Mutagenic Alkylating Agents 313 



Many radiomimetic effects such as growth inhibition can 

 probably be explained on a DNA "deletion" hypothesis, i.e. 

 they are the result of rendering DNA, in one way or another, 

 useless to the cell. Obviously, crosslinking both intra- and inter- 

 molecular will be most effective for this purpose since a very 

 few reactions per molecule entirely alter its physical character 

 and are bound to render it biologically inactive. Monofunc- 

 tional reagents, on the other hand, will only esterify isolated 

 phosphate groups and it seems very reasonable that the 

 blocking of a few of these may be sufficient to spoil the mole- 

 cule and that a number of these random esterifieations are 

 necessary before biological activity is lost. This would explain 

 why higher concentrations of the monofunctional compounds 

 have to be used to produce the same effect. 



Probably the lethal as well as some visible and biochemical 

 mutations can be satisfactorily explained by this DNA 

 "deletion" hypothesis since they are manifestations of the 

 loss of a gene. The major difficulty is to explain how non- 

 selective reactions like the alkylation of phosphate groups can 

 produce a change in gene function as opposed to a general 

 destruction of the molecule for biological purposes. We must 

 look to much more subtle processes than crosslinking for 

 causing a rearrangement of a DNA molecule so that it retains 

 its biological integrity but carries a different code. Although 

 we are in no position to put forward any hypothesis, experi- 

 ments on the production of isolated breaks in one of the 

 chains of the twin molecules may provide the basis for a 

 possible mechanism. 



The possible significance of "hidden breaks'* 



A break produced in one of the twdn chains of DNA does 

 not lead to a disruption of the molecule since this is main- 

 tained intact by the other chain. A decrease in the particle 

 size will only occur when there are interruptions in both 

 chains close together (i.e. within about 50 A) (Fig. 10). The 

 hidden breaks can, however, be revealed if the twin chains are 



