The protonation of the cyclic nitrogen atoms of cytosine and adenine is also facilitated by possible delocalization of the positive charge:

It is evident from Table 4-1 that adenine and cytosine as nucleoside and nucleotide constituents are weaker bases than aniline (pK_a 4.6).

Bases which can exist in nueleosides or nucleotides in a tautomeric hydroxy form and lack amino groups in the heterocyclic nucleus (uracil, thymine) easily give away their proton in an alkaline medium; that is, they exhibit acid properties:

As can be seen from Table 4-1, uracil and thymine are close to phenol in terms of protonation capacity (pK_a 9.99).

Guanine can display both basic and acid properties: it accepts a proton in an acid medium and loses it in an alkaline one. Just as in the above cases, the site of proton attachment is the nitrogen of the heterocyclic nucleus, rather than the amino group. Calculations have shown that the electron density is maximum at N7 Protonation of the latter is confirmed by spectroscopic data and also by increased lability of the glycosidic bond in the cation of guanosine, caused by delocalization of the positive charge at both nitrogens in the imidazole ring:

The pKa values for guanine (see Table 4-1) indicate that it is a weaker base,, as compared to cytosine and adenine, and closer to uracil and thymine in acid behavior (all bases are nucleoside or nucleotide constituents). The nature of the pentose moiety exerts only a minor influence on the ionization constant. The observed decrease in pKa from the base to deoxyribonucleoside and further to ribonucleoside can be explained by the negative inductive effect of pentose, which is weaker in the case of deoxyribose, as opposed to ribose.

The pKa value for the base is strongly affected by the presence of phosphate groups in nucleotides, albeit in a complicated manner. Since it might appear that the inductive effect of the phosphate group would lead to a further decrease in the pKa value for the base, whereas in fact the opposite is true (it is greater in nucleotides than in nucleosides), one may assume that the effect of the phosphate group manifests itself in steric interactions. Proceeding from the structural formulas of nucleosides, it would be logical to assume that such interactions between the base and phosphate group must be especially pronounced in the case of nucleoside 5'-phosphates whose phosphate groups may be sterically close to the base. The 2'(3’)-phosphate group which is in transposition with respect to the heterocycle must exert a weaker influence on the properties of the base.

The pK, values corresponding to the first and second steps of ionization of the phosphate groups in nucleotides are given in Table 4-1. As can be inferred from the tabulated data, the pKa value corresponding to the first step of ionization ranges from 0.7 to 0.9 and that corresponding to the second step ranges from 6.0 to 6.4.

Nucleoside cyclic phosphates are essentially monobasic acids with a pKa ~ 0.7.

The pKa decreases in the order uridine 5’-phosphate > adenosine 5’-phosphate > cytidine 5’-phosphate > guanosine 5’-phosphate which is indicative of the influence exerted by the cation emerging during protonation of the base on the first step of ionization of the phosphate group. Among the possible reasons for this phenomenon is intramolecular hydrogen bonding, for example:

Comparison of the pKa values for the phosphate group of nucleotides (0.7-1.0) with that for methyl phosphate (1.5) shows that in nucleotides the first step of dissociation of the phosphate group is strongly affected by the nucleoside.

Comparative pKa values for the bases and phosphate groups in four major nucleotides are represented in Figure 4-1 as dissociation curves at pH values ranging from 0 to 8.

Proceeding from the differences in the pKa values for the amino group, which, as can be seen from Figure 4-1, become manifest in the pH interval of 2 to 5, one can segregate a mixture of the four major nucleotides into individual compounds.

 In the above pH interval, the phosphate group is completely dissociated at the first hydroxyl group, whereas at the second one it is still deprived of any charge whatsoever (noticeable ionization begins after pH 5). The widest differences in the net charges of nueleotides are observed at pH 3.5, when the positive charge at the bases (see curves in Fig. 4-1) equals 0.54 for adenosine 5'-phosphate, 0.05 for guanosine 5'-phosphate, 0.84 for cytidine 5’-phosphate, and 0.0 for uridine 5'-phosphate. Since each nucleotide has a negative charge equal to unity at pH 3.5, the free negative charges (net charges) are, respectively, 0.46 for adenosine 5'-phosphate, 0.95 for guanosine 5’-phosphate, 0.16 for cytidine 5'-phosphate, and 1.0 for uridine 5'-phosphate. The molecules of these four nucleotides are almost of the same size and, consequently, this factor does not produce any decisive effect on their mobility. Thus, the above values of net charges indicate that the electrophoretic mobility of the molecules is relatively high at pH 3.5. The dissociation curves can be used to calculate the relative mobility of nucleotides for any other pH value, which is a widely used technique in the separation of various nucleotide mixtures both by electrophoresis and ion-exchange chromatography.

Thus, the nucleotide molecule as a whole is capable of displaying both base and acid behavior at virtually any pH value. The implication is that, depending on local proton concentrations, a nucleotide can act as their donor and also as their acceptor.