ALLELIC AND NONALLELIC GENES CONTROLLING HOST SPECIFICITY IN A BACTERIOPHAGE

Abstract

Preliminary crosses ( HERSHEY and ROTMAN 1948, and unpublished) between pairs of independently arising host-range mutants of the same virus, however, have failed to yield recombinants, even when the two mutants were clearly different. LURIA (personal communication) has had the same experience with host-range mutants of the related phage T2L. If the results with the two kinds of mutant are to be given a consistent interpretation, it has to be assumed that different host-range mutations, unlike the r mutations, tend to occur at a single genetic locus. This assumption is not implausible, inasmuch as the host-range character can be readily combined with the r character in the appropriate genetic crosses (HERSHEY and ROTMAN 1949). Nevertheless, we felt that the genetic interpretation could be greatly strengthened by establishing one authentic example of multiple allelism in the virus. This we have done, by three independent means. The demonstration is all the more convincing because one mutant was found that proved to be carrying a mutation in a second locus controlling host specificity. M ETH ODs The technique of crossing viral mutants has been modified slightly since last described (HERSHEY and ROTMAN 1949). The bacteria are now grown in nutrient broth, suspended in buffered saline, admixed with virus, and diluted in broth after allowing five minutes for adsorption to occur. Adsorption in the absence of nutrient is physiologically equivalent to simultaneous adsorption of the added virus particles ( BENZER et al. 1950), and the effects we observe are understandable on the assumption that this causes a more