r and /-+ phages usually 

 yield progeny phages that produce plaques of 

 one or the other type. However, two per cent 

 of the plaques observed are mottled, that is, 

 are plaques which appear partly r and partly 

 r+. When the phages in mottled plaques are 

 tested, both (and only) r and r+ types are 

 found. Accordingly, such mottled plaques 

 cannot be explained as being derived from 

 single haploid recombinant progeny. These 

 mottled plaques can be shown not to be due 

 to some action by clumps of phage particles. 

 Finally, such plaques cannot be explained as 

 being due to mutations in unstable r mutants, 

 for such unstable r phages are known to pro- 

 duce sectored, rather than mottled, plaques, 

 whose phage contents when tested either 

 yield sectored plaques again or r plaques, but 

 none of r+ type. From these results and 

 others, it is proved ^ that the two per cent of 

 phage that produce mottled plaques are 

 heterozygotes for a short region including the 

 r locus. It is also found that such hetero- 

 zygotes are recombinant for genetic markers 

 on opposite sides of the short heterozygotic 

 region.^ This suggests that, in phage, the 

 processes leading to heterozygosis are the 

 same as those leading to recombination. 



One of the ways that heterozygosis and 

 genetic recombination in phage can be visual- 

 ized to occur is as follows. In a mixed infec- 

 tion, the DNA's of different vegetative phage 

 particles repeatedly pair and unpair until the 

 host lyses. It is during these "matings" that 

 replication of new vegetative phage occurs. 

 Suppose replication proceeds by a copy- 



7 A. D. Hershey and M. Chase (1951). 

 ^ As indicated by C. Levinthal. 



choice mechanism, and the two mating 

 strands are genetically different at two closely 

 linked points. If new phage DNA was 

 formed by using the DNA of only one vege- 

 tative phage as template, the progeny phage 

 receiving this would, of course, be non- 

 recombinant. If, however, as diagrammed in 

 Figure 42-4, one vegetative phage is used as 

 a template to make a partial replica that 

 extends through one marked locus («+) and 

 the other mate is used as a template to make 

 a partial replica that extends through the 

 other marked locus (/)+), synthesis may con- 

 tinue so that both partial replicas contain the 

 short segment between the markers. A 

 progeny phage which receives such over- 

 lapping partial replicas would be recombin- 

 ant for the markers we are following and 

 diploid for the segment between them. It is 

 possible that the ends of the partial replicas 

 may join to produce a single strand with a 

 tandem duplication. Had the parent phages 

 differed in the genie alternatives (xi vs. X2) 

 present in the segment which became diploid, 

 the recombinant phage would also be partially 

 heterozygous. Of course, should partial 

 replicas overlap in some other region, the 

 progeny phage containing the overlap would 

 not be recombinant for the markers we have 

 been following, and recombination would be 

 identified only if this new region had been 

 suitably marked genetically. It is known that 

 a single phage particle can be heterozygous 

 for several loci, provided these are located 

 far enough apart. In fact, it is unlikely, 

 contrary to the possibility already mentioned, 

 that any phage is completely haploid. Con- 

 sidered in this way, phage recombination is a 

 consequence of the occurrence of copy-choice 

 in the formation of partial DNA replicas. 

 This explanation receives support '^ from the 

 fact that the observed frequency of hetero- 

 zygotes is great enough to explain the ob- 

 served frequency of recombination. 



The preceding explanation of the mecha- 

 ' From work of C. Levinthal. 



