Viruses : Recombination in Bacteriophage {II) 



395 



estimated by the summation of a number of 

 small distances. On these bases, then, the 

 total genetic map of phage T4 can be calcu- 

 lated to be approximately 800 recombination 

 units long, i.e., shows 800% recombination 

 with respect to its total genetic content. 

 (Recall that a recombination map based upon 

 crossing over can be longer than 100 re- 

 combination [in this case, crossover] units.) 

 With reference to the primary assumption 

 (2) it may be recalled that there are about 

 400,000 nucleotides per phage, represented by 

 a double helix of 200,000 linearly arranged 

 nucleotides. 



The fraction 800% recombinations/200,000 

 nucleotides equals 0.004%, and expresses the 

 percentage of recombinations which occurs 

 per linear nucleotide. On the reasonable 

 assumption that recombination cannot take 

 place within a nucleotide, there would be 

 199,999 points between nucleotides where 

 exchange may occur if the phage chromosome 

 is a rod, and 200,000 such points if the phage 

 chromosome is a ring. Expressed in this 

 particular way, we can say that if two r 

 mutants each differ from their wild-type form 

 by only a single nucleotide, and if the two 

 nucleotides changed are adjacent in /•+, then 

 recombinants (r+ and double mutant) would 

 be expected to occur among 0.004% of the 

 progeny obtained by crossing the mutants. 



Suppose that the lowest recombination 

 frequency observed, 0.02%, is actually the 

 minimal rate. This means that, on the aver- 

 age, only every fifth (.02/.004) internucleotide 

 point is capable of undergoing recombination, 

 so that the recon is estimated to be as small 

 as five nucleotides in length. Because tests 

 fail to give evidence that phage nucleotide 

 sequence is interrupted by non-nucleotide 

 material, it would seem reasonable that re- 

 combination could occur at any internucleo- 

 tide position. Accordingly, in view of the 

 fact that the observed value of 0.02% is a 

 maximal value for the least amount of re- 

 combination, and in view of the uncertainties 



which exist with regard to the length of the 

 genetic map and the number of nucleotides 

 in the phage genome, we can entertain the 

 working hypothesis that one recon equals one 

 nucleotide. ^ This is consistent with our expec- 

 tation regarding the polarity of a recon (see 

 p. 303). On this hypothesis, it should be 

 clear that the term allele is properly restricted 

 to the alternatives that occur for a particular 

 recon. In this way we are describing reconic 

 alleles, within which there can be no recombi- 

 nation or crossing over (see Chapter 22 and 

 p. 290 for previous usage of the term allele 

 in this sense). 



Finally, let us consider the functional 

 characteristics of the rll region. So long as 

 we are considering the production of the r 

 and r+ phenotypes, the rll region has a single 

 function. In this respect the whole region 

 behaves as a single cistron. But the rll region 

 is composed of two subregions, A and B, a 

 mutant in one subregion being able to 

 functionally complement a mutant in the 

 other. Such complementation suggests that 

 A and B are independent at this level of 

 functioning, and therefore each may be con- 

 sidered to be a separate cistron having a dif- 

 ferent function. The whole rll region 

 contains about 1000 linearly arranged nu- 

 cleotides, each subregion containing hundreds 

 of nucleotides. We do not know, but A and 

 B may order the amino acid sequence in dif- 

 ferent polypeptide chains, both of which must 

 be correctly specified in order to produce the 

 r+ phenotype. While the specification of a 

 polypeptide chain is accepted as one kind of 

 primary effect possible for a cistron, it should 

 be noted again (refer to Chapter 32) that the 

 primary effect of a cistron may not always be 

 to specify a polypeptide. Substances other 

 than polypeptides may be specified by the 

 primary action of a cistron; these substances 

 might conceivably be simpler than a polypep- 

 tide, for example single amino acids. In this 



^ Support for this view is found in work by D. R. 

 Helinski and C. Yanofsky (1962). 



