Antioxidant activity tests on novel triterpenoids from Salvia macrochlamys

The methanol extract of Salvia macrochlamys Boiss. and Kotschy was fractionated on a silica gel column to yield a group of terpenic compounds. After separation and cleaning, seven known and three new terpenoid compounds were isolated, and their structures were elucidated by spectroscopic methods, including intensive NMR and MS studies. The crude extract was tested in five different systems for antioxidant activity. The extract and monogynol A ( 1 ) and its three derivatives ( 2-4 ) were found to be highly active in a metal chelating test system on ferrous ions.


Introduction
Salvia species have been used in folk medicine since ancient times to cure tuberculosis, 1,2 cancer, 3 diabetes, 4 coronary heart diseases, angina pectoris and myocardial infarction. 5,6Skin diseases such as psoriasis and eczama 7 could be treated by Salvia species and they also exhibit oestrogenic activities. 8In Turkey there are about 90 Salvia species, half of them being endemic, and some species are consumed as tea, especially in rural areas. 9Our group has studied over 50 Salvia species for their chemical contents and their biological activities.
From the whole plant, after extraction and chromatographic separation seven known and three new terpenoids were isolated.The known compounds were germanicol, germanicol acetate, 10 lupeol, lupeol acetate, 11 monogynol A, 12,13 ursolic acid 14 and caryophyllene oxide. 15he HSQC and HMBC experiments also allowed us to determine the methyl groups at δ 0.94 and 1.07 must belong to Me-27 and Me-26, respectively, particularly by the observation of twoand three-bond away correlations between both methyl protons with C-8 (δ 41.60) and C-14 (δ 43.74) in the latter experiment.In addition, methyl protons at δ 0.85 and 0.87 showed three-bond away correlations with the carbon at δ 55.5 (C-5) which led to their assignments as C-23 and C-25 methyls, respectively.The stereochemistry of the hydroxyl group at C-22 was deduced as β, by studying on a Dreiding model, and from the J value of the peak at δ 3.53 (dd, J= 7.80 and 9.75 Hz) which was verified by a NOESY experiment following from NOESY correlations between H-22 and H-18.Thus, the structure of compound 3 was determined as 3β-acetyl,22βhydroxymonogynol A based on mass and intensive NMR spectral data including 1 H, APT, HSQC, HMBC and NOESY experiments.The third new compound was also a derivative of monogynol A and its structure was deduced as 3β-acetyl,21β,22β-dihydroxymonogynol A (4) on the basis of spectral data.The HRMS of 4 exhibited a molecular ion peak at m/z 518.3996 corresponding to the molecular formula C 32 H 54 O 5 which correlated with the 13 C NMR findings consisting of 9 methyl, 8 methylene, 8 methine and 7 quaternary carbons.The 1 H NMR spectrum of compound 4 showed similar skeleton profile to those of 2 and 3 with eight methyl signals at δ 0.80, 0.82, 0.83, 0.84, 0.97, 1.06, 1.24 and 1.34 as singlets.There were three methine signals which should be adjacent to oxygenated substituents, followed from 1 H and 13 C NMR spectra at δ 4.48 (1H, dd, J=5.9, 10.5 Hz), 4.20 (1H, dd, J=9.36, 9.60 Hz) and 3.53 (1H, d, J=9.36 Hz) and corresponding carbons at δ 81.17 (C-3), 79.18 (C-21) and 84.20 (C-22) respectively, determined by HSQC spectrum.Their locations were identified through COSY and HMBC experiments (Fig. 2).The COSY experiment allowed us to observe a COSY correlation between the signals at δ 3.53 and 4.20 indicating that they must be vicinal protons.However, there was no COSY relation between these signals and the signal at δ 4.48 which attached to C-3 proton.This signal gave a COSY correlation with the signal at δ 1.60 (m) indicating that the latter signal should belong to one of the C-2 protons.The signal at δ 1.60 further gave a COSY correlation with its geminal proton at δ 1.90 as well as the signal at δ 1.00 which is assigned to one of the C-1 protons (H-1α).
Observation of direct correlations between C-1 signal at δ 38.68 and proton signals at δ 1.00 and 1.71 (H-1β) verified that both signals belong to C-1 protons.Since C-5, C-6 and C-7 are resonated at exactly the same ppm with those observed in monogynol A, and there was no interaction in the COSY experiment between the signal at δ 4.48 (H-3) and the methine protons at δ 3.53 and 4.20, the latter two protons should take place on other than A and B rings by attaching to the oxygenated carbons.On the other hand, these two protons, therefore their adjacent oxygenated substituents could not be located at ring C or D because of their multiplicities, the only possible place for them was found to be on ring E. The observation of COSY correlation between the signal at δ 4.20 dd (J= 9.36 and 9.60 Hz) and a triplet signal at δ 1.93 (H-19) is attributed to their vicinity, and therefore the location of the former signal at C-21 on ring E. A three-bond away correlation between C-19 (δ 46.22) with Me-29 and Me-30 (δ 1.24 and 1.34) were clearly observed, as well as correlation between C-18 with Me-28 (δ 0.80) by HMBC experiment (Fig. 2).In fact, the location of two secondary hydroxyl groups at C-21 and C-22 were also determined by HMBC experiments following two-, three-and four-bond away correlations.A two-bond away correlation was observed between H-19 with C-21, and a threebond away correlation between proton signal of Me-28 (δ 0.80) with C-22 (δ 84.20), and fourbond away correlation Me-28 with C-21 (δ 79.18) which indicating the location of both secondary hydroxyl groups should be on ring E. Since their vicinity, both H-21 and H-22 showed a two-bond away HMBC correlation with each other carbons.Two-bond correlations were further observed between C-20 signal at δ 74.79 with two methyl singlets at δ 1.24 and 1.34.Assignments of the all methyl groups were made based on their HSQC and HMBC correlations (Fig. 2).Three-bond away correlations were observed between C-3 (δ 81.17) with the methyl signals at δ 0.82 and 0.83, which allowed the assignments of the latter signals to Me-24 and Me-23, respectively.The stereochemistry of the hydroxyl groups located at C-21 and C-22 were deduced from their J values and by studying on a Dreiding model as well as by NOESY experiments.Thus, both methine protons, attached to the secondary hydroxyl groups on ring E, showed a NOESY correlation with each other, as well as with H-18, while no correlation was observed between them with either H-19 or Me-28 protons.The extract of Salvia macrochlamys was tested in five different methods for potential antioxidant activity consisting of free radical scavenging activity by DPPH (Table 2), antioxidant activity by β-caroten-linoleic acid (Fig. 4), superoxide anion radical scavenging activity (Fig. 5) in PMS-NADH system, CUPRAC test for Cu (II) reducing capacity (Fig. 6), and metal chelating activity with ferrozine (Fig. 7).While the extract was not found to be active in the DPPH test system it showed moderate activity, in general (Table 2).Particularly, in inhibition of superoxide anion radical generation, the extract showed better activity than standard BHT and a similar activity to that of ascorbic acid at 12.5-25 µg/mL concentrations, while in inhibition of lipid peroxidation it exhibited less activity, however, there was a linearity with increasing concentration, and inhibition was observed to be the same with that of BHT at 50 µg/mL concentration.It is noteworthy that the plant extract showed high activity in metal chelating system giving better results than those of tested standards TOC, BHT and quercetin (Fig. 7).Therefore, isolated monogynol A (1) and its new derivatives (2-4) were investigated in the same test system for their metal chelating activity, and they were found not as active as the crude extract, however, they showed comparible activity results with those of the standards (Fig. 7).

Antioxidant Activity Tests Determination of the antioxidant activity with the β-carotene bleaching method
The antioxidant activity of S. macrochlamys extract was evaluated by β-carotene-linoleic acid model system. 18,19β-carotene (0.5 mg) in 1 mL of chloroform was added to 25 µL of linoleic acid, and 200 mg of Tween 40 emulsifier mixture.After evaporation of chloroform under vacuum, 100 mL of distilled water saturated with oxygen, were added by vigorous shaking.Four thousand microliters of this mixture were transferred into different test tubes containing different concentrations of the sample.As soon as the emulsion was added to each tube, the zero time absorbance was measured at 470 nm using a spectrophotometer.The emulsion system was incubated for 2 h at 50 0 C. A blank, devoid of β-carotene, was prepared for background subtraction.BHT and α-tocopherol were used as standards (Fig. 4).

Free radical scavenging activity
The free radical scavenging activity of S. macrochlamys extract and triterpenes (1-4) was determined by the DPPH assay described by Blois. 20In its radical form, DPPH absorbs at 517 nm, but upon reduction by an antioxidant or a radical species its absorption decreases.Briefly, 0.1 mM solution of DPPH in methanol was prepared, and 4 mL of this solution was added to 1 mL of sample solution in methanol at different concentrations.30 minutes later, the absorbance was measured at 517 nm.Lower absorbance of the reaction mixture indicates higher free radical scavenging activity (Table 2).The capability to scavenge the DPPH radical was calculated using the following equation. 20,21PH Scavenging Effect (%) = The reaction started by adding 1 mL of PMS solution (10 µM) to the mixture.The reaction mixture was incubated at 25 o C for 5 min, and the absorbance at 560 nm was measured against blank samples.Decreased absorbance of the reaction mixture indicates increased superoxide anion scavenging activity.Results were given as percentage inhibition 21 (Fig. 5).

Cupric reducing antioxidant capacity (CUPRAC)
The cupric reducing antioxidant capacity of the extract and triterpenes (1-2, 4) was determined according to the method of Apak et al. 23 To a test tube, 1 mL each of 10 mM Cu (II), 7.5 mM neocuprine, and NH 4 Ac buffer (1 M, pH 7.0) solutions were added.Extract and triterpenes (1-2, 4) at different concentrations were added to the initial mixture so as to make the final volume 4.1 mL.The tubes were stoppered, and after 1 h, the absorbance at 450 nm was recorded against a reagent blank (Fig. 6).

Metal chelating activity
The chelating activity of S. macrochlamys extract and triterpenes (1-4) on Fe 2+ was measured as reported by Decker, & Welch. 24The extract was added to a solution of 2 mM FeCl 2 (0.1 mL).The reaction was initiated by the addition of 5 mM ferrozine (0.2 mL).The mixture was shaken vigorously and left standing at room temperature for 10 min.After the mixture reached equilibrium, the absorbance was determined at 562 nm, results were given as percentage inhibition 21 (Fig. 7).

Statistical analysis
Experimental results concerning this study were mean ± S.D. of two parallel measurements.Analysis of variance was performed by ANOVA procedures.Significant differences between means were determined by student's-t test, p values <0.05 were regarded as significant, p values <0.01 were regarded as very significant.

Figure 3 .
Figure 3.The mass fragment patterns of the three new compounds 2-4.
Measurement of superoxide anion scavenging activity of S. macrochlamys extract was based on the method described by Liu et al. with slight modification.22Superoxide radicals are generated in PMS-NADH systems by oxidation of NADH and assayed by the reduction of NBT.In this experiment, the superoxide radicals were generated in 3 mL of Tris-HCl buffer (16 mM, pH 8.0) containing 1 mL of NBT (50 µM) solution, 1 mL NADH (78 µM) solution and sample solutions.