I. REPLICATION OF DNA IN CELL-FREE SYSTEMS 49 



1 dGMP residue per 10^ nucleotide units. Within the limits of the assay, 

 no significant incorporation of dGMP occurred in a reaction mixture 

 made up of d-AT primer, dOTP^^-, dATP, dTTP, Mg^ and E. coli 

 polymerase. However, when a polymer composed of dAMP and 5- 

 bromodeoxyuridylate (d-ABU) was used in place of d-AT as the primer, 

 there was a significant incorporation of dGMP. For example, in one 

 experiment less than 0.10 /xfimoles of dGMP^- was incorporated using 

 d-AT as a primer, whereas 2.82 ^^moles of dGMP''- was incorporated 

 using d-ABU as primer. 



Mistakes in base-pairing could result from tautomeric shifts in the 

 bases during replication (Watson and Crick, 1953b). The specific muta- 

 genic effect of 5-bromouracil which is incorporated into DNx\ in place 

 of thymine (Dunn and Smith, 1954) has been attributed to its facilitated 

 tautomerization to a fonii w^hich could pair with guanine instead of 

 adenine (Watson and Crick 1953b; Freese, 1959). If guanine only substi- 

 tuted for adenine in the d-ABU polymer described above, a nearest 

 neighbor analysis of a product prepared dGMP^- as the labeled deoxy- 

 ribonucleotide should reveal only one labeled 3'-deoxyribonucleotide, 

 that is, 5-bromodeoxyuridylate. When Trautner et al. (1962) analyzed 

 the product they found that all three possible 3'-deoxyribonucleotides 

 were labeled (Table XXIV). These data show there are just as many 



TABLE XXIV 



Nearest Neighbors of the Guanixe Residues Incorporated 



INTO THE d-ABU Polymer" 



" From Trautner et al. (1962). 



GG pairs as GBU pairs in the product and about one-half as many GA 

 pairs. This would suggest that the mutagenic effect of 5-bromouracil is 

 not entirely explained on the basis of a mistake in pairing with adenylate 

 instead of guanylate, but that the analog exerts a more complicated 

 effect. The authors are careful in pointing out that these results in vitro 

 with the d-ABU polymer do not necessarily represent the in vivo effect 

 of 5-bromouracil on whole DNA. Also, the argument outlined above 

 is predicted on the basis that the d-AT (and hence d-ABU) polymer 



