46 MAURICE J. BESSMAN 



clearly by .Io.xsjc et al. (1961) by iicaix'st iici<i;lib()r analysis. Within the 

 liinits of the niothods, (k'uxycytitlylatc was only found next to deoxy- 

 oytidylatc. No jiliosphodicster linkages were evident between deoxy- 

 ^nanylate and deoxycytidylate. Besides this dif'fei'ence, d-dC j)olyniei' 

 is distinct fi'oni d-AT in having uneciuai i)roj)ortions of deoxycytidylate 

 and deoxyjiuanylate. Usually, the molar ratio of cKlMP to dCMP is 

 greatei' than I. 



H. SUBSTITUTION OF ANALOGS 



An analog is any purine oi" pyriniidine base not normally found in 

 the DNA of a particulai- oi-ganism. P"or example, although 5-inethyl- 

 cytosine is a normal constituents of wheat germ DNA, it would be con- 

 sidered an analog if incorjjorated into E. coli DNA, because it is not 

 normally detected there. Several analogs have been synthesized as the 

 deoxyribonucleoside triphosphates and tested for incorporation into 

 DNA (Bessman et al, 1958b). When primed with E. coli DNA, the 

 following analogs support DNA synthesis: uracil, hyjioxanthine, 5- 

 bromouracil, 5-bromocytosine, and 5-methylcytosine. Xanthine was not 

 incorporated. Recent experiments from Romberg's laboratory (personal 

 comnnmication) have shown that 5-fluorocytosine and 5-fluorouracil are 

 incorporated into DNA whereas A^-methyl 5-fluorocytosine is not. 5- 

 Hydroxymethylcytosine, the pyrimidine base found exclusively in the 

 T-even bacteriophages (Wyatt and Cohen, 1953), is also incorporated 

 into DNA by E. coli polymerase primed with DNA prepared from several 

 different sources including calf thymus, E. coli and phage ^X174 (Korn- 

 berg et al, 1959). 



According to the restrictions imposed by the base-pairing rules of 

 the Watson-Crick model (1953b), these base analogs should be incorpo- 

 rated into specific positions in the DNA molecule. A diagram of these 

 relationships is shown in Fig. 16. The specificity in replacing a given 

 base by an analog was tested by determining which deoxyribonucleoside 

 triphosphate could be omitted from a reaction mixture in the presence 

 of a given analog and still support DNA synthesis. For example, in order 

 to find out which substrate could be replaced by deoxyuridine triphos- 

 phate, four reactions were set up, each lacking one of the normal 

 triphosphates. DNA synthesis proceeded only in the reaction mixture 

 lacking dTTP, suggesting that dUTP was replacing the dTTP require- 

 ment. Table XXII contains the results of several such experiments. It 

 can be seen that in each case the given analog substitutes for the purine 

 or pyrimidine base predicted by the Watson-Crick model. Table XXIII 

 shows direct evidence for the incorporation of deoxyuridylate into DNA. 

 The results are consistent with a specific substitution of deoxyuridylate 



