SEPARATION BY PAPER CHROMATOGRAPHY 257 



alcohol system of Carpenter*' (solvent d) should give adequate separation of the 

 nucleotides from PNA. Use of this mixture in analytical work has not yet been re- 

 ported. 



The isobutyric acid-ammonium isobutyrate solvent of Magasanik et al.*" (solvent 

 e) has been used for quantitative analysis of PNA's.** [Cf. Magasanik, Chapter 11.] 

 Since guanylic and uridylic acids occupy the same position on the chromatogram, 

 however, the amount of each must be calculated from extinction measurements taken 

 at two wavelengths. It is reported*^''' ^^ that the five deoxynucleotides from th.ymus 

 DNA are completely resolved from one another in this solvent, the respective dis- 

 tances travelled (relative to deoxyadenylic acid taken as 100) being: deoxyguanylic 

 acid, 53; thymidylic acid, 66; deoxycytidylic acid, 80; deoxyadenylic acid, 100; de- 

 oxy-5-methylcytidylic acid, 137. 



The phenolic solvent system of Boulanger and MontreuiP" (solvent /) is capable 

 of separating the ribonucleotides with excellent spacing, and was used for quanti- 

 tative analysis of PNA's. Because of the ultraviolet absorption of the phenol, how- 

 ever, it is necessary to detect and estimate the nucleotides by reactions of the phos- 

 phoryl group, and anj' other nucleic acid derivatives would of course be missed. 



For fractionation of the products of digestion of PNA by ribonuclease, Markham 

 and Smith''*' ''^ have used (1) 70% (bj- vol.) aqueous isopropanol and (2) 70% aqueous 

 isopropanol with concentrated ammonia added to the solvent in the bottom of the 

 tank at the rate of 0.35 ml. for each liter of gas space. The latter system proved espe- 

 cially useful for initial fractionation (prior to electrophoresis or chromatography in 

 other solvents) of the mixture of mono-, di-, and trinucleotides; in it the cyclic 

 nucleotides (riboside-2'3'-monohydrogen phosphates) move more rapidly than the 

 riboside-2'- and -3 '-phosphates. It also resolves nucleotides from their benzyl es- 

 ters.^' 



With systems based on salt solutions instead of organic solvents (Table II, solvents 

 g and h) , the purine ribonucleotides, but not those of the pyrimidines, have been sepa- 

 rated. Carter's^^ sodium and potassium phosphate systems resolve adenosine-2'-phos- 

 phate and -3'-phosphate. The resolution of adenosine phosphates in dibasic sodium 

 phosphate is said to be improved by substituting for isoamyl alcohol 0.5% lauryl 

 amine in n-amjd alcohol.^'' The ammonium sulfate mixture of Markham and Smith*" 

 gives better spacing of the isomeric nucleotides and separates also those of guanylic 

 acid and 8-azaguanylic acid;'" in this solvent the cyclic nucleotides move more slowly 

 than the corresponding nucleoside-2'- and -3 '-phosphates. ''^ 



With water as the solvent, the Rp values of all the nucleotides are close to 0.9.^* 



V. Quantitative Estimation of the Nitrogenous Components of 



Nucleic Acids 



1. Hydrolysis of Deoxypentose Nucleic Acids 



The purine and pyrimidine components of deoxypentose nucleic acids 



are satisfactorily estimated as free bases, since these may be obtained in 



*' E. Chargaff, B. Magasanik, E. Vischer, C. Gi'een, R. Doniger, and D. Elson, J. 



Biol. Chem. 186,51 (1950). 

 «' E. Chargaff, Federation Proc. 10, 654 (1951). 



'" P. Boulanger and J. Montreuil, Bull. sac. chim. biol. 33, 784, 791 (1951). 

 " R. Markham and J. D. Smith, Biochem. J. 52, 552 (1952). 

 " R. Markham and J. D. Smith, Biochem. J. 52, 558, 565 (1952). 

 " D. M. Brown and A. R. Todd, /. Chem. Soc. 1953, 2040. 

 " O. Snellmann and B. Gelotte, Nature 168, 461 (1951). 



