Selective protection of catechin gives access to the intrinsic reactivity of the two phenol rings during H-abstraction and photo-oxidation

Article abstract of DOI:10.1016/S0040-4039(00)00944-8

Selective protection of the catechol ring of catechin has been achieved. From this key compound, catechin analogues protected either on the catechol or the resorcinol rings were synthesized. From these analogues, phenoxyl radicals on the catechol or on the resorcinol rings were produced by photo- oxidation (direct irradiation at 308 nm) of the phenolate and by H- abstraction experiments. Spectra of the radicals were recorded at short times before any further chemical evolution. Investigation of catechin behavior itself and comparison with the reactivity of models show that H-abstraction is unselective, whereas photo-oxidation is selective on the catechol ring monoanion establishing that this ring has the lowest pKa. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00944-8

Tetrahedron Letters 41 (2000) 5847±5851
Selective protection of catechin gives access to the intrinsic
reactivity of the two phenol rings during H-abstraction and
photo-oxidation
Cecile Cren-Olive,a Stephane Lebrun,a Philippe Hapiot,b Jean Pinsonb
and Christian Rolandoa,*
a
 Â
Universite des Sciences et Technologies de Lille, UPRESA 8009, Chimie Organique et Macromoleculaire,
Â
Equipe Polyphenols, Batiment C4, 59655 Villeneuve d'Ascq, France
 Â
Universite Paris 7-Denis Diderot, Laboratoire d'Electrochimie Moleculaire, UMR 7591, 2 Place Jussieu,
Ã
b
75251 Paris Cedex 05, France
Received 17 March 2000; accepted 2 June 2000
Abstract
Selective protection of the catechol ring of catechin has been achieved. From this key compound,
catechin analogues protected either on the catechol or the resorcinol rings were synthesized. From these
analogues, phenoxyl radicals on the catechol or on the resorcinol rings were produced by photo-oxidation
(direct irradiation at 308 nm) of the phenolate and by H-abstraction experiments. Spectra of the radicals
were recorded at short times before any further chemical evolution. Investigation of catechin behavior
itself and comparison with the reactivity of models show that H-abstraction is unselective, whereas photo-
oxidation is selective on the catechol ring monoanion establishing that this ring has the lowest pKa.
# 2000 Elsevier Science Ltd. All rights reserved.
Keywords: catechin; polyphenols; radical; oxidation; dissociation constant.
Epidemiological studies suggest that catechin which is relatively abundant in food1,2 is a key
compound in the relationship between health and diet. Indeed, catechin and related compounds
such as epi-catechin or epi-gallocatechin are known to combat ageing pathologies such as
cancers,3,4 cardiovascular5,6 and neurodegenerative7 diseases in which oxidative stress is involved.
These properties can be attributed to the ability of ¯avan-3-ols to protect biomolecules from
oxidative degradation, in particular by reactive oxygen species (e.g. hydroxyl, alkoxyl and alkyl-
peroxyl radicals, singlet dioxygen, etc.). Therefore, the ability of ¯avonoids to act as antioxidants
in vivo and in vitro has been extensively studied.8 16 But there is much discussion and many
contradictions regarding the catechin dissociation constants17,18 and the structure±activity
* Corresponding author. Fax: 333 20 33 61 36; e-mail: christian.rolando@univ-lille1.fr
0040-4039/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00944-8

5848
relationship of the antioxidant activity.8,12 16 Most of the problems encountered when describing
catechin antioxidative activity are due to the lack of information on the intrinsic reactivity of each
ring and more precisely of each phenol function.
Therefore, we decided to synthesize selectively protected catechin analogues to investigate more
thoroughly their free radical chemistry and determine, for the ®rst time, the reactivity of each
phenol function. Although (30,40,5,7)-tetramethyl catechin is readily accessible by alkylation of
catechin 1 in the presence of dimethylsulfate,19 partial methylation gives a complex untractable
mixture of products in low yield.20,21 Even benzylation of catechin is dicult leading to only fair
yield.22,23 Also, many classical protective groups for catechol are incompatible with catechin at the
introduction or removal step due to its instability both in slightly acidic or basic medium.24 Among
the various catechol-protecting groups investigated in our work, only dichlorodiphenylmethane
gave access to the protected catechol 2 albeit with a moderate yield (20%) (1.1 equivalent of
dichlorodiphenylmethane, 5 equivalents of triethylamine at room temperature for 18 hours).
Under partial methylating conditions of key intermediate 2 with dimethylsulfate (1 equivalent,
1.1 equivalent of K2CO3 at room temperature), a mixture of two monomethylethers 3, 4 with a
ratio 9:1 was obtained and could be separated by column chromatography. NOE experiments
clearly indicate that methylation takes place mostly on the 5-position. Exhaustive methylation
(3 equivalents of dimethylsulfate, 3 equivalents of K2CO3, re¯ux) of 2 lead nearly quantitatively
(94%) to 5,7-dimethylated compound 5. The quantitative deprotection was then achieved by
hydrogenolysis on palladium hydroxide in methanol at room temperature a€ording pure
compounds 6, 7 or 8.y Compounds 2 and 8 are appropriate models for, respectively, A and B ring
study whereas 3 allows the precise study of the 7 hydroxy reactivity (Scheme 1).
Scheme 1. Synthesis of selectively methylated catechin
Phenoxyl radicals have been generated in aprotic media by two di€erent methods: (i) direct
photo-oxidation by direct irradiation of the phenolate; (ii) H-atom abstraction from phenolic OH
by tert-butoxyl radical. A ¯ash photolysis experimental set-up (XeCl excimer laser
y
For 8:25,26 1H NMR (300 MHz, acetone-d6) ꢀ: 2.50 (1H, dd, J=16, 8 Hz), 2.86 (1H, dd, J=16, 6 Hz), 3.73 (3H, s),
3.78 (3H, s), 4.00 (1H, m), 4.59 (1H, d, J=8 Hz), 6.03 (1H, d, J=2 Hz), 6.11 (1H, d, J=2Hz), 6.75 (1H, dd, J=8, 2 Hz),
6.79 (1H, d, J=8 Hz), 6.88 (1H, d, J=2 Hz). 13C NMR (75 MHz, acetone-d6) ꢀ: 160.6, 159.5, 156.5, 145.6, 145.5, 131.9,
119.9, 115.7, 115.1, 102.6, 93.9, 91.9, 82.7, 68.0, 55.7, 55.4, 28.5.

5849
lirradiation=308 nm) was used. It allows the recording of UV±visible spectra of radicals at short
times (down to 100±200 ns after the laser pulse) and before further chemical evolution can occur.
Experiments were performed under strict deoxygenated conditions, as solutions of basic catechin are
not stable under air. In direct irradiation experiments, only the ionized ring absorbed the laser light at
308 nm with the typical concentrations of phenol used (10^4±5.10^4 mol L^1), whereas the neutral
compounds displayed no or negligible absorption at this wavelength. The appearance of radical
species at the shortest observation time and the low catechin concentration ruled out a mechanism
involving an electron transfer between a photo-exited state and traces of oxygen or other impurities.
Inspection of Table 1 shows that radicals from model compounds for rings A and B can
unambiguously be characterized by their di€erent UV-spectra. In agreement with previous
studies12,13 catechol-like phenoxyl radicals generated from 4-methylcatechol and 8 show an
absorption band at 380 nm whereas resorcinol-like phenoxyl radicals generated from 5-methoxy-
resorcinol, 2 and 3 present an absorption at 495 nm. Another band at 550 nm is visible for the
radicals from 5-methoxyresorcinol and 2 during photo-oxidation experiments which can be
ascribed to a further fast deprotonation of the neutral resorcinol radical to the corresponding
radical-anion in basic medium.27,28
Table 1
Absorption bands of phenoxyl radicals obtained upon direct ¯ash photolysis and H-abstraction experiments.
Delay after laser pulse respectively: 200 ns, 5 ms
Catechin exhibits a more complex behavior. H-abstraction of catechin by tert-butoxyl radicals
gives a spectrum characterized by two absorption bands, respectively, at l=380 and 495 nm (Fig. 1a).
The 308 nm laser-induced photo-oxidation of catechin phenolate in the presence of 1 equivalent
of base (when the two rings can partially be ionized due to the close pKa values for ring A and B)
produces a similar spectrum (l=380 and 495 nm), whereas in the presence of 0.5 equivalent of
base only one absorption band at l=380 nm is obtained (Fig. 1b). In both cases, experiments
performed at the shortest available detection time show that both types of radicals are already
produced 100 ns after the laser pulse. This fast production of radicals excludes an intermolecular
or long-range intramolecular H or proton transfer which are observed to take place at much
longer times.13
On one hand, H-abstraction by tert-butoxyl radical on catechin is clearly not selective leading
to an approximate equal reactivity between the two rings. This result can be explained by the
strong H-abstracting ability of tert-butoxyl radical toward most phenols.29 On the other hand,
the selectivity of direct irradiation experiments is related to the possibility of deprotonating a
single or several phenolic functions as only phenolates absorb the laser light and are much more easily
oxidized than neutral phenols. The 380 nm absorption is therefore related to the photo-oxidation of
the ®rst deprotonated phenolic group. Even if the successive catechin pKa values have been

5850
Figure 1. UV±vis spectra of catechin phenoxyl radical obtained upon: (a) H-abstraction experiments (solution of
8.27Â10^1 mol L^1 of di-tert-butylperoxide and 3.21Â10^3 mol L^1 of catechin in acetonitrile; average of ®ve spectra,
time after pulse: 2 ms) and (b) direct irradiation (solution of 2.55Â10^4 mol L^1 of catechin in acetonitrile with
1 equivalent (dashed line) or solution of 4.28Â10^4 mol L^1 of catechin with 0.5 equivalent (unbroken line) of tetra-
methylammonium hydroxide 0.1 mol L^1 to form the phenolate (see text); average of 10 spectra, time after pulse:
200 ns). Curves were obtained by high frequency ®ltering of experimental data displayed by dots on the graphs
determined by various methods17,18 agreeing with ®rst and second pKas at about 8.7þ0.1 and
9.4þ0.1 in water, the protonation sequence between the two rings has not been clearly established
and is still under discussion. Our results indicate that the ®rst pKa of rings A and B are close:
addition of 1 equivalent of base gives a mixture of phenolates leading to a mixture of phenoxyl
radicals after irradiation. The selectivity observed with 0.5 equivalent of base indicates that the B
catechol ring is slightly more acidic in acetonitrile. A similar order between the two rings must
occur in water as suggested by Slabbert17 as the relative acidities of phenols are generally not
a€ected by solvents e€ects.30 32
The synthesis of selectively protected catechin analogues allows the investigation by direct
photo-oxidation and H-abstraction experiments of the free radical chemistry of catechin: two
families of radicals have been characterized. H-abstraction kinetic measurements are under
investigation in order to evaluate the relative ability of catechin to intercept or inactivate
damaging free radicals before they reach vital cell components, or to chemically repair damaged
biomolecules and bioradicals thus reversing the oxidative damages.
Acknowledgements
This work was supported by the European Community (FAIR CT Contract 95-0653 Complex
Phenols Tannins and Health).
References
1. Kunhau, J. World Rev. Nutr. Diet 1976, 24, 117±191.
2. Bravo, L. Nutrition Reviews 1998, 56, 317±333.
3. Jankun, J.; Selman, S. H.; Swiercz, R.; Skrypczak-Jankun, E. Nature 1997, 387, 561.
4. Garbisa, S.; Biggin, S.; Cavallarin, N.; Sartor, L.; Benelli, R.; Albini, A. Nature Medicine 1999, 5, 1216.

5851
5. de Lorgeril, M.; Salen, P. Lancet 1999, 353, 1067.
6. Stein, J. H.; Keevil, J. G.; Wiebe, D. A.; Aeschlimann, S.; Folts, J. D. Circulation 1999, 100, 1050±1055.
7. Smith, M. A.; Perry, G.; Richey, P. L.; Sayre, L. M.; Anderson, V. E.; Beal, M. F.; Kowall, N. Nature 1996, 382,
120±121.
8. Rice-Evans, C. A.; Miller, N. J.; Paganga, G. Free Radic. Biol. Med. 1996, 20, 933±956.
9. Hapiot, P.; Neta, P.; Pinson, J.; Rolando, C.; Schneider, S. New J. Chem. 1993, 17, 211±224.
10. Hapiot, P.; Pinson, J.; Neta, P.; Francesch, C.; Mhamdi, F.; Rolando, C.; Schneider, S. Phytochemistry 1994, 36,
1013±1020.
11. Hapiot, P.; Neudeck, A.; Pinson, J.; Fulcrand, H.; Neta, P.; Rolando, C. J. Electroanal. Chem. 1996, 405, 169±176.
12. Steenken, S.; Neta, P. J. Phys. Chem. 1982, 86, 3661±3667.
13. Jovanovic, S. V.; Hara, Y.; Steenken, S.; Simic, M. G. J. Am. Chem. Soc. 1995, 117, 9881±9888.
14. Jovanovic, S. V.; Steenken, S.; Hara, Y.; Simic, M. G. J. Chem. Soc., Perkin Trans. 2 1996, 2497±2504.
15. Bors, W.; Michel, C.; Schikora, S. Free Radic. Biol. Med. 1995, 19, 45±52.
16. Bors, W.; Michel, C. Free Radic. Biol. Med. 1999, 27, 1413±1426.
17. Slabbert, N. P. Tetrahedron 1977, 33, 821±824.
18. Kennedy, J. A.; Munro, M. H. G.; Powell, H. K. J.; Porter, L. J.; Foo, Y. Aust. J. Chem. 1984, 57, 885±892.
19. Clark-Lewis, J. W.; Koryntnyk, W. J. Chem. Soc. 1958, 2367±2372.
20. Donavan, J. L.; Luthria, D. L.; Stremple, P.; Waterhouse, A. L. J. Chrom. B 1999, 726, 277±283.
21. Cren-Olive, C.; Deprez, S.; Coddeville, B.; Rolando, C. unpublished.
22. Tuckmantel, W.; Kozikowsky, A. P.; Romanczyk, L. J. J. Am. Chem. Soc. 1999, 121, 12073±12081.
23. Bu€noir, S. PhD Thesis, University Pierre et Marie Curie (Paris 6), 1998.
24. Kiatagrajai, P.; Wellons, J. D.; Gollod, L.; White, J. D. J. Org. Chem. 1982, 47, 2910±2912.
25. Balas, L.; Vercauteren, J. Magn. Reson. Chem. 1994, 32, 386±393.
26. Kashiwada, Y.; Nonaka, G. I.; Nishioka, I. Chem. Pharm. Bull. 1986, 34, 3208±3222.
27. Steenken, S.; Neta, P. J. Phys. Chem. 1979, 83, 1134±1137.
28. Steenken, S.; O'Neill, P. J. Phys. Chem. 1977, 81, 505±508.
29. Foti, M.; Ingold, K. U.; Lusztyk, J. J. Am. Chem. Soc. 1994, 116, 9440±9447.
30. Corse, J.; Ingraham, L. L. J. Am. Chem. Soc. 19, 73, 5706±5707.
31. Cohen, L. A.; Jones, W. M. J. Am. Chem. Soc. 1963, 85, 3397±3402.
32. Pearce, P. J.; Simkins, J. J. Can. J. Chem. 1968, 46, 241±248.