Reactions of Flavan-3-ols

Studies of the reactions of flavan-3-ols and particularly those of catechin have been central to the elucidation of the structure and development of uses for the condensed tannins. This work, initiated by Freudenberg and his colleagues at Heidelberg in the 1920s [see Weinges et al. (377) for a thorough review], continues to be an important aspect of condensed tannin chemistry. A wide range of electrophilic aromatic substitution reactions has been examined to obtain definitive evidence for the location of substitution (i.e. C-6 or C-8) of proanthocyanidins and to establish the influence of steric hindrance on the relative reactivity of these nucleophilic centers in flavan-3-ols. 

The first aspect of these reactions that must be considered is the difference between kinetic control (i.e. conditions are such that products are not cleaved back to the flavan-4-carbocation or quinone methide from the chain extender units and the flavan-3-ol from the terminal unit) and thermodynamic control (i.e condensation and cleavage reactions are permitted to occur repeatedly) (177, 316). 

The proanthocyanidins containing 5-deoxy upper units are known to be far more stable to acid-catalyzed cleavage than the 5,7-dihydroxy procyanidins. Therefore, provided that reaction conditions are sufficiently mild (i.e. 0.1 M HCl, 2 to 4 hours at 25 °C) - such as those employed by Botha et al. (26-29) - an assumption of kinetic control of the regioselectivity of condensation of these compounds seems appropriate. Botha et al. (26-29) have done a great deal of work to define relationships of the structures of flavan-4-carbocations or quinone methides and flavan-3-ols on the regioselectivity of their condensations in the synthesis of proanthocyanidins (Thble 7.6.6). 

The situation is more complex in defining the regioselectivity of procyanidin synthesis. Here differences between kinetic control and thermodynamic equilibrium ratios become particularly important because of the lability of the in-terflavanoid bond in these compounds and because of differences in both the relative rate of acid-catalyzed cleavage and rate of condensation for the C-6 and C-8 substituted isomers (148). Work of Haslam s group indicated that the C-8 substituted procyanidins predominated over their C-6 linked pairs by a factor of 8 to 9 to 1 (177, 352). However, Botha et al. (28) obtained catechin-(4a- 8)-catechin and catechin-(4a- 6)-catechin in relative yields of 3.2 to 1 from a 2-hour bio-mimetic synthesis with 0.1 M HCl at ambient temperature. Similarly, Hemingway et al. (143) obtained epicatechin-(4)ff- 8)-catechin and epicatechin-(4)ff- 6)-cate-chin in relative yields of about 2.5 to 1 through synthesis by reaction of Pinus taeda proanthocyanidins with excess catechin for 48 hours at 25 °C using HCl as a catalyst. This ratio was similar to the yield of the two isomers isolated from the phloem of Pinus taeda. The extreme lability of the interflavanoid bond in the procyanidins causes one to wonder if true kinetic control ratios can be obtained from acid-catalyzed reactions of the procyanidins. 

Two other approaches to determination of a kinetic control ratio for substitution at the C-6 and C-8 positions of procyanidins have been attempted (139,148). Determination of the relative rate of cleavage of C-8 and C-6 regio-isomers (2.6-1) and a determination of their relative yields at the thermodynamic equilibrium (1.3-1) permitted an estimation of a kinetic control ratio of 3.3 1 for the C-8 and C-6 linked isomers (148). The other approach was to determine the ratio of the C-8 and C-6 linked dimers after synthesis by reaction of epicatechin-(4)ff)-phenylsulfide and catechin at pH 9.0 and ambient temperature through quinone methide intermediates (139). Here the C-8 and C-6 linked isomers were obtained in an approximate ratio of 3.5-4 1. These later results, together with evidence for interflavanoid linkage isomerism in trimeric and polymeric procyanidins (141,143) (Sect. 7.6.3.3), show that substitution is not as heavily favored at the C-8 position in procyanidins as had been thought. 

---

Van der Westhuizen, J.H., Steenkamp, J.A., and Ferreira, D., An unusual reaction of flavan-3-ols with acetone of relevance to the formation of the tetracyclic ring system in peltogynoids, Tetrahedron, 46, 7849, 1990. 

The discussions of the reactions of flavan-3-ols (Sect. 7.6.3.1.3) and of the flavan-3,4-diols (Sect. 7.6.3.2.3) are, for the most part, directly applicable to those of the oligomeric proanthocyanidins and condensed tannins, the critical difference, of course, being reactions at the interflavanoid bond. Although this difference is obvious, the interflavanoid bond, particularly its lability to either acid- or base-catalyzed solvolysis (107, 144, 148, 152, 225), has not been given adequate consideration in many instances. Reactions of condensed tannins are sometimes incorrectly postulated to parallel those of the flavan-3-ols. It seems far more appropriate now to use dimeric proanthocyanidins as model compounds for study of the reactions of condensed tannins, because these compounds are easily synthesized and their spectral properties are well known. 

Flavan-3,4-diols FIavan-3,4-diols, also known as leucoanthocyanidins, are not particularly prevalent in the plant kingdom, instead being themselves precursors of flavan-3-ols (catechins), anthocyanidins, and condensed tannins (proanthocyanidins) (see Fig. 5.4). Flavan-3,4-diols are synthesized from dihydroflavonol precursors by the enzyme dihydroflavonol 4-reductase (DFR), through an NADPH-dependent reaction (Anderson and Markham 2006). The substrate binding affinity of DFR is paramount in determining which types of downstream anthocyanins are synthesized, with many fruits and flowers unable to synthesize pelargonidin type anthocyanins, because their particular DFR enzymes cannot accept dihydrokaempferol as a substrate (Anderson and Markham 2006). 

For phloroglucinolysis, a solution of 0.1 N HCl in MeOH, containing 50 g/L phloroglucinol and 10 g/L ascorbic acid, is prepared. The PA of interest is reacted in this solution at 50°C for 20 min and then combined with 5 volumes of 40 mM aqueous sodium acetate to stop the reaction. After acid-catalyzed cleavage in the presence of phloroglucinol, the fraction is depolymerized and the terminal subunits released as flavan-3-ol monomers and the extension subunits released as phloroglucinol adducts of flavan-3-ol intermediates. These products are then separated and quantified by HPLC [25]. 

There are several caveats associated with this assay that may affect accuracy and precision. (+)-Catechin is the natural form in proanthocyanidins. Part of (+)-catechin epimerizes at the C2 position to form (+)-epicatechin during depolymerization. Similarly, part of (—)-epicatechin epimerizes to form (—)-catechin as an artifact. (+)-Catechin and (+)-epicatechin are an epimer pair in solution (similar to a- and 3-glucose in solution), i.e. they are chiral isomers that cannot be separated on a common reversed-phase HPLC column. The degree of epimerization increases with reaction temperature and time. Depolymerization at room temperature for 10 h caused less than 10% of flavan-3-ols to undergo epimerization. Toluene-a-thiol also causes the heterocyclic ring fission of flavan-3-ols to form adducts that... 

For analyzing molecular sizes of proanthocyanidins, depolymerization methods suffer from side reactions, such as epimerization and heterocyclic ring fission of flavan-3-ols. The depolymerization method estimates the average DP of proanthocyanidins in a mixture, but it is not able to determine the ratio of proanthocyanidins of different sizes. MALDI-TOF-MS is not able to analyze proanthocyanidins with a DP value higher than 12. The signal intensity of proanthocyanidins on MALDI-TOF-MS is not proportional to their amount, because proanthocyanidins of different sizes and structures differ in their... 

The factors that control the feasibility and the stereochemical course of the coupling process, as well as the methods to establish the configuration at C(4) of the condensation products and the mode of interflavanyl linkage were sufficiently reviewed (4, 27-29). Acid-catalyzed reactions to produce flavan-4-carbocations or A-ring quinone methides that may react with the A-rings of flavan-3-ols to produce oligo- and polymeric proanthocyanidins have been so successfully employed that they were called biomimetic syntheses (39, 40). The most recent variations of this theme are now briefly discussed. The nomenclature delineated in ref. (1) will be consistently employed. 

The redox potential of interest to understand the biological effects of flavan-3-ols is the one related to phenoxyl radical-phenate couple, as this potential is roughly 1 V lower than the potential of the phenoxyl radical-phenol couple, which furthermore may transiently involve the oxidation of the aromatic atoms. Standard potential can be measured by electrochemistry [49] or pulse radiolysis [40 4]. However, determining the redox potential of polyphenolic compounds is a real challenge since for these methods the measurement must be faster than the subsequent reactions induced by the oxidation of the phenol group in order to obtain the thermodynamic value. By using ultramicroelectrodes (electrodes with a micrometer diameter), it has been shown that a very high scan rate, up to 1 milUon... 

Other electrophilic aromatic substitutions of flavan-3-ols have been examined, primarily regarding the use of proanthocyanidins in adhesives. Hillis and Urbach (157-158) studied the reactions of catechin and condensed tannins with formaldehyde early. Later studies (147) through H-NMR spectra and GPC profiles of... 

Caution must be exercised in use of flavan-3-ols as models for the reactions of proanthocyanidins, because substitution at the C-2 or the C-4 positions of... 

The vast majority of recent works on the reactions of flavan-3,4-diols has centered on their role as precursors to flavan-4-carbocations or quinone methide electrophiles in proanthocyanidin syntheses. The regioselectivity of their condensation with flavan-3-ols or oligomeric proanthocyanidins is discussed in Sect. 7.6.3.1.2. Our concern here is with the factors governing the stability of the electrophile and the stereospecificity or stereoselectivity of its reaction with phenolic nucleophiles. Roux and coworkers have made extensive studies of these reactions and have prepared reviews of their recent results (316, 317). The intent here is to summarize some of the more important principles that have been developed. 

Flavanols and procyanidins Flavanols, or flavan-3-ols, are synthesized via two routes, with (+) catechins formed from flavan-3,4-diols via leucoanthocyanidin reductase (LAR), and (—) epicatechins from anthocyanidins via anthocyanidin reductase (ANR) (see Fig. 5.4). These flavan-3-ol molecules are then polymerized to condensed tannins (proanthocyanidins or procyanidins), widely varying in the number and nature of their component monomers and linkages (Aron and Kennedy 2008 Deluc and others 2008). It is still not known whether these polymerization reactions happen spontaneously, are enzyme catalyzed, or result from a mixture of both. 

It is important to monitor the time that the proanthocyanidin is allowed to react with the phloroglucinol solution. The products formed are not stable under acidic conditions, and it is therefore critically important that the reaction not exceed 20 min. Of particular concern are the flavan-3-ol monomers, which degrade more rapidly than the phloroglucinol adducts (Kennedy and Jones, 2001). Excessive degradation of the flavan-3-ol monomers will result in reduced amount of terminal subunits. This in turn will reduce the conversion yield and increase the average degree of polymerization calculated. 

---