Antioxidant activity of O-protected derivatives of (-)-epigallocatechin-3-gallate: inhibition of soybean and rabbit 15-lipoxygenases

(-)-Epigallocatechin-3-gallate (EGCG) was protected at all eight phenolic groups as the methyl ether or acetate, propionate or butyrate esters. The synthetic derivatives were examined for antioxidant activities. The compounds were tested as potential diphenylpicrylhydrazyl (DPPH) radical scavengers and as inhibitors of 15-lipoxygenase (15-LO) enzymes from soybeans and rabbit reticulocytes. The O-protected EGCG derivatives were essentially inactive as DPPH scavengers but showed profound inhibiting effects on both types of 15-LO, and the inhibitory effects towards the two enzymes were significantly correlated. The O-protected EGCG derivatives inhibited 15-LO more strongly that EGCG itself. EGCG and its O-protected derivatives were also screened for antibacterial activity against M. tuberculosis


Introduction
Lipoxygenases (LO) have been found in a large number of higher plants. 1 It is believed that this enzyme class is involved in plant wound healing and it has been suggested that lipoxygenase metabolites are involved in plant hormonal activity, 2,3 growth regulation and senescence induction. 3We have previously reported that 6-substituted purines as well as indolizines act as soybean 15-lipoxygenase (15-LO) inhibitors. 4,5Nevertheless, our knowledge of effective 15-LO inhibitors is limited.From our ongoing studies on different new classes of 15-LO inhibitors we now wish to report that fully O-protected (-)-epigallocatechin-3-gallate derivatives (Figure 1) strongly inhibit both soybean and rabbit reticulocyte 15-LO.
In recent years, there has been considerable interest in the antioxidant, antitumor and antimutagenic properties of green tea and its constituents, particularly the flavan-3-ols. 6The main flavan-3-ol in green tea, epigallocatechin 3-gallate (EGCG), has been implicated as a major factor in the physiological effects of green tea consumption.Several in vivo experiments have shown chemopreventive effects of EGCG against cancer initiation, promotion and progression in animal models of oral, lung, duodenal, prostate, liver and colon cancers. 7Recently, it has been suggested that the peracetate of EGCG could be useful as a prodrug due to the limited stability and bioavailability of EGCG itself. 8t has been reported that overexpression of 15-LO in human prostate cancer cells increases tumorigenesis. 9Furthermore it is generally hypothesized that dietary polyphenols may reduce the risk of disorders associated with enhanced production of reactive oxygen and nitrogen species, such as coronary heart disease, stroke and inflammatory diseases as well as cancer. 102][13][14] Recently, it has been shown that some catechins may have

Results and discussion
We have previously synthesized a number of indolizines and azaindolizines and screened their ability to act as scavengers of lipid peroxidation in vitro, [16][17][18] and also recently investigated their role as 15-LO inhibitors. 5We found that an oxygen atom connected to the indolizine C-1 was required for antioxidant activity, but that a variety of oxygen substituents were tolerated.It has been shown previously that radical scavenging and 15-LO inhibition are unrelated properties. 19e have now investigated if the 15-LO inhibiting and radical scavenging properties are related to the free phenolic functionalities of EGCG or to other properties of the molecule.Consequently EGCG was permethylated by dimethyl sulfate, 20 or peracylated with acetic anhydride, propionic anhydride or butyric anhydride in the presence of N,N-dimethylaminopyridine in dichloroethane, and purified by flash chromatography on silica gel giving molecules II-V in moderate to good yields.(Figure 1) Substances II 21,22 and V 23,24 have been reported previously, but substance V has been characterized only incompletely.Substances III and IV have been prepared by us for the first time. 25,26he radical scavenging properties of catechins towards diphenylpicrylhydrazyl have been reported and oxidation products identified. 27Furthermore the reaction of EGCG with diphenylpicrylhydrazyl has been investigated using EPR spectroscopy. 28ur results are summarized in Tables 1-2.Linear regression gives the equation y = 2.428x -48.55 where y is the IC 50 value for the rabbit reticulocyte enzyme and x is the IC 50 value for the soybean enzyme.The R value for this correlation is 0.957.If substance I is omitted from the calculation (since its IC 50 value for the rabbit reticulocyte enzyme is determined by extrapolation), the correlation is still present, although somewhat weakened (R=0.850).
Quercetin was employed as positive control in both experiments, giving IC 50 values of 51±3 µM for the soybean enzyme and 37±4 µM for the rabbit reticulocyte enzyme.Our compounds are thus seen to compare well in activity with quercetin, a known 15-LO inhibitor. 5rom our results, it seems that derivatives without free hydroxyl groups of EGCG I are nearly devoid of radical scavenging activity in the DPPH assay, while I itself is an excellent scavenger as expected from previous knowledge.The derivatives are, however, considerably better inhibitors of 15-LO than EGCG.Furthermore, the correlation between inhibition of mammalian and soybean 15-LO shows that, at least for this type of compounds, the inexpensive and easily available soybean enzyme can be used for screening purposes, and that results obtained with this enzyme can be predictive for inhibition of the mammalian enzyme.
0][31] Our results are in accord with this.Recently, however, a lack of correlation was reported, based on the inability of epicatechin and caffeic acid to inhibit soybean 15-LO. 32][35][36] Differences in methodology could conceivably be the reason for this discrepancy.
Inhibition of soybean 15-LO by epicatechin, epicatechin gallate, epigallocatechin and EGCG has been reported. 34Apparently, the 3,4,5-trihydroxyphenyl moiety gave much higher inhibitory activity than the 3,4-dihydroxyphenyl one; epicatechin being one order of magnitude less active than the others.The stronger inhibitory effect of epigallocatechin compared to epicatechin was also observed by Goupy et al, 37 although differences in methodology makes it difficult to compare results directly.The effects of alkylation or esterification on the lipoxygenase inhibiting effect of EGCG or related compounds such as catechin or epicatechin have apparently not been reported previously.
The purpose of this part of the work was to synthesize derivatives of EGCG, the major flavonoid in tea, and to study the lipoxygenase inhibiting activity of EGCG and its derivatives in two different systems, soybean 15-LO and mammalian 15-LO.The results reported are in vitro, and at present, there are no data on in vivo activity of these substances.Since the fate of these substances in the GI tract and their bioavailability is not well known, 8 studies of their bioactivity in vivo would seem to be worth while in view of previously reported epidemiological results on the prevention of coronary heart disease by catechins. 15uberculosis caused by Mycobacterium tuberculosis (Mtb) is the leading cause of death in the world, greater that cancer or cardiovascular disease. 38As early as in 1949 flavonoids were reported to inhibit Mtb and it was reported that the action of streptomycin against the tubercle bacillus was markedly potentiated by the investigated catechins. 39Flavonoids have also very recently been reported as anti-tuberculosis agents. 40,41These recent reports prompt us to include our recent results for molecules I-V, (Table 3) in the present publication.Table 3. Inhibition of M. tuberculosis H 37 Rv at 6.5 µg/ml in BACTEC 12B medium using the microplate alamar blue assay. 42,43 pound Inhibition at 6.5 µg/ml (%) It was reported 40 that some O-methylated flavonoids inhibit Mtb, which is in accordance with the inhibition (38 % at 6.5 µg/ml) shown by the polymethylated compound V (Table 3).EGCG (I) itself shows 0 % inhibition.The finding that the polypropionylated compound III exhibits similar activity seems to have no precedent in the literature.

Conclusions
We conclude that EGCG and alkyl and acyl derivatives of EGCG are inhibitors of 15-LO from soybean and mammals in vitro.The derivatives are better inhibitors than the parent substance, but in contrast to EGCG, they are not radical scavengers.The permethylated and

Experimental Section
(-)-Epigallocatechin gallate (I) (from green tea, minimum 80% (HPLC)) was purchased from Sigma (St. Louis, MO, USA) and used as received or isolated from Formosa Gunpowder Tea. 44iphenylpicrylhydrazyl (DPPH) and soybean 15-lipoxygenase type 1-B (EC 1.13.11.12) were purchased from Sigma (St. Louis, MO, USA).Rabbit reticulocyte 15-lipoxygenase (Catalog No. PL-015, lot B2559, 5000 U/ml, solution in 10 mM potassium phosphate buffer (pH 6.0), supplied frozen) was purchased from BIOMOL Research Laboratories (Plymouth Meeting, PA, USA).Butyric anhydride and propionic anhydride were purchased from Fluka (Buchs, Switzerland).Silica gel for flash chromatography was available from Merck (Darmstadt, Germany) (Merck No. 9385) or Fluka (Fluka No 60752).HPLC grade acetone and HPLC grade acetonitrile were used as obtained from supplier, dichloroethane was distilled from CaH 2 .HPLC was performed on a Spectra System P 2000 equipped with an LC-18 Semi-Prep column, length 25 cm, ID-10 mm and a Spectra System UV 3000 detector. 1H NMR spectra were recorded at 500 MHz with a Bruker Avance DRX 500 instrument or at 300 MHz with a Bruker Avance DPX 300 instrument.The 13 C NMR spectra were recorded at 125 or 75 MHz using the instruments mentioned above.Unless otherwise stated, the spectra were recorded at ambient temperature.Chemical shifts (δ) are referenced to solvent shifts.Mass spectra were recorded with a Micromass QTOF2 instrument using electrospray ionization (solution for III and IV: MeCN + 2% formic acid, for II and V MeOH:H 2 O 50/50 + 2% formic acid) and presented as m/z (% rel.int.).Melting points are uncorrected.All measurements of DPPH scavenging and 15-LO activity were carried out in a Shimadzu UV-160A spectrophotometer (Shimadzu, Kyoto, Japan).
The frozen rabbit reticulocyte enzyme (in phosphate buffer, used as received) was thawed immediately before experiment and subsequently kept on ice throughout the experimental period (maximum one day).

ARKAT
Enzyme activity was measured according to the assay procedure supplied with the enzyme, using 2-methoxyethanol (which had much less effect on enzyme activity than DMSO) as a solvent for the test substances: A mixture of substrate solution (linoleic acid in abs.ethanol; 0.4% (w/v), 20 µL), 2-methoxyethanol (in uninhibited control experiments, 50 µL) or a methoxyethanolic solution of test substance (50 µL), and phosphate-buffered saline (0.1 M phosphate, pH 7.60, 2.80 mL) was stirred in a quartz cuvette.The oxidation reaction was started by addition of 8 µL enzyme solution.After brief stirring, the increase in absorbance at 234 nm was measured from 30 to 90 s after enzyme addition.

Table 1 .
Scavenging of the DPPH radical by epigallocatechin gallate (I) and derivatives.Final concentration of test substance was 667 µM.Results are shown ± SD.

Table 2 .
Inhibition of soybean and rabbit reticulocyte 15-lipoxygenase by epigallocatechin gallate (I) and derivatives.Results are shown ± SD.IC 50 denotes the concentration needed for 50% inhibition of the enzyme.