Characterization and Quantification of Compounds in the Hydroalcoholic Extract of the Leaves from Terminalia catappa Linn. (Combretaceae) and Their Mutagenic Activity

Terminalia is a genus of Combretaceous plants widely distributed in tropical and subtropical regions. Thus, the aim of this study was to quantify the majority compounds of the hydroalcoholic extract (7 : 3, v/v) of the leaves from T. catappa by HPLC-PDA, chemically characterize by hyphenated techniques (HPLC-ESI-IT-MSn) and NMR, and evaluate its mutagenic activity by the Salmonella/microsome assay on S. typhimurium strains TA98, TA97a, TA100, and TA102. The quantification of analytes was performed using an external calibration standard. Punicalagin is the most abundant polyphenol found in the leaves. The presence of this compound as a mixture of anomers was confirmed using HPLC-PDA and 1H and 13C NMR. Mutagenic activity was observed in strains TA100 and TA97a. As the extract is a complex mixture of punicalagin, its derivatives, and several other compounds, the observed mutagenicity may be explained in part by possible synergistic interaction between the compounds present in the extract. These studies show that mutagenic activity of T. catappa in the Ames test can only be observed when measured at high concentrations. However, considering the mutagenic effects observed for T. catappa, this plant should be used cautiously for medicinal purposes.


Introduction
As part of our efforts to identify the substances responsible for the pharmacological activities of plants utilized in Brazilian popular medicine (Biota-Fapesp Project number 09/52237-9), we have analyzed the hydroalcoholic extract of the leaves from Terminalia catappa L., which belongs to the family Combretaceae. The trivial names of these plants in Brazil include "amendoeira-da-praia, " "cuca, " and "chapéu de sol. " In several Asian countries, physicians have used the leaves, bark, and fruit of T. catappa to treat dermatitis and pyresis [1][2][3][4]. Moreover, the antimicrobial and antifungal [5][6][7][8], antioxidative [9][10][11], anti-inflammatory [12,13], hepatoprotective [3,4], antidiabetic [14], carcinogenesis-preventing 2 Evidence-Based Complementary and Alternative Medicine as esophageal cancer, have been reported to be related to consumption of tannin-rich foods such as betel nuts and herbal teas, suggesting that tannins might be carcinogenic [22]. However, other reports have indicated that the carcinogenic activity of tannins might be related to components that are associated with tannins, rather than the tannins themselves [23,24]. Tannins have also been reported to exert other physiological effects, such as the acceleration of blood clotting, a reduction in blood pressure, a decrease in serum lipid levels, liver necrosis, and modulation of the immune response [21,25].
Ellagitannins have been isolated along with gallotannins from various other plant families and show little chemotaxonomic significance. The tannins punicalagin and punicalin, both of which contain a gallagyl unit and were first isolated from the pericarps of pomegranate (Punica granatum, Punicaceae) [26], were a characteristic component in some Terminalia species. This supports the hypothesis that Punicaceae is chemotaxonomically proximate to Combretaceae as classified by Cronquist [27].
Despite its biological properties, the chemical composition may vary according to geographiclocation [28,29], a fact impairing the standardization of raw material and commercial products for medicinal purposes [30].
Considering that there are no data in the literature on quantification of the majority secondary metabolites of the hydroalcoholic extract of the T. catappa by HPLC-PDA and the popular use of this plant, in this study the punicalagin anomers and ellagic acid were isolated and quantified.
Moreover, the mutagenic effects were evaluated using a Salmonella/microsome assay (Ames test) in the presence and in the absence of in vitro metabolizing systems in the concentrations 22.24, 16.68, 11.12, 5.56, and 2.78 mg/plate.

2.2.
Apparatus. The HPLC system used was a JASCO 2010 HPLC (Jasco, Tokyo, Japan) equipped with PU-2089S Plus pump, a MD-2018 Plus Photodiode Array Detector (PDA), an AS-2055 Plus auto sampler, and column oven (CO-2065 plus). The software ChromNav (Workstation JASCO-ChromNav 1.18.03) was used to control the analytical system and data collection and processing. To purify the substances, a semipreparative JASCO HPLC equipped with two PU-2086 pumps, a high pressure mixer (Model MX-2080-32), a PDA detector (model MD-2018 Plus), and a Rheodyne 500 L manual injector loop was used. The software used to operate the equipment and data processing was ChromNAV version 1 : 18 : 03.
The mass spectrometry experiments were performed on LCQ Fleet equipment (Thermo Scientific) equipped with a dispositive of direct insertion of the sample via flow injection analysis (FIA). The studied matrix was analyzed by electrospray ionization (ESI), and fragmentation in multiple stages (MS 2 , MS 3 , and MS n ) was performed at an ion trap (IT) interface. The negative mode was selected for the generation and analysis of the mass spectra for the first order (MS) and for the remaining multistage experiments under the following conditions: capillary voltage, −25 V; voltage spray, −5 kV; and capillary temperature, 275 ∘ C. A carrier gas (N 2 ) with a flow of 8 arbitrary units (A.U.) was used, and the collision gas was helium (He). The track acquisition was 100-2000 / . Xcalibur version 1.3 software (Thermo Finnigan) was used to acquire and process the data.
The 1 H-and 13 C-NMR experiments were performed using a 300 MHz (11.7 T) nuclear magnetic resonance spectrometer (Varian Inova). CD 3 OD was used as the solvent (Sigma Aldrich).

Extraction.
The infusion and hydroalcoholic extract were prepared according to the experimental protocol established in Biota-Fapesp Project, which this work is linked (process number 09/52237-9).

Preparation of the Infusion of the Leaves from T. catappa.
The leaves (1 g) of T. catappa, after drying, were prepared by infusion as their use in folk medicine [31].

Preparation of the Hydroalcoholic
Extract of the Leaves from T. catappa. The dried and powdered leaves (524.6 g) of T. catappa were extracted by percolation at room temperature with a mixture of EtOH : H 2 O (7 : 3, v/v). The solvents were evaporated to dryness at low pressure to give 33.1 g of the crude hydroalcoholic extract (8.7%). The extract was analyzed by FIA-ESI-IT-MS n and HPLC-PDA.

Identification of the Punicalagin Anomers and Ellagic
Acid in the Hydroalcoholic Extract. Identification of the punicalagin anomers was performed using FIA-ESI-IT-MS n and compared with the literature data [32], 1 H and 13 C NMR [33] and HPLC-PDA. For the FIA-ESI-IT-MS n assay, 10 mg of the crude hydroalcoholic extract was dissolved in 1 mL of MeOH : H 2 O (1 : 1, v/v) after using an ultrasonic bath for 5 minutes. The samples were then filtered through a 0.22 m PTFE filter, and aliquots of 20 L were directly injected into the FIA-ESI-IT-MS n system.
For the HPLC-PDA a clean-up step was performed to remove any contaminants; the solution was purified by solid phase extraction (SPE) using Phenomenex Strata C 18 cartridges (500 mg of stationary phase) that were previously activated with 5 mL of MeOH and equilibrated with 5 mL of MeOH : H 2 O (1 : 1, v/v). The compounds were eluted from cartridges using 1 mL of MeOH : H 2 O (1 : 1, v/v) with a final volume of 5 mL. The samples were then filtered through a 0.22 m PTFE filter and dried. The dry extract was diluted to 10 mg/mL in HPLC solvent. Aliquots of 20 L were injected directly into the HPLC-PDA with detection at 270 nm. The identification of the different compounds in the chromatographic profile of the hydroalcoholic extract was done by comparing their retention times ( ) and UV spectra with those obtained for the pure standards (ellagic acid, Sigma) and the compounds isolated from the hydroalcoholic extract of T. catappa (punicalagin anomers).
For the confirmation of the presence of ellagic acid experiments were done by coinjection. The hydroalcoholic extract (15 mg) was dissolved in 1 mL of MeOH : H 2 O (1 : 1, v/v) and subjected to a clean-up procedure using SPE cartridge RP18 (50 mg). The sample was then filtered through a PTFE filter (0.22 m) and dried at room temperature. The resulting solid was redissolved in MeOH : H 2 O to a concentration of 10 mg/mL. Two milligrams of ellagic acid standard (Sigma) was then added to this solution, and the solution was centrifuged for 2 min and filtered through a 0.22 m filter. A hydro column with a flow rate of 1 mL/min was used with a gradient of 5-60% MeOH for 60 minutes, and the absorbance at 270 nm was detected. The chromatograms of the hydroalcoholic extract were compared before and after the addition of the ellagic acid standard. From the coinjection of ellagic acid and the hydroalcoholic extract, we confirmed the presence of ellagic acid from the significant increase in the peak eluting at 47.30 min (peak 3).

Conditions for Quantification of Punicalagin (Anomers)
and Ellagic Acid. The quantification of analytes was performed using an external calibration standard [34,35]. The curves were constructed using standards of the commercial ellagic acid (Sigma) and the isolated punicalagin anomers. We evaluated the content of each punicalagin anomer and the total content of punicalagins in leaves of T. catappa. A stock solution of 1000 g/mL ellagic acid was prepared, and  Adapted from Seeram et al. [32] Adapted from Seeram et al. [32] Adapted from Seeram et al. [32]     dilutions to 500, 250, 125, 62.5, 15.6, 7.8, and 3.9 g/mL were made. For punicalagin, a stock solution at concentrations of 4000 g/mL of the -anomer (peak 1) and 4000 g/mL of the -anomer (peak 2) was prepared. From the stock solution, serial dilutions were made to reduce the concentration in 2 : 1 increments (2000, 1000, 500, 250, 125, 62.50, and 31.25 g/mL) for both the -and -anomers. Each concentration level was analyzed in triplicate. The mean areas of the chromatographic peaks obtained were interpolated as a function of concentration using linear regression and were used to generate the calibration curves. The linear coefficients (a) and angle (b) were obtained from the calibration curves. The correlation coefficient ( 2 ) and limits of detection (LOD) and quantitation (LOQ) for peaks 1 and 2 corresponding to the -and -anomers of punicalagin [36] were obtained. The accuracy of the HPLC method was estimated from the recovery tests. The recovery tests were performed by adding known concentrations (low, medium, and high) of the reference ellagic acid (15, 60, and 250 g/mL) and isolated punicalagin anomer (250, 500, and 2000 g/mL) standards. The average recovery value was calculated using the following formula: recovery % = [( f / nf + pd )] × 100, where f corresponds to fortified concentration, nf corresponds to the unfortified concentration and pd corresponds to the concentration of the added standard. Repeatability of intra-and interday was done to determine the accuracy of the developed method. The relative standard deviation (RSD) was taken as a measure of accuracy. Repeatability of intra-and interday was determined in six replicates within the range found in the extract in a day and on three consecutive days, respectively.

Salmonella/Microsome Assay.
Mutagenic activity was evaluated using the Salmonella/microsome assay with the Salmonella typhimurium tester strains TA98, TA100, TA97a, and TA102, which were kindly provided by Dr. B. N. Ames (Berkeley, CA, USA), with (+S9) and without (−S9) metabolization using the preincubation method [37]. The strains were grown overnight from frozen cultures for 12-14 h in Oxoid Nutrient Broth number 2. The metabolic activation mixture (S9 fraction), prepared from the livers of Sprague-Dawley rats treated with the polychlorinated biphenyl mixture Aroclor 1254 (500 mg/kg), was purchased from Molecular Toxicology Inc. (Boone, NC, USA) and freshly prepared before each test. The metabolic activation system consisted of 4% S9 fraction, 1% 0.4 M MgCl 2 , 1% 1.65 M KCl, 0.5% 1 M D-glucose-6-phosphate disodium, 4% 0.1 M NADP, 50% 0.2 M phosphate buffer, and 39.5% sterile distilled water. For the determination of the mutagenic activity, five different concentrations of extract (1.56 to 22.24 mg/plate), diluted in DMSO, were assayed. The concentrations of the sample were selected on the basis of a preliminary toxicity test. In all subsequent assays, the upper limit of the dose range tested was either the highest nontoxic dose or the lowest toxic dose determined in the preliminary assay. Toxicity was detected either as a reduction in the number of histidine revertants (His+) or as a thinning of the auxotrophic background (i.e., background lawn).
The various concentrations of extract to be tested were added to either 0.5 mL of 0.2 M phosphate buffer or 0.5 mL of 4% S9 mixture with 0.1 mL of bacterial culture and then incubated at 37 ∘ C for 20-30 min. Next, 2 mL of top agar was added, and the mixture was poured onto a plate containing minimal agar.
The plates were incubated at 37 ∘ C for 48 h, and the His+ revertant colonies were counted manually. All experiments were analyzed in triplicate. The results were analyzed with the statistical software package Salanal 1.0 (U.S. Environmental Protection Agency, Monitoring Systems Laboratory, Las Vegas, NV, from the Research Triangle Institute, RTP, NC, USA), adopting the Bernstein et al. [38] model. The data (revertants/plate) were assessed by analysis of variance (ANOVA) followed by linear regression. The mutagenic index (MI), which is the average number of revertants per plate with the test compound divided by the average number of revertants per plate with the negative (solvent) control, was also calculated for each concentration tested. A sample was considered mutagenic when a dose-response relationship was detected and a twofold increase in the number of mutants (MI ≥ 2) was observed for at least one concentration [39]. The standard mutagens used as positive controls in experiments without the S9 mix were 4-nitro-o-phenylenediamine (10.0 g/plate, TA98), sodium azide (1.25 g/plate, TA100), mitomycin (0.5 g/plate, TA102) and 2-anthramine (1.25 g/plate, TA98, TA100), and 2aminofluorene (10.0 g/plate, TA102) in the presence of S9. DMSO served as the negative (solvent) control.

Results and Discussion
The analysis of mass spectra of the hydroalcoholic extract ( Figure 1) showed the precursor ions [M-H] − / 1083 (punicalagin), / 781 (punicalin), / 601 (gallagic acid), and / 301 (ellagic acid) [32]. Second-order fragmentation (MS 2 ) confirmed the presence of the metabolites shown in the scheme of Figure 2 for substances 1-4 (Table 1). Seeram et al. [32] studying the commercial juice industry of the pomegranate husk identified the same fragmentation of these same compounds. The chromatographic profile by HPLC-PDA of the hydroalcoholic extract and the infusion are shown in Figures  3 and 4. In the chromatogram, three major peaks (1-3) eluting at 14.75 min, 18.80 min, and 47.30 min are observed. Figures 3 and 4 show the UV region of the spectrum obtained using a PDA detector for the peaks eluting at 14.75 min (peak 1) and 18.80 min (peak 2). The data confirm the presence of both punicalagin anomers with absorption maxima at max 218, 260, and 379 nm. The compound eluting at 47.30 min (peak 3) corresponds to ellagic acid with max values of 250, 306, and 368 nm, as confirmed by coinjection experiments with a standard analyte ( Figures  3 and 4). The data obtained for UV anomers punicalagin and ellagic acid are consistent with those found in the literature [40] and are characteristic of gallagyl chromophore [41].
Peaks 1 and 2 ( Figure 3) were partitioned (item 2.4.3), collected separately, and analyzed by HPLC-PDA ( Figure 5). The analysis of the peaks by HPLC-PDA with 14.75 min and 18.80 min, respectively, display an interesting behavior, as they resolve into two peaks with distinct retention times when analyzed under the same conditions to obtain peaks at similar retention times with the same UV spectra. This indicates that the compounds are the mixture of anomers ( -and -punicalagin). These compounds have absorption maxim at max 218, 260, and 379 nm and are characteristic of the chromophore gallagyl [41]. For peak 1, the HPLC-PDA analysis shows the presence of two peaks (1a and 1b) with retention times of 15.29 min and 21.07 min, respectively ( Figure 5).
In solution, punicalagin rapidly interconverts between the -and -anomers (Figure 2, compound 1). The 13 C NMR ( Table 2) spectrum reveals the presence of two anomeric carbons at 89.87 and 93.77, corresponding to punicalagins (present in lesser proportion in peak 1) and (present in a greater proportion in peak 2), respectively. The stereochemistry at the anomeric position can be easily determined by measuring the coupling constant between the anomeric proton H-1 and the adjacent proton H-2 in the 1 H NMR spectrum. A large coupling constant ( H-1,H-2 > 5 Hz) generally indicates a -glycosidic configuration (substituent in the equatorial orientation), whereas a small coupling constant ( H-1,H-2 = 0-5, Hz) indicates an -glycosidic configuration (substituent in the axial orientation). The chemical shift and the coupling constant corresponding to the -anomer is 5.08 ( = 3.5 Hz) and for the -anomer is 4.66 ( = 8.0 Hz). NMR data of the isolated anomers ( -and -punicalagin) are shown in the Table 2 (enumeration of carbons and hydrogens in Figure 2) compared with data found in the literature [33].  (Table 1) and the UV spectra exhibited by -and -punicalagin confirmed this identification ( Figure 5).
The linear regression analysis of the peak area forand -punicalagin, the total punicalagins, and ellagic acid are showed in the Table 3. Thus, it becomes necessary to use sufficient number of standard solutions to adequately define the relationship between concentration and response [42]. These results demonstrate the linearity of the method over the analyzed concentration range.
The recovery percentages for ellagic acid and total punicalagins were found to be 87% and 86%, respectively ( Table 4). The results were presented as mean and standard deviation of the percentages recovered. The results showed that the accuracy of the method is very good for all analytes (RSD < 2%). The relative standard deviations also showed that the process of sample preparation is reproducible [43].
The relative standard deviations (RSDs) for repeatability testing intra-and interday analyses are shown in Table 5. The result is presented as relative standard deviation from the average. The method showed excellent repeatability, with all RSD values lower than 1% [44]. Table 6 shows the mean number of revertants/plate (M), the standard deviation (SD), and the mutagenic index (MI) after the treatments with the T. catappa hydroalcoholic extract, observed in S. typhimurium strains TA98, TA100, TA102, and TA97a, in the presence (+S9) and absence (−S9) of metabolic activation.
In the TA98 strain, the T. catappa hydroalcoholic extract did not induce any increase in the number of revertant colonies under the conditions used in this study. The other strains were more sensitive to the toxic effects of extract, and it was thus necessary to decrease the doses.
Only in the absence of the external metabolizing system, S9 mix, in the S. typhimurium strains TA100 and TA97a, the T. catappa hydroalcoholic extract induced an increase in the number of revertant colonies relative to the negative  control, with a MI higher than 2.0 at the concentration of 3.12 mg/plate in strain TA100 and 9.38 mg/plate in strain TA97a, indicating the direct mutagenic activity for these strains.
The S. typhimurium test strain TA97a detects frameshift mutations in C-C-C-C-C-C; +1 cytosine and TA98 frameshift in DNA target-C-G-C-G-C-G-C-G, the S. typhimurium tester strain TA100 is capable of revealing base-pair substitution point mutations and the S. typhimurium tester strain TA102 is normally used to detect cross-linking agents and base-pair substitution mutations [39]. Thus, according to the strains involved, the T. catappa hydroalcoholic extract mainly induces substitution of base pairs (TA100) and frameshift mutations (TA97a).
Ko et al. [17] evaluated the toxicity and mutagenicity of supercritical carbon dioxide (SC-CO 2 ) extracts of T. catappa leaves at a dose of 0.5 mg/plate in the strain TA100 (detect base-pair substitution) and observed absence of mutagenic effect. The absence of mutagenic effect was also observed for TA102 and TA98 when the ethanolic extract was evaluated.
These studies show that mutagenic activity of T. catappa in the Ames test can only be observed when measured at high concentrations. The results of this study are consistent with a review of the literature indicating that high concentrations of tannins (15 and 60 mM) can cause DNA damage. These   Evidence-Based Complementary and Alternative Medicine data corroborate the observations of Gupta et al. [45], which indicated that all tannins result in polyphenol toxicity at high concentrations [46].

Conclusions
In the hydroalcoholic extract of T. catappa, tannins were the most abundant compounds observed, and the anomersand -punicalagin were the major compounds. Moreover, considering that the extract is a complex mixture of several unknown organic compounds, the mutagenicity observed may be explained in part by a synergy between compounds present in the extract. The dosage and type of tannin involved are critical for these effects. Thus, the results obtained in this study are useful for better understanding the pharmacological activities of T. catappa. However, considering the mutagenic effects observed in this study, this plant should be used cautiously for medicinal purposes.
Based on the obtained results, it can be concluded that the developed method (HPLC-PDA) is suitable for its purpose, namely, the determination of anomers punicalagin and ellagic acid in extract of T. catappa.