218 A. GAREN AND L. M. KOZLOFF 



(Puck, 1953). A mutation to resistance to one phage strain frequently de- 

 creases the rate of attachment of another phage, although not sufficiently to 

 impart resistance to the other phage as well (Luria, 1945a,b), Thus, the effect 

 of a mutation may be more extensive than is indicated by the sole criterion 

 of resistance. The various patterns of resistance that can arise in a sensitive 

 cell make it apparent that the structure of the ceU surface is subject to 

 many different modifications through mutations. 



Kesistance to Tl in E. coli B can rise in two ways (Garen and Puck, 1951), 

 In one kind of mutant, B/1, irreversible attachment is blocked but reversible 

 attachment remains unimpaired. In another mutant, B/1, 5, neither attach- 

 ment reaction can occur. All other resistant mutants that have been studied 

 are, like B/1, 5, incapable of interacting either reversibly or irreversibly with 

 the phage. It should be recalled that since these mutants are selected for total 

 resistance to a phage, a selection is at the same time being made for muta- 

 tions that drastically affect the attachment reaction. There is hkely to exist 

 another important class of mutants having more subtle surface modifications. 

 The rate of phage attachment to such mutants might be slow but still suffi- 

 cient to prevent colony formation in the presence of the phage. These mutants 

 would therefore remain undetected by the usual procedures. 



Attachment reactions may also be affected when a cell is lysogenized 

 (Burnet and Lush, 1936; Bradley and Boyd, 1952; Boyd, 1954). This process 

 differs from the random mutations discussed above, in that the mutagenic 

 agent is known to be the lysogenizing phage, and each lysogenized cell is 

 modified in the same ways. We shall not be concerned here with the general 

 aspect of phage resistance in lysogenic cells, since this usually involves intra- 

 cellular reactions subsequent to attachment and injection (see Chapter 5). 

 The specific effects of lysogeny on attachment may be twofold: the lysogenic 

 cell may lose the capacity to react with the phage strain used to lysogenize, 

 and it may also gain the capacity to react with a different phage. For example, 

 Salmonella phage E15 can attach only to cells of S. anatum that are non- 

 lysogenic for E15, while another phage, E34, can attach only to the cells 

 that are lysogenic for E15 (Uetake et al., 1958). In another system, that of S. 

 typhimurium and phage P22, lysogenization decreases the rate of (but does 

 not completely block) attachment of P22 (Garen and Zinder, personal com- 

 munication). However, lysogeny does not always influence attachment of the 

 carried phage strain; coliphage lambda can attach equally well to lysogenic 

 (for lambda) and nonlysogenic E. coli K12 (Lieb, personal communication). 



It was pointed out many years ago by Burnet (1930) that the surface 

 antigens of the cell frequently determine attachment specificity. A particu- 

 larly clear example is provided by the serological changes in S. anatum which 

 result from lysogenization and are associated with the development of resist- 

 ance to phage. Cells lysogenic for phage E15 lack a surface antigen carried by 



