AU1410401A - S-allylmercaptoglutathione and uses thereof - Google Patents

Description

WO 01/36450 PCT/ILOO/00761 S-ALLYLMERCAPTOGLUTATHIONE AND USES THEREOF Field of the Invention 5 The present invention relates to a new glutathione derivative, namely S allylmercaptoglutathione, its preparation and pharmaceutical compositions comprising it. Abbreviations 10 BAPNA: N-c-benzoyl-DL-arginine p-nitroanilide; Biradical: bis (2,2,5,5 tetramethyl-3-imidazoline-1-oxyl-4-yl) disulfide; CSSA: S-allylmercapto cystein; DAD: diallyl disulfide; DTT: dithiothreitol; DMPO: 5,5'-dimethyl-1-pyrroline N oxide; DTNB: 5,5'-dithio-bis (2-nitrobenzoic acid); ESR: electron spin resonance; GSH: reduced glutathione; GSSA: S-allylmercaptoglutathione; GSSG: oxidized 15 glutathione; LPO: lipid peroxides production; NTB: 2-nitro-5-thiobenzoate; PBS: phosphate-buffered saline; RBC: red blood cell; SUV: small unilamellar phospholipid vesicles; TBA: thiobarbituric acid; TBAD: alcohol dehydrogenase from Thermoanaerobium brockii; TBARS: thiobarbituric acid reactive substance. 20 Background of the Invention Many beneficial health related biological effects of garlic (Allium sativum) are attributed to its characteristic organosulfur compounds. The best known and most extensively studied is allicin (diallyl thiosulfinate), the principal active substance of fresh garlic extract, which is responsible for garlic's typical pungent 25 smell. Allicin is produced during the crushing of garlic cloves by the interaction between the non-protein amino acid alliin (S-allyl-L-cysteine sulfoxide) and the enzyme alliinase. Allicin is a precursor of a number of secondary products formed in aged garlic and crushed garlic preparations. Allicin possesses various biological 30 activities among which antibacterial, antifungal and antiparasitic effects are included [Ankri et a., 1997; Koch and Lawson, 1996; Feldberg et al., 1988; 1 WO 01/36450 PCT/ILOO/00761 Cavallito et al., 1944). In addition to that it reduces serum cholesterol and triglycerides levels -as well as atherosclerotic plaque formation and platelet aggregation, it inhibits cancer promotion and decreases ocular pressure [Koch and Lawson, 1996; Chu et al., 1993]. Allicin rapidly disappears after injection into the 5 blood [Lawson and Wang, 1993; Freeman and Kodera, 1995]. This rather unstable compound has been suggested by Lawson and coworkers to transform rapidly into secondary products (in vivo), such as allylmercaptan and others [Koch and Lawson, 1996; Lawson and Wang, 1993]. At present a variety of biological effects of allicin is attributed to both antioxidant activity and modification of SH 10 dependent activities [Wills, 1956; Prasad et al., 1995]. These activities have been confirmed by us in model systems [Rabinkov et al., 1998; Miron et al., 1998]. Recently the inhibitory effect of allicin on NO formation was also demonstrated [Dirsch et al., 1998] . In addition to that, allicin affects the processing of DNA and RNA synthesis [Feldberg et al., 1988], signal transduction and apoptosis. All the 15 effects described above are mainly intracellular. Therefore the question arose: how does allicin permeate through the plasma membrane and take part in intracellular processes? The observation of fast disappearance of allicin from the blood on one hand and its high efficacy on the other, lead to the question about the ability of allicin 20 transformed products to generate extra- and intracellular processes. The most abundant non-protein thiols which can potentially interact with allicin are reduced glutathione (GSH) and cysteine. S-allylmercaptocysteine (CSSA) is known already as the product formed when cysteine reacts with allicin [Cavallito et al., 1944] . Aged garlic extract contains this compound as one of its active components. 25 Recently antioxidant, antiproliferative, decreasing ocular pressure activities of CSSA have been demonstrated [Imai et al., 1994; Lee et al., 1994; Sigounas et al., 1997a, 1997b; Pinto et al., 1997; Chu et al., 1999]. CSSA revealed antiproliferative effect on different cell lines whereas S-allylcysteine had no effect [Lee et al., 1994] 2 WO 01/36450 PCT/ILOO/0076 1 The concentration of GSH in the blood is about 100-fold higher than that of cysteine, therefore GSH remains as the main candidate for the interaction with allicin in vivo. 5 Summary of the Invention According to the present invention, while studying the permeability of allicin through lipid bilayers of phospholipid vesicles as well as through natural membranes of red blood cells (RBCs), using its interaction with internally entrapped thiol containing compounds such as glutathione (GSH), it was found that 10 when allicin interacts with GSH according to the Reaction 1 below, the reaction product formed is S-allylmercaptoglutathione (hereinafter GSSA): 0 2 GSH + CH 2

=CH-CH

2

-S-S-CH

2

-CH=CH

2 2 G-s-s-CH 2

-H=CH

2 + H20 Glutathione Allicin S-Allylmercaptoglutathione Reaction 1, wherein G is the residue of glutathione. 15 The complete formula of GSSA is as follows:

H

2

NCHCH

2

CH

2

CONHCHCONHCH

2 COOH COOH CH 2 -S-S-CH2-CH=CH 2 The present invention thus relates to the novel synthetic glutathione 20 derivative, S-allylmercaptoglutathione, and to salts thereof, to its preparation by reaction of glutathione with allicin or with diallyl disulfide, and to pharmaceutical compositions comprising it. Also included in the invention are salts of GSSA. The term "salts", as used herein, refers both to salts of carboxyl groups and to acid addition salts of the 25 amino group. Salts of a carboxyl group may be formed by methods known in the art and include inorganic salts such as salts with alkali metals, e.g. sodium or potassium, and salts with ammonia or with organic bases such as with amines. 3 WO 01/36450 PCT/ILOO/0076 1 Acid addition salts include salts with mineral acids such as hydrochloric acid or sulfuric acid and salts with organic acids such as acetic acid. The pharmaceutical compositions comprising S-allylmercaptoglutathione are useful for the treatment of several disorders including atherosclerosis, coronary 5 artery diseases, thrombosis, high levels of cholesterol and blood lipids, high blood pressure, control of weight, Alzheimer disease, glaucoma, cancer and inflammatory disorders such as colitis. Brief Description of the Drawings 10 Fig. 1 shows the kinetics of the reaction of allicin with GSH, monitored by appearance of S-allylmercaptoglutathione (GSSA) (white circles) and disappearance of allicin (black circles). The initial concentration of GSH was 1 .2 mM and that of allicin 0.7 mM. The reaction was carried out at pH 7.0. Fig. 2 shows estimation of concentration of intracellular thiols under 15 incubation, as a function of allicin added to a 1% RBC suspension. Concentration of SH-groups was estimated by ESR with Biradical. Fig. 3 shows the kinetics of appearance of GSSA in RBC at 37*iC after treatment with allicin. Allicin-treated RBC were washed with cysteine and PBS to remove external allicin, then treated with trichloroacetic acid. The concentration of 20 GSSA was assayed in the supernatant. Fig. 4 shows GSSA formation rate in the allicin/GSH reaction mixture (1 mM GSH incubated with 0.7, 1.0 or 1.3 mM allicin at pH 5.0, at room temperature) as function of allicin concentration. The initial amount of GSSA produced was followed by HPLC analysis at different time intervals after dilution 25 with HPLC acidic running buffer. Fig. 5 shows the kinetics of formation of GSSA in the allicin/GSH reaction mixture (0.25 mM GSH incubated with 1.068 mM allicin) at pH 6.0 and 7.0. At different time intervals, aliquots were diluted with 0.1% formic acid in 60%/o methanol, and assayed by HPLC. 30 Fig. 6 shows GSSA formation rate in the allicin/GSH reaction mixture (0.25 mM GSH incubated with 1 mM allicin at room temperature) as a function of pH. 4 WO 01/36450 PCTIL00/00761 Each column represents the mean ± S.D. of three determinations done at the same pH. Fig. 7 shows papain inhibition by GSSA. The residual activity of the inhibited enzyme was assayed at pH 6.5. Activity is expressed as % of enzyme 5 activity of non-inhibited papain. Figs. 8A-8B show ESR spectra of spin-adduct of DMPO with OH radical formed in the Fenton system, in the absence of an antioxidant (Fig. 8A) and in the presence of 1 mM GSSA (Fig. 8B). Fig. 9 shows that allicin, CSSA and GSSA inhibit the appearance of -OH 10 radicals and thus decrease the formation of DMPO- -OH spin adduct. Fig. 10 shows lipid peroxides (LPO) production (%) in fetal rat brain slices during 30 min incubation in presence of 0.1 mM of alliin, allicin, GSSA, GSH or vitamin E, as determined by the thiobarbituric acid assay (production of thiobarbituric acid reactive substance - TBARS). 15 Fig. 11 shows the antioxidant effect of alliin, allicin, GSSA and vitamin E on LPO production by fetal rat brain slices, determined as in Fig. 10 above, as function of reagent concentration. Fig. 12 shows the effect of concentrations of allicin (squares) and GSSA (diamonds) on the proliferation of MCF-7 human mammary cancer cells. 20 Measurements were done after three days of treatment at various concentrations. Fig. 13 shows the effect of 32 ptM of GSSA, 32 tM of allicin or medium only, on the proliferation of MCF-7 cells. Detailed Description of the Invention 25 We have recently shown that the beneficial effects of allicin on health may stem from its interaction with SH-containing molecules such as enzymes and/or its high antioxidant activity [Ankri et al., 1997; Rabinkov et al., 1998; Miron et al., 1998]. Free thiol-containing compounds are usually intracellular, since extracellular thiols are in most cases oxidized to disulfide by the environmental 30 molecular oxygen. Allicin rapidly interacts with free thiols abundantly present in tissues such as GSH, the major intracellular thiol. The high SH-modifying 5 WO 01/36450 PCT/ILOO/00761 reactivity of allicin can also be the main reason for its fast disappearance from the blood. Therefore, the manner in which allicin can penetrate cell membranes and interact with intracellular thiols was investigated according to the present 5 invention. Low molecular weight thiol-containing compounds such as GSH and 2 nitro-5-thiobenzoate (NTB) were entrapped in lipid vesicles and red blood cells were used as natural containers of GSH. Thus the interaction of GSH and NTB with externally added allicin could be studied. Based on the kinetics of this reaction, it was clearly shown that the interaction of allicin with the low molecular 10 weight thiols entrapped in the vesicles was not dependent on the release of GSH and NTB from vesicles, since the reaction product, GSSA, was detected in the vesicle fraction after gel filtration at the end of the reaction. It was further observed that allicin freely permeates through phospholipid bilayers and interacts with the SH groups. The reaction rate of allicin with the SH 15 containing molecules after crossing the membrane was the same as in solution. Fast diffusion and permeation of allicin across human red blood cell membranes was also demonstrated. The high permeability of allicin through membranes may greatly enhance the intracellular interaction with thiols. According to the present invention, GSSH was synthesized for the first time 20 and its SH-modifying and antioxidant properties were demonstrated. GSSA is prepared by reaction between allicin and GSH. This reaction is rather fast with K of bimolecular reaction apparently 3.0 M-Isec-l. It is pH dependent revealing strong dependence on real concentration of GS-, indicating that products of allicin transformation which include a thioallyl radical possess 25 some of allicin activities due to its SH-modifying effect. Study of the antioxidant effect of GSSA in vitro using the Fenton system revealed that GSSA has high ability to prevent formation of OH radicals. In experiments on fetal rat brain slices, the Fe2+-induced TBARS production assay was utilized as an indicator of oxidative stress in the brain. The study of the effect 30 of allicin and GSSA on the capacity of fetal brain to generate peroxides, both under normal and after ischemia-induced stress conditions, showed that addition of 6 WO 01/36450 PCT/ILOO/00761 0.1 mM GSSA was sufficient to reduce TBARS production by about 40%, while 0.1 mM of either vitamin E or allicin exhibited similar inhibition effect (about 50 %) against iron-induced lipid peroxidation in fetal brain slices. This indicates that the SH-modifying activity of GSSA may be attributed to antioxidant capacity as 5 well and that GSSA may be clinically useful as potential antioxidant. The invention further includes pharmaceutical compositions comprising GSSA and a pharmaceutically acceptable carrier. For this purpose GSSA can be incorporated in conventional, solid and liquid pharmaceutical formulations (e.g. tablets, capsules, caplets, injectable and orally administerable solutions) for use ir 10 treating mammals, including humans, suffering from several disorders including atherosclerosis, coronary artery diseases, thrombosis, high levels of cholesterol and blood lipids, high blood pressure, control of weight, Alzheimer disease, glaucoma, cancer and inflammatory disorders such as colitis. In principle, the GSSA compositions can be used for all uses known for allicin, excepting as antibacterial. 15 In a preferred embodiment, the pharmaceutical composition of the invention is in a form for oral administration, for example as an aqueous solution. The invention will now be illustrated by the following non-limiting Examples. 20 EXAMPLES Materials and Methods (a) Materials Reduced glutathione (GSH), L-cysteine, DTT, N-a-Benzoyl-DL-arginine p nitroanilide (BAPNA), 5,5'-dimethyl-1-pyrroline N-oxide (DMPO), FeSO4. 25 (7H20), a-Tocopherol (Vitamin E), oxidized glutathione (GSSG), dimyristoyl phosphatidylcholine (DMPC), calcein and 5,5'-dithiobis-(2-nitrobenzoic acid) (DTNB) were purchased from Sigma Chemical Co. (St. Louis, MO, USA). Diallyl disulfide (DAD) was purchased from Aldrich. Papain (EC 3.4.22.2) was obtained from Boehringer Mannheim (Germany). Alcohol dehydrogenase frori 30 Thermoanaerobium brockii (TBAD) (EC 1.1.1.2) was the kind gift of Dr. M. Peretz and Dr. Y. Burstein, Weizmann Institute of Science, Rehovot, Israel. Egg 7 WO 01/36450 PCT/ILOO/00761 phosphatidylcholine (PC) was purchased from Lipid Products (South Nutfield, UK). Cholesterol (extra pure) was from Merck (Darmstadt, Germany). Symmetrical stable nitroxyl biradical containing disulfide bond, bis (2,2,5,5 tetramethyl-3-imidazoline-1-oxyl-4-yl) disulfide (Biradical) synthesized according 5 to [Kitz and Wilson, 1962] was a kind gift of Dr.V. Martin (Lipitek Int. Inc, San Antonio, Texas, USA). All other reagents were of analytical grade. (b) Synthesis of alliin, allicin, S-allylmercaptocystein (CSSA), and NTB Alliin was synthesized as described in Rabinkov et al., 1998. Allicin was 10 produced by applying synthetic alliin onto immobilized alliinase as described in published PCT Patent Application No. WO 97/39115. CSSA was synthesized from cysteine and allicin and isolated as described in Rabinkov et al., 1998. NTB was prepared according to Degani and Patchornik, 1971. 15 (c) Separation of alliin, GSSA and allicin by HPLC Quantitative determinations of alliin, GSSA and allicin, were performed in an LKB HPLC system with the SP 4290 integrator (Spectraphysics). The separation was achieved on a LiChrosorb RP-18 (7 im) column using 60% methanol in water containing 0.1 % formic acid as an eluant. Flow rate was 0.56 20 ml/min. (d) Assay of papain activity. Papain activation was carried out by dilution (1:10) of papain suspension with 50 mM Na acetate, 2 mM EDTA, pH 6.1 (Na acetate/EDTA buffer) in 25 presence of 2.5 mM dithiothreitol (DTT) for 30 min at room temperature. Excess of DTT was removed by gel filtration on Sephadex G-25 pre-equilibrated with Na acetate/EDTA buffer. The activity of papain was determined at room temperature by following the hydrolysis of BAPNA at pH 6.1, in room temperature at 382 nrm [Angelides and Fink, 1979]. 30 8 WO 01/36450 PCT/ILOO/00761 (e) Assay of TBAD activity TBAD activity was assayed at room temperature by following the fon-nation rate of NADPH from NADP+ at 340 rn (E 340 = 6.225 mM- 1 cm -1) [Peretz and Burstein, 1989] 5 (f) Inhibitory analysis Kinetic analysis of irreversible inactivation of the SH-containing enzymes papain and TBAD were calculated according to Kitz and Wilson, 1962. 10 (g) Determination of rate constants The rate constant for the reaction of allicin (0.7-1.3 mM) and GSH (1 mM) was obtained in 10 mM phosphate/citrate buffer (pH 5.0) or in 0.1 M NaCl, 0.0 1 M Na phosphate buffer (pH 7.0). The initial rate of GSSA appearance was monitored by HPLC. At various time intervals, samples were diluted with HPLC running 15 buffer and analyzed by HPLC. For calculation the following equation was used: d[GSSA]/dt = K[Allicin]x[GSH] where: d[GSSA]/dt is the initial rate of GSSA appearance; K is the bimolecular rate constant; [Allicin], [GSH] are the initial concentrations of allicin and GSH. 20 (h) Preparation of phospholipid vesicles Small unilamellar phospholipid vesicles (SUV) were made either from egg phosphatidylcholine (PC) or dimyristoylphosphatidylcholine (DMPC) and cholesterol. In the case of SUV made from PC, the concentration of cholesterol was 17% mol/mol. In the case of SUV made from DMPC, cholesterol content was 25 25%. mol/mol. Cholesterol was dissolved in CHCl 3 /MeOH (2:1 v/v). The cholesterol solution was added to the dry lipid. A film of 2.5-5 mg lipid/glass vial was prepared by evaporating the solvents under a stream of nitrogen, followed by 3 h drying under high vacuum. SUV containing GSH were obtained by sonication according to Shin et al., 1996, as follows: GSH solution (0.2 M in 0.05 M NaCI, 30 pH 7.0) was added to the dried film, mixed by vortex and sonicated for 10-30 min 9 WO 01/36450 PCT/ILOO/00761 in a bath-type sonicator (GI125SP1, Laboratory Supplies Co. Inc., New York, NY). Final phospholipid concentration was either 5 or 10 mg/ml. (i) Gel filtration 5 Separation of SUV from free reagents was done by gel filtration on Sephadex G-50 (15 ml bed volume) pre-equilibrated with 0.1 M NaCl in 0.01 M Na phosphate buffer (pH 7.0). Vesicles were eluted in the void volume and were monitored by absorption at 340 nm. 10 (j) Preparation of red blood cells (RBC) Fresh human blood (in the presence of heparin) was washed three times with PBS. RBCs were collected after centrifugation (2000 rpm, 5 min) and resuspended in PBS. A 10% or 1% (vol/vol) suspension of RBC in PBS was used in the experiments. 15 (k) Electron spin resonance (ESR) experiments Measurements were performed in a flat cell of the Bruker ER-200 D-SRC spectrometer. The experimental conditions included the following: field, 3500 G; sweep width, 100 G; receiver gain, 2x10 5 ; microwave power, 20 mW; modulation 20 amplitude, 0.8 G. DMPO was purified as described in Buettner and Oberley, 1978. In experiments of hydroxyl radical generation, the sample contained H202 (1 mM), Fell(EDTA)2 (0.8 mM ), DMPO (100 mM), in 20 mM sodium phosphate buffer (pH 7.4)) and NaCl (0.2 M), (final volume 0.2 ml) The hydroxyl radical formation was measured from the ESR spectrum of spin adduct DMPO--OH 25 radical formed in the Fenton system. (1) NMR Spectroscopy NMR spectra were collected on a Bruker AMX-400 spectrometer. Analysis of allicin and CSSA structures was performed as described in Rabinkov et al., 30 1998. The product of the interaction between allicin and glutathione was isolated 10 WO 01/36450 PCTIL00/00761 and dissolved in deuterated water and solutions of 10 mM were prepared. The pH was adjusted to 6.5 using KOD. ID IH (with water signal suppression) and 13 C spectra were collected at 250 C. Structure analysis of the product obtained was performed as described previously (Miron et al., 1998; Rabinkov et al., 1998). 5 Resonance multiplicities for 13 C were established by acquiring DEPT spectra. For the DEPT sequence, the width of a 13 C 900 pulse was 7ms, that of a 1 H 900 was 12.8 ms, and the (2J)-I delay was set to 3.45 ms. The 2D COSY45 IH-lH shift-correlated spectra was recorded using a data size of 512 (tl)x 2048 (t2) with a spectral width of 1400 Hz. The HMQC spectra 10 were recorded using a pulse sequence (invbtp in the Bruker software) which included the bilinear rotational decoupling (BIRD) pulse to invert the magnetization of protons not coupled to 13 C. The spectra were collected with 2048 (t2) x 256 (t1) data points. Spectral widths of 1400 and 11000 Hz were used in the F2 (l1H1) and F1 ( 13 C) domains, respectively. Data sets were multiplied in both 15 dimensions by a 90'- shifted sine bell or Gaussian transformation function and generally zero-filled to 512 in tl dimension prior to Fourier transformation. The delay DI was set to 3.4 ms while D2 was empirically optimized at 600ms. (M) Assay of antioxidant effect on lipid peroxides production (LPO) 20 The preparation of fetal rat brain slices and the determination of LPO concentration in fetal rat brain slices by the thiobarbituric acid (TBA) method were conducted as described in Glozman and Yavin, 1997. Freshly prepared FeSO4(7H20) was added to induce lipid peroxidation. Slices were routinely incubated for 15 min at 370 C with gentile shaking under permanent oxygen 25 aeration. At the end of the incubation period, medium was separated from tissue by centrifugation at 3500x g for 5 min and 0.5 ml of medium was taken for TBARS content determination. 11 WO 01/36450 PCT/ILOO/00761 Example 1. Reaction of allicin with GSH entrapped in SUV Small unilamellar phospholipid vesicles (SUV) containing GSH were prepared in order to show allicin's propensity to permeate the lipid bilayer 5 membrane. GSH, the main intracellular thiol of mammalian cells, is highly hydrophilic and therefore cannot pass lipid membranes. The reaction between GSH and allicin was followed by HPLC and the product was characterized by 'H and "C NMR (Table 1). The elution peak of the reaction product, GSSA (retention time 5.3 min), is positioned between that of 10 allicin (retention time 8 min) and GSH (retention time 4.6 min) (Table 2). The kinetics of the reaction between GSH and allicin (~2:1 ratio) in water is shown in Fig. 1. Formation of the product and disappearance of allicin were determined at various time intervals. The reaction was also monitored at different concentrations of allicin and GSH and at different temperatures. Bimolecular rate constant values 15 for the reaction of allicin with GSH entrapped in phospholipid vesicles or in solution, calculated according to Reaction 1 above, are presented in Table 3. Allicin reacts with GSH entrapped in liposomes at a high rate. The rate of GSSA formation in the vesicles (Table 3) was only slightly slower than in solution. The vesicle composition, DMPC or PC, shows no significant effect on 20 allicin permeability. A decrease in temperature to less than 23*C, below the phase transition of DMPC/cholesterol vesicles, did not hinder allicin penetration. These findings indicate that allicin permeates very easily through the lipid bilayers. 25 12 WO 01/36450 PCT/ILOO/00761 Table 1. IH and 13 C NMR of GSSA in D20 at 298K 11 12 C 13 S CH 2 CH=

CH

2 H 2 C0 2 H 0 H 'CH2 C 3 4 6! H l 89H H

H

2 N-. -CH 2 - CH 2 - - - C-N -'CH 2 .... CO2H H H 5 No IH (ppm) 13 C (ppm) 1 3.74 (t; 6.6) 55, 121 2 171.19- 172.99 3 2.14 (q; 7.58) 27.244 4 2.52 (q; 3.91) 32.495 5 171.19- 172.79 6 4.72 (dd; 5.4; 4.4) 53.844 7 5.88 (m) 39.822 8 171.19- 172.99 9 3.76 (d; 4.9) 44.372 10 171.19- 172.99 11 3.36 (dd; 7.34; 0.5) 41.924 12 5.88 (m;) 134.502 13 5.2 (m;) 119.904 13 WO 01/36450 PCT/ILOO/0076 1 Table 2. HPLC analysis of the reaction mixture of GSH and CSSA with allicin. Compound Retention time (min.) GSH 4.6 CSSA 5.1 GSSA 5.3 Allicin 8.1 5 Table 3. Bimolecular rate constants for the reaction of allicin with GSH: entrapped in phospholipid vesicles or in solution. [GSH] Allicin Temperature (*C) K (M- s7 10- 3 M 10-3 M Solution 0.22 0.80 4 26.28 ± 1.25 (3) 0.22 0.80 24 48.66 ±2.45 (3) 0.42 0.58 35 88.13 ± 5.32 (3) Vesicles: 0.18 2.90 10 21.29 ±2.20 (2) DMPC/cholesterol 3:1 mol/mol 0.34 1.35 15 23.38 ±2.75 (2) 0.34 1.35 24 33.57 ±4.95 (2) 0.34 1.35 35 78.45 ± 4.13 (2) PC: cholesterol 0.16 1.92 24 30.40 ±2.47 (2) 5:1 mol/mol 0.11 0.68 24 31.28 ±3.15 (2) Reaction was performed in 0.1M NaCI, 0.01M Na phosphate buffer (pH 7.0). 10 The results are the mean value of n independent experiments, where (n) represents the number of experiments, as is indicated in brackets. 14 WO 01/36450 PCT/ILOO/00761 Example 2. Interaction of allicin with red blood cells; end products Red blood cells (RBC) are naturally occurring membrane vesicles in which GSH is present at a concentration of 1-2 mM. The amount of thiols in PBS-washed 5 human RBC was measured either by the noninvasive ESR technique with Biradical developed earlier by Weiner (1995), or directly after trichloroacetic acid (TCA) precipitation of the RBC and assay with DTNB at pH 7.0 (Table 4). The total concentration of thiol-containing compounds estimated by the two independent methods were very similar (about 2 mM) and are in a good agreement with data 10 reported by others (Vina et al., 1995). The advantage of the ESR method is that it enables monitoring of the decrease of thiol concentration in cells after short exposure to allicin. Allicin (up to 2.5 x 10- M final concentration) was added to a 1% washed RBC suspension in an Eppendorf tube. After 1-2 min at room temperature the Biradical was added (10~4 M final concentration) and the content of 15 residual thiol concentration was monitored. As shown in Fig. 2, allicin penetrates through the RBC membrane and reacts with intracellular thiols in a concentration dependent manner For the identification of the reaction products, washed human RBC were treated at 370 C with allicin (5x10- 4 M). At different time intervals, a fresh solution 20 of cysteine (20 mM final concentration) was added to block external allicin, and the suspension was washed three times with PBS. The washed RBC were treated with trichloroacetic acid (final concentration 5%), mixed by vortexing and centrifuged. The supernatant was diluted 1:12.5 with HPLC running buffer and analyzed by HPLC. The main product appeared as GSSA, which was formed 25 immediately after allicin penetration into the RBC (Fig. 3). In this experiment, allicin was not in excess with respect to internal thiols. 30 15 WO 01/36450 PCT/ILOO/00761 Table 4. Concentration of intracellular thiols (R-SH) determined in human RBC Method [R-SH] in human RBC Biradical (ESR) 1.78x10 3 M± 0.20 DTNB * 1.69x10 3 M± 0.34 5 *Assay of RSH was done after protein TCA precipitation of RBC suspension, and neutralization of the supernatant to pH 7.0 with NaOH. The results are means of three independent experiments. Example 3. Synthesis and isolation of S-allvimercaptogglutathione (GSSA) 10 GSSA was synthesized from glutathione (GSH) and allicin using an excess of allicin. GSH (200 mg in 5 ml water, pH 6.0) was added dropwise to allicin (130 mg), dissolved in 50% methanol (2 ml) and kept at room temperature for 2 hours. The product of the reaction was detected by HPLC analysis. Excess of allicin was removed by extraction with ether. The water phase was dried by lyophilization. 15 The product was re-dissolved in water and dried again by lyophilization. The structure of GSSA was confirmed by 'H and ' 3 C NMR in D20 at 298K (Table 1) and MS (not shown). The elution pattern of the reaction product was as in Example 1 above (see Table 2). 20 Example 4. GSSA formation rate in the allicin/GSH mixture as a function of allicin concentration and of pH GSH (1 mM) was incubated with allicin (0.7 mM to 1.3 mM) in 10 mlA citrate phosphate buffer pH 5.0, at room temperature. The initial amount of GSSA produced was followed by HPLC analysis at different time intervals after dilution. 25 with HPLC acidic running buffer. The appearance of the product on HPLC during the reaction enabled us to follow the kinetic of interaction of allicin with GSH. The bimolecular kinetic constant of the reaction at pH 5 was determined at allicin concentrations of 0.7, 1.0 and 1.3 mM. The results are shown in Fig. 4 and in Table 5. 16 WO 01/36450 PCT/ILOO/00761 The kinetics of formation of GSSA in the allicin/GSH mixture was then determined at pH 6.0 and 7.0. Starting concentrations were GSH 0.25 mM and allicin 1.068 mM. At different time intervals, aliquots were diluted with 0.1% formic acid in 60% methanol, and assayed by HPLC. The results are shown in Fig. 5 5. Allicin (1 mM) was incubated with GSH (0.25 mM) in citrate-phosphate (10 mM) buffer at various pH: 4.5, 5, 5.5, 6. 6.5 and 7. The initial rate for each condition was followed at room temperature. The results are shown in Fig. 6. Each column represents the mean ± S.D. of three determinations done in the same pH. 10 As shown in Figs. 5 and 6, the rate of the reaction observed clearly indicates a strong dependence on pH. The pH dependence of the reaction rate indicates that GSH is involved in reaction with allicin in the form of mercaptide ion (GS-). Table 5. Rate of the reaction and kinetic constants of the reaction GSH/allicin 15 at pH 5.0 and at room temperature [allicin] xl0- 3 M v, 10- 6 x M sec-I K, M-I sec-1 0.7 2.00 2.87 1.0 2.92 2.92 1.3 3.50 2.69 17 WO 01/36450 PCT/ILOO/00761 Example 5. SH-modifying activity of GSSA: Papain and TBAD inhibition The SH-modifying activity of GSSA and CSSA was determined in the same enymatic systems as was described earlier for allicin (Rabinkov et al., 1998), 5 namely, in SH-protease papain, where a SH-group (Cys 2 5 ) is located in the active site, and in alcohol dehydrogenase from the thermophylic bacteria Thermoanaerobium brockii (TBAD), where a SH-group (Cys 2 0 3 ) is located very close to the NADP+- binding site (Korkhin et al., 1998). According to X-ray data, the Cys 2 5 of papain is located on the protein surface, on the groove between the 10 two lobes of the protein, and therefore is available for chemical modification. Each TBAD subunit contains one free SH-group on the protein surface. Chemical modification of that group inhibits dramatically the enzymatic activity (Peretz et al., 1997; Rabinkov et al., 1998). To test the inhibition of papain, DTT-activated and gel-filtered papain 15 (papain-SH) was inactivated by GSSA (2.0 mM, 4.0 mM. 9.0 mM) in 50 mM Na acetate buffer pH 6.2 containing 2 mM EDTA. The residual activity of the inhibited enzyme was assayed at pH 6.5 and recorded as described in Material and Methods, section (d). Activity is expressed as % of enzyme activity of non inhibited papain. As shown in Fig. 7, incubation of activated papain with GSSA 20 led to rapid loss of enzymatic activity. The same results are obtained with CSSA (data not shown) as well as with allicin (Rabinkov et al., 1998). The loss of activity was time dependent. The rates of enzyme inactivation depend on the initial disulfide's concentration. As follows from these experiments, CSSA and GSSA, the natural products of allicin transformation, show pronounced SH modifying 25 activity. The kinetics of irreversible inhibition of the enzymes papain and TBAD by allicin, diallyl disulfide (DAD), GSSA and CSSA were evaluated according to Kits and Wilson, 1962. The kinetic constants for SH modification of papain and TBAD are presented in Tables 6 and 7, respectively. Allicin was more reactive, followed 30 by CSSA and GSSA, which was the slowest due to its bulky glutathionyl moiety. 18 WO 01/36450 PCT/ILOO/00761 Table 6. Analysis of irreversible inhibition of papain by allylmercapto derivatives. Inhibitor k3 [min.- 1 ] KI [M] Allicin 1.03 ±0.15 8.59 x10- 5 ± 1.29 x10-5 CSSA 0.69 ±0.14 4.13 x10- 3 ± 0.83 x10-3 DAD 1.03 ± 0.20 2.20 x10- 2 ± 0.33x10-2 GSSA 1.38 ± 0.22 7.58 x10- 3 ± 1.25 x10-3 Each point represents the mean value for n=3. 5 Table 7. Analysis of irreversible inhibition of TBAD by allylmercapto derivatives. Inhibitor k3 [min.- 1 KI [M] Allicin 0.405 ± 0.15 1.30 x10- 3 ± 0.30 x10-3 CSSA 0.295 ± 0.08 0.83 x10- 3 ± 0.22 x10-3 GSSA 0.09 ± 0.04 0.91 x10- 3 ± 0.15 x10-3 Each point represents the mean value for n=3. 10 Example 6. Antioxidant activity of GSSA in vitro The antioxidant properties of GSSA, CSSA and allicin were determined using the Fenton system as a source of hydroxyl radicals (Lloyd et al., 1997).

H

2 0 2 + Fe(II) - -OH + ~OH + Fe(III) 15 The rates of appearance and amount of hydroxyl radicals were determined with the spin trap DMPO that interacts with OH radicals at a high rate. The bimolecular rate constant for this interaction is 2x10 9

M-

1 sec- 1 (Finkelstein et al., 1980). We observed a spin adduct with a characteristic ESR spectrum: a quartet with intensity ratio 1:2:2:1 and hyperfine splitting constants aN = 14.9 G 19 WO 01/36450 PCT/ILOO/0076 I (Fig. 8A). This is a typical spectrum for a spin adduct of DMPO with -OH radical (Buettner, 1987). As shown in Fig. 8B and Fig. 9, the addition of GSSA, CSSA or allicin to the samples decreased significantly the formation of DMPO-OH spin adduct, 5 indicating a pronounced antioxidant activity of all three compounds. These data thus demonstrate that both GSSA and CSSA have high antioxidant properties. This mechanism explains the antioxidant activity of allicin even after its fast disappearance in 'real' biological systems. 10 Example 7. Effect of GSSA on LPO production by fetal rat brain slices er vivo The antioxidant activity of GSSA was investigated in tissues using fetal rat brain slices model ex vivo under iron-induced oxidative stress by determination of the lipid peroxides level as described in Materials and Methods, section (in). The 15 effect of GSSA was compared with activities of alliin, allicin, GSH, GSSG, 2 deoxyribose and vitamin E. LPO production (%) in fetal rat brain slices was determined during 30 min incubation in presence of 0.1 mM of the various reagents by the TBA assay as described in Materials and Methods, section (m). As shown in Fig. 10, no significant antioxidant activity was observed after incubation 20 of brain slices with alliin. In addition, no antioxidant activity was observed after 2 deoxyribose or GSSG treatments (data not shown). In contrast, significant decrease of LPO production (p<0.05) was demonstrated under GSSA treatment. This antioxidant activity of GSSA is similar to the effects of GSH, about 30% decrease of TBARS production. Among all tested compounds, the most prominent 25 inhibition of TBARS production (about 60%) was observed after allicin or vitamin E treatment. The concentration curves for the antioxidant effect of GSSA, allicin, alliin and vitamin E on LPO production by fetal rat brain slices were examined. As shown in Fig. 11, the antioxidant effects of GSSA, allicin and vitamin E are dose 30 dependent, whereas alliin does not demonstrate any antioxidant activity at all. However, the rate of inhibition of LPO production by allicin is similar to that of 20 WO 01/36450 PCT/ILOO/00761 vitamin E, indicating that the lipid permeability of both is a factor responsible for inhibition of LPO production in this ex vivo model. Example 8. Antiproliferative activity of GSSA 5 Cell culture and cell proliferation assay MCF-7 human mammary cancer cells were grown in DMEM mediurn containing 0.6 pg/ml insulin. The medium was supplemented with penicillin (100 U/ml), streptomycin (0.1 mg/ml) nystatin (12.5 pg/ml), and 10% FCS. Cells were seeded into 6-multiwell plates (170,000 cells per well) or 96-multiwell plates 10 (5,000 cells per well) or 100 mm dish (1,500,00 cells per dish) in mediurn containing 3% FCS. One day later the medium was changed to one containing either only medium or allicin or GSSA, and the media were replaced daily. At the end of the incubation, 1.25 pCi/well of [3H] thymidine (specific radioactivity 5 iCi/mmol) was added for 1 hour. The nucleotide incorporation was stopped by 15 adding unlabeled thymidine (0.5 pmol). The cells were then trypsinized and collected on a glass-fiber filter using a cell harvester (Inotech, Switzerland). Radioactivity was determined by a radioactive image analyzer (BAS 1000, Fuji, Japan). As shown in Fig. 12, inhibition of proliferation was observed after treating 20 the MCF-7 cells with either allicin (squares) or GSSA (diamonds). This inhibition was dose dependent. About 50% inhibition of proliferation was observed for both allicin and GSSA at 32 pM. When the MCF-7 human mammary cancer cells were treated with either allicin or GSSA at 32 pM, for several days, the observed inhibition of cell proliferation continuously increased and reached a plateau at 25 about 50% inhibition after 3 days of treatment (Fig. 13). No significant differences were found between allicin and GSSA. References 1. Angelides, K.J. and Fink, A.L. (1979) Mechanism of thiol protease 30 catalysis: detection and stabilization of tetrahedral intermediate in papain catalysis, Biochemistry, 18: 2363-2369. 21 WO 01/36450 PCT/ILOO/00761 2. Ankri, S., Miron, T., Rabinkov, A., Wilchek, M. and Mirelman, D. (1997) Allicin from garlic strongly inhibits cysteine proteinases and cytopathic effects of Entamoeba histolytica., Antimicrob Agents Chemother., 41: 2286-2288. 3. Buettner, G.R. and Oberley, L.W. (1978) Considerations in the spin 5 trapping of superoxide and hydroxyl radical in aqueous systems using 5, 5 dimethyl-1-pyrroline- 1-oxide, Biochem. Biophys. Res. Commun., 83: 69-74. 4. Buettner, G.R. (1987) Spin trapping: ESR parameters of spin adducts, Free Radic. Biol. Med., 3: 259-303. 5. Cavallito, C.J.and Bailey, J.H. (1944) Allicin, the antibacterial principle of 10 Allium sativum. II. Isolation, physical properties and antibacterial action, J. Am. Chem. Soc., 66: 1950-1951. 6. Chu, T.C., Ogidigben, M., Han, G. and DE, Potter. (1993) Allicin-induced hypotension in rabbit eyes, J Ocul Pharmacol, 9: 201-209. 7. Chu, T.C., Han, P., Han, G. and DE, Potter. (1999) Intraocular pressure 15 lowering by S-allylmercaptocysteine in rabbits., J Ocul Pharmacol Ther, 15(1): 9 17. 8. Degani, Y. and Patchomik, A. (1971) Selective cyanylation of sulfhydryl groups. II On the synthesis of 2-nitro-5-thiocyanatobenzoic acid, J. Org. Chem., 36: 2727-2728. 20 9. Dirsch, V.M., Kiemer, A.K., Wagner, H. and Vollmar, A.M. (1998) Effect of Allicin and Ajoene, 2 Compounds of Garlic, on Inducible Nitric-Oxide Synthase, Atherosclerosis, 139: 333-339. 10. Feldberg, R.S., Chang, S.C., Kotik, A.N., Nadler, M., Neuwirth, Z., Sundstrom, D.C. and Thompson, N.H. (1988) In vitro mechanism of inhibition of 25 bacterial cell growth by allicin, Antimicrob. Agents Chemother., 32: 1763-1768. 11. Finkelstein, E., Rosen, G.M. and Rauckman, E.J. (1980) Spin trapping of superoxide and hydroxyl radical: practical aspects, Arch. Bioch. Biophys., 200: 1 16. 12. Freeman, F. and Kodera, Y. (1995) Garlic Chemistry - Stability of S-(2 30 Propenyl) 2-Propene-1-Sulfinothioate (Allicin) in Blood, Solvents, and Simulated Physiological Fluids., J. Agricultural and Food Chem., 43: 2332-2338. 22 WO 01/36450 PCT/ILOO/00761 13. Glozman, S. and Yavin, E. (1997) Lipid peroxides are generated by the fetal rat brain after episodes of global ischemia in utero, Neurochemical Research, 22: 201-208. 14. Imai, J., Ide, N., Nagae, S., Moriguchi, T., Matsuura, H. and Itakura, Y. 5 (1994) Antioxidant and radical scavenging effects of aged garlic extract and its constituents., Planta Med., 60(5): 417-420. 15. Kitz, R. and Wilson, I.B. (1962) Esters of methanesulfonic acid as irreversible inhibitors of acetylcholinesterase, J. Biol. Chem., 237: 3245-3249. 16. Koch, H.P. and Lawson, L.D. (1996) Garlic: The science and therapeutic 10 application of Allium sativum L. and related species, 2nd edn., Williams & Wilkins, Baltimore, MD, USA. 17. Korkhin, Y., Kalb(Gilboa), A.J., Peretz, M., Bogin, 0., Burstein, Y. and Frolow, F. (1998) NADP-dependent Bacterial Alcohol Dehydrogenases: Crystal structure, cofactor-binding and cofactor specificity of the ADHs of Clostridium 15 beijerinckii and Thermoanaerobacter brockii, J. Mol. Biol., 278: 967-981. 18. Lawson, L.D. and Wang, Z.J. (1993) Pre-hepatic fate of the organosulfur compounds derived from garlic (Allium sativum), Planta Med. Suppl., 59: A688 A689. 19. Lee, E.S., Steiner, M. and Lin, R. (1994) Thioallyl compounds: potent 20 inhibitors of cell proliferation., Biochim Biophys Acta, 1221(1): 73-77. 20. Lloyd, R.V., Hanna, P.M. and Mason, R.P. (1997) The origin of the hydroxyl radicals in the Fenton reaction, Free Radic. Biol. Med., 22: 885-888. 21. Miron, T., Rabinkov, A., Mirelman, D., Weiner, L. and Wilchek, M. (1998) A spectrophotometric assay for allicin and alliinase (Alliin lyase) activity: reaction 25 of 2-nitro-5-thiobenzoate with thiosulfinates., Anal Biochem., 265: 317-325. 22. Peretz, M. and Burstein, Y. (1989) Amino acid sequence of alcohol dehydrogenase from thenrophilic bacterium Thermoanaerobiurm brockii, Biochemistry, 28: 6549-6555. 23. Peretz, M., Weiner, L.M. and Burstein, Y. (1997) Cysteine reactivity in 30 Thermoanaerobacter brockii alcohol dehydrogenase, Protein Science, 6: 1074 1083. 23 WO 01/36450 PCT/ILOO/00761 24. Pinto, J.T., Qiao, C., Xing, J., Rivlin, R.S., Protomastro, M.L., Weissler, M.L., Tao, Y., Thaler, H. and Heston, W.D. (1997) Effects of garlic thioallyl derivatives on growth, glutathione concentration, and polyamine formation of human prostate carcinoma cells in culture., Am J Clin Nutr, 66(2): 398-405. 5 25. Prasad, K., Laxdal, V.A., Yu, M. and Raney, B.L. (1995) Antioxidant Activity of Allicin, an Active Principle in Garlic., Molecular and Cellular Biochemistry, 148: 183-189. 26. Rabinkov, A., Miron, T., Konstantinovski, L., Wilchek, M., Mirelman, D. and Weiner, L. (1998) The mode of action of allicin: trapping of radicals and 10 interaction with thiol containing proteins, Biochim. Biophys. Acta, 1379: 233 -244. 27. Shin, I., Silman, I. and Weiner, L. (1996) Interaction of partially unfolded forms of Torpedo acetylcholinesterase with liposomes. Protein Sci. 5: 42-51. 28. Sigounas, G., Hooker, J.L., Li, W., Anagnostou, A. and Steiner, M. (1997a) S-allylmercaptocysteine, a stable thioallyl compound, induces apoptosis in 15 erythroleukemia cell lines., Nutr Cancer, 28(2): 153-9. 29. Sigounas, G., Hooker, J., Anagnostou, A. and Steiner, M. (1997b) S allylmercaptocysteine inhibits cell proliferation and reduces the viability of erythroleukemia, breast, and prostate cancer cell lines., Nutr Cancer, 27(2): 186 191. 20 30. Vina, J., Sastre, J., Asensi, M. and Packer, L. (1995) Assay of blood glutathione oxidation during physical exercise, Methods Enzymol., 251: 237-243. 31. Weiner, L.M. (1995) Quantitative determination of thiol groups in low and high molecular weight compounds by electron paramagnetic resonance, Meth. Enzymol., 251: 87-105. 25 32. Wills, E.D. (1956) Enzyme inhibition by allicin, the active principle of garlic, Biochem. J, 63: 514-520. 24

Claims (4)

1. S-Allylmercaptoglutathione of the formula: 5 H 2 NCHCH 2 CH 2 CONHCHCONHCH2COOH COOH CH 2 -S-S-CH,-CH=CH 2 10 and salts thereof.

2. A method for the preparation of S-allylmercaptoglutathione which comprises reacting glutathione with allicin or with diallyl disulfide. 15

3. A pharmaceutical composition comprising S-allylmercaptoglutathione or a salt thereof and a pharmaceutically acceptable carrier.

4. The pharmaceutical compositions according to claim 3 for the treatment of atherosclerosis, coronary artery diseases, thrombosis, high levels of cholesterol 20 and blood lipids, high blood pressure, control of weight, Alzheimer disease, glaucoma, cancer and inflammatory disorders. 25 30 25