Many organic π-electron heterosystems, including natural products, display a particular case of the constitutional isomerism of functional groups called prototropy. This structural phenomenon has been clearly explained, more than eighty years ago, by Pauling [1], who not only showed the fundamental relation between prototropy and resonance, but also indicated the important difference between tautomeric and resonance structures. According to his explanation, prototropic conversions are reversible processes that can run intra- or intermolecularly. During tautomerization, labile proton(s) can move between two or more conjugated functional groups together with the delocalization of π-electrons, leading to the mixture of two or more constitutional isomers, called tautomers.

Prototropic tautomers always differ by the positions of labile proton(s) and π-electrons ^[1][2][1,2]. The number of possible tautomeric forms is an internal property of the tautomeric molecule. It is a consequence of the number of labile protons and the number of conjugated tautomeric sites. Although the most favored tautomer is very often selected to determine the name and formula of the tautomeric compound, which cannot be identified only with one Lewis structure. Each tautomeric derivative can be described by means of two (or more) structures (tautomers) being in equilibrium, whereas electron delocalization in each tautomer can be described by the corresponding resonance hybrid. For a single tautomer, the number of possible resonance structures results from the position of labile protons and double bonds. A different situation takes place for the relative stabilities of individual tautomers. They strongly depend on various internal and external factors that affect tautomeric preferences. Among the internal factors, the polarity, resonance stability (aromaticity), acidity–basicity of conjugated tautomeric sites, stability of functional groups, and substituents effects, as well as intramolecular interactions, play a particular role. For the external factors, usually, the solvent, pH, excess electron(s), other molecules, ions, radicals, oxidizing or reducing agents, ultraviolet (UV), and γ- and X-ray are considered.

The Pauling explanation of the prototropy phenomenon [1] has been employed in the IUPAC definition of prototropic tautomerism (IUPAC—International Union on Pure and Applied Chemistry) [3]. Only proton-transfers accompanied by the migration of double bonds refer to prototropic conversions in the tautomeric molecule. In other words, prototropic rearrangements always run in relation with electron delocalization ^[1][2][3][1,2,3]. The labile proton(s) can move from proton-donor site(s) to proton-acceptor site(s) separated by different conjugated spacers according to 1,3-, 1,5-, 1,7-, 1,9-proton shift, etc. Other intramolecular transfer(s) of H⁺ or H^● leading to a separation of positive and negative charges or to a separation of paired electrons cannot be considered as prototropy, and, consequently, zwitterions or polyvalent radicals formed in these processes cannot be classified as prototropic tautomers.

Prototropic conversions in aromatic heterocompounds, including nucleic acid bases, have been reviewed by Katritzky (died in 2014) and his co-workers in the 1960–2010 period (see, for example, refs. ^[4][5][4,5]). They compiled experimental and computational results mainly for favored tautomers (percentage contents > 1%), and considered most minor (<1%) and all rare tautomers (<0.01%) as negligible in tautomeric mixtures. This kind of treatment of tautomeric systems has led to some discrepancies in the literature, particularly for ionized, protonated, and deprotonated forms, for which prototropy has been usually neglected. Experimental and/or theoretical investigations have been carried out for tautomers that are favored in neutral isomeric mixtures. In the case of pyrimidine nucleic acid bases, the canonical forms or their major tautomers (two or three structures) have been the most frequently considered. These kinds of investigations for the selected isomers are usually partial.

2. Principles of Prototropic Equilibria

Four types of prototropic conversions {keto-enol, imine-enamine, imine-amine (amidine), and/or amide-iminol} can be distinguished for pyrimidine bases, uracil (U), thymine (T), cytosine(C), isocytosine {iC—structural part of guanine (G)}, and 4-aminopyrimidine {4APM—structural part of adenine (A)}, as well as for bicyclic purine bases, G and A, and for their metabolites such as hypoxanthine (HX), xanthine (X), and uric acid (UA). These equilibria are summarized in Table 1 for selected tautomeric moieties. For all of them, the labile proton can move between the conjugated sites according to the analogous scheme of reversible inter- or intramolecular rearrangement accompanied by the migration of the corresponding π-electrons ^[1][2][3][1,2,3]. Keto-enol tautomerism occurs in derivatives containing the C=O group and at least one H atom at the neighboring ^αC-sp³ or other C-sp³ atom separated from C=O by a conjugated spacer {e.g., –(CH=CH)_i–, i = 1, 2, 3, etc.}. In prototropic conversion, H is transferred as a proton from the conjugated C-sp³ to O-carbonyl, and vice versa, leading to two tautomeric forms, called keto and enol tautomers being in equilibrium ^[1][2][3][6][1,2,3,6]. For the majority of the neutral aliphatic carbonyl compounds, the labile proton prefers C-sp³, while the labile π-electrons favor O-carbonyl [7]. This means that the keto isomer predominates in carbonyl compounds more frequently than the enol form. The enol tautomer requires an extra stabilization, e.g., intramolecular H-bond formation and electron conjugation that occur in the enol isomers of β-ketoaldehydes, β-diketones, β-ketoacids, β-ketoesters, and β-ketoamides. The extra intramolecular interactions decrease considerably the enol-form energy in comparison to the keto one that the enol tautomer is favored in the gas phase and non-polar environments ^[6][8][6,8]. Intermolecular interactions with polar solvents (e.g., water) destruct intramolecular H-bonding and diminish the enol-isomer amount in favor of the keto one [9]. Analogous isomeric phenomena, such as imine-enamine, imine-amine (amidine), and amide-iminol conversions, occur for other neutral simple tautomeric systems containing heteroatom(s) in the conjugated tautomeric parts [2]. The labile proton can move from one to the other conjugated site, i.e., from C to N, from N to N, or from N to O, respectively, and vice versa. In parallel to proton-transfer, the migration of π-electrons takes place. Compounds containing one tautomeric part without any conjugated spacer are called triad conjugated systems. They display particular amphoteric properties. The sp² atom (with labile π-electrons) is a protonation site (base center) and the sp³ atom (with labile proton) is a deprotonation site (acid center). However, there is a principal difference between tautomeric systems and classic amphoteric compounds such as amino acids. Intra- or intermolecular proton-transfer (PT) in amino acid leads to the zwitterionic form, whereas that in a tautomeric system leads to the other isomer (tautomer) without the positive- and negative-charge separation. In the case of simple aliphatic tautomeric triad systems, the more electronegative atom (less basic) prefers the labile π-electrons and forms the double bond with the neighboring atom ^{[2][7][10][11]}[2,7,10,11]. The other conjugated atom (more basic) favors the labile proton. Generally, the acid–base properties of the conjugated sites dictate the tautomeric preferences ^[2][4][5][2,4,5]. In the case of heterocompounds possessing one labile proton and three, four, five, or more conjugated sites, tautomeric conversions become more complex. The labile proton can be transferred according to a 1,3-, 1,5-, 1,7-, 1,9-, or 1,n-proton shift, leading to the tautomeric mixture consisting of more than two tautomers [2]. The same is true for tautomeric compounds containing more than one labile proton and more than one pair of conjugated sites. The number of possible tautomers, exactly defined by the number of labile protons and by the number of conjugated sites, is also larger than two. Nevertheless, the principles of proton-transfer and π-electron delocalization are always analogous to those for the triad heterosystems ^[1][2][3][1,2,3]. Prototropic preferences depend on the acid–base properties of the conjugated sites. Generally, the least basic (least acidic) tautomer predominates in the tautomeric mixture ^[2][4][2,4]. Some exceptions can be found for prototropic derivatives displaying intramolecular interactions between functional tautomeric groups or intermolecular interactions of these groups with other molecules, ions, or radicals ^[2][4][6][9][2,4,6,9]. For these kinds of systems, the general rule on tautomeric preferences for neutral isolated compounds can change, because the acid–base properties of tautomeric sites can be different for neutral, ionic, radical, and associated species. Tautomeric aromatic derivatives, for which aromaticity can play a more important role than the general rule of acid-base properties, are very exciting derivatives [4]. For example, mono-hydroxy arenes exhibit keto-enol tautomerism; however, the enol forms most frequently predominate in the tautomeric mixtures ^[2][12][2,12]. The presence of the endo C-sp³ atom in the keto forms destructs electron delocalization in the ring, increases the energy of the keto forms, and reduces their amounts in the isomeric mixtures. In other words, the lower-percentage content of the keto forms is a consequence of the higher stability (aromaticity) of the enol isomer. Unsubstituted phenol (C₆H₅OH in Figure 1 for X = O), a parent system of mono-hydroxy arenes, is a classic derivative displaying this trend. For phenol, the energy of aromatic stabilization is higher than that of prototropy [13], and, consequently, the aromaticity of the ring is a pivotal factor that dictates the higher amount of the enol isomer (hydroxybenzene, a) than the keto ones (cyclohexa-2,4- and -2,5-dienones, b–d; note that b has an identical constitution to d). The experimental tautomeric equilibrium constants for enolization in an aqueous solution for cyclohexa-2,4- and -2,5-dienones (generated by flash photolysis) [14] are almost the same as those estimated theoretically in the gas phase by various quantum-chemical methods ^{[14][15][16][17][18]}[14,15,16,17,18]. Owing to the insignificant amounts of the keto isomers (<10⁻¹⁰%), cyclohexadienones are usually not considered in the structural and acid-base chemistry of neutral phenol. However, in organic chemistry, the keto forms of phenols are frequently used as intermediate structures to explain the mechanism and product(s) formation of various organic reactions such as oxidative metabolism, electrophilic substitution, ionization processes, and the Kolbe-Schmitt and Reimer-Tiemann reactions ^{[12][17][18][19]}[12,17,18,19]. More complex keto-enol conversions occur for mono-hydroxy arenes containing two or more condensed rings ^[16][20][16,20]. For example, 9-anthrol is only slightly less stable than its keto isomer, 9-anthrone. In the isomeric mixture, the keto form coexists with the enol tautomer. An extra stability of 9-anthrone originates from an aromatic character of two marginal rings. A slight amount of the keto form (<1%) can also be found in 9-phenanthrol. However, the keto isomers of 1- and 2-naphthols can be neglected in the isomeric mixtures. The same is true for those of 1- and 2-anthrols, as well as of 1-, 2-, 3-, and 4- phenanthrols. Nevertheless, the percentage contents of some of them are considerably higher (>10⁻⁶%) than those of the keto forms of unsubstituted phenol (<10⁻¹⁰%). Mono-hydroxy azulenes (azulenols), constitutional isomers of naphthols, are also interesting arene derivatives. Contrary to 1- and 2-naphthols, at least one keto form containing the labile proton at the C atom of the five-membered ring significantly contributes to the isomeric mixtures of azulenols [21]. For example, the keto and enol isomers of 2- and 5-hydroxyazulenes coexist in almost equal amounts in their gaseous isomeric mixtures. The percentage contents of the keto forms in other azulenols (1-, 4-, and 6-hydroxyazulenes) are not larger than 1%. The high amounts of the keto tautomers are strongly related with the polarity of the azulene system. Other factors, such as the acidity-basicity of endo CH/CH₂, also seem to play an important role in the tautomeric composition of azulenols. Owing to the azulene-system polarity, larger amounts of the keto-isomers are found in the gas phase (non-polar environment) than in the aqueous solution (polar medium). The six-membered aromatic ring is also present in the mono-amino arene—aniline (C₆H₅NH₂). Like C₆H₅OH, it contains one labile proton and four conjugated sites (Figure 1 for X = NH). However, amino benzene displays the other type of prototropy (enamine-imine tautomerism) compared to phenol [22]. The labile proton can move according to the 1,3-, 1,5, or 1,7-proton shift from the exo NH₂ group to the endo C-sp² atom being at the 2-, 4-, or 6-position vis-à-vis NH₂. Four tautomers are thus possible for aniline (a–d, where b and d possess an identical constitution). The aromatic enamino form a dictates the tautomeric preference in aniline. The transfer of the labile proton to the ring C-sp² and its transformation into C-sp³ strongly changes the delocalization of π-electrons in the imino tautomers b–d, reduces the stability of the ring, and increases the energies of b–d, like for the keto forms of phenol. From a physicochemical point of view, the imino tautomers can be neglected in the tautomeric mixture. They are exceptionally rare isomers of aniline (<10⁻¹⁵% [22]). However, the imine forms are often used as intermediate structures to understand the mechanism of various chemical reactions and to confirm the product formation ^{[19][22][23][24][25][26]}[19,22,23,24,25,26]. The resonance energy (aromatic stability > 30 kcal mol⁻¹ [27]), being considerably higher than the energy of prototropy in aniline, can only explain the change of isomeric preference when going from simple aliphatic to aromatic enamine-imine tautomeric systems. Biomolecules, such as nucleic acid bases, their metabolites, and model compounds, consist of aromatic ring(s) with the exo NH₂ and/or OH groups and endo N atoms. They exhibit various types of prototropic equilibria (given in Table 1) that originate mainly from the presence of the exo groups, conjugated with the corresponding endo atoms (N and C) ^[2][4][5][2,4,5]. In the case of pyrimidine bases (U, T, C, and iC) containing one six-membered ring with two endo N atoms and two exo groups (NH₂ and/or OH), two labile protons can move between the conjugated sites according to the analogous scheme of proton-transfers as that given in Figure 1 [28]. For bicyclic purine bases (A and G) and their metabolites (HX, X, and UA) containing the six-membered pyrimidine fragment structurally fused with the five-membered imidazole ring, the imidazole part contains additional labile proton(s) and additional conjugated sites that can also participate in prototropy ^{[29][30][31][32][33][34]}[29,30,31,32,33,34]. Consequently, the tautomeric equilibria for purine derivatives are more complex than those for pyrimidine bases. For adenine and its model compounds (imidazole, 4APM, and purine), the complete prototropic conversions have been already reviewed [35].