Plant samples

Annona muricata leaves were collected from two Mexican states, Nayarit (Tepic, 21°30'0" N, 104°54'0" W) and Chiapas (Cantón el Carmen, 14°46'24" N, 92°13'24" W), in July-August 2020. The selection of the plant material was made considering their healthy phytosanitary characteristics. The leaves were freezedried (Labconco, LYPH Lock 4.5, USA), milled (electric mill, NutriBullet, Los Angeles, Grijalva-Verdugo, Rodríguez-Núñez, Núñez-Colin, Aguirre-Mancilla, Montoya-Anaya, Villareal-Fuentes, Balois-Morales, and Rodríguez-Carrillo: Total polyphenols, antioxidants, and cytotoxic activity of infusions from soursop (finnona muricata) leaves from two Mexican regions USA) and stored (two weeks) in closed plastic bags (Ziploc) at room temperature in the absence of light until analyses.

Infusion preparation

Tea bags for herbal teas were used for the infusion preparations. Each bag of tea, containing 3 g of lyophilized soursop leaves, were infused in 240 ml of boiling water and allowed to steep for 5 min, 10 min, and 15 min; afterwards the infusions were allowed to cool to room temperature and TSP, FC, CT, AC and antioxidant activity were determined (^{Coz-Bolaños et al., 2018}).

Methanolic extraction yields

The polyphenolic compounds from soursop leaves were extracted using solutions of water/methanol (ME), ratio 70:30 and 40:60 (v:v) for total soluble phenols and flavonoids, and absolute methanol was used for tannins. The mixtures were prepared in a ratio 10:1 (m:v), then stirred for 10 min, centrifuged at 5000 rpm for 10 min, and the supernatant were filtered using a Whatman filter paper and the same analyses as for infusions were made (^{Deshpande & Cheryan, 1985}; ^{Singleton et al., 1999}; ^{Dewanto et al, 2002}).

Total soluble phenols

Total soluble phenols (TSP) were determined by spec-trophotometric methods using Folin-Ciocalteu's reagent (Hycel), according to ^{Singleton et al. (1999)} with some modifications as follow: an aliquot (60 μl) of soursop infusion was mixed with 240 of distilled water and 60 of Folin-Ciocalteu reagent were added. The samples were then stirred and left to steep for 6 min. Finally, 600 of 7% (w/v) Na₂CO₃ (Meyer) solution and 480 of distilled water were added and maintained in dark conditions for 90 min at room temperature. Absorbance was read at 750 nm using a Multiskan GO (Thermo Fisher Scientific, 51119200, USA). A gallic acid (Fermont) standard curve (acid gallic = 0.2643(Abs₇₅₀) - 0.006, R²=0.9898) was elaborated using known concentrations (0.200 - 0.020 mg ml^-1). The results were expressed as milligrams of gallic acid equivalents per milliliter (mg GAE ml^-1).

Flavonoid content

The flavonoid content (FC) was determined following the procedure described by ^{Dewanto et al. (2002)} with some modifications: 150 μl of soursop infusion was mixed with 45 nl of 5% (w/v) NaNO₂ (J. T. Baker) solution, 90 μl of recent prepared 10% (w/v) AlCl₃ (J. T. Baker) and 300 of1 M NaOH (J. T. Baker) were added. The mixture was brought to 1.5 ml with distilled water, mixed and left to repose for 5 min. Absorbance was measured against the blank at 510 nm using a Multiskan GO. Catechin (Sigma-Aldrich®, USA) was used as a standard (catechin=0.5803(Abs₅₁₀) + 0.0538, R²=0.9817), and the solutions were prepared at 0.300-0.062 mg ml^-1. The results were expressed as milligrams of cate-chin equivalents per milliliter (mg CE ml^-1).

Condensed tannins

The methodology reported by ^{Deshpande and Cheryan (1985)} was used with some modifications to determine condensed tannins (CT) in the infusions. A total of 200 μl of sample was reacted with 1 ml of a solution prepared in proportion 1:1 (v/v) from 1% vanillin (Sigma-Aldrich®, USA) (w/v, dissolved in methanol) and 8% HCl (J. T. Baker) (v/v, dissolved in methanol) and incubated at a temperature of 30°C for 20 min. Absorbance was read at 500 nm against the blank using a Multiskan GO. To estimate the concentration of tannins, a calibration curve with catechin (Sigma-Aldrich®, USA) (catechin=5.5325(Abs₅₀₀) - 0.1577, R² =0.9866) was obtained using known concentrations of catechin (0.750-0.031 mg ml^-1). Condensed tannin contents were expressed as milligrams of catechin equivalents per milliliter (mg CE ml^-1).

Total anthocyanins

The anthocyanin contents (AC) were studied according to the ^{Abdel-Aal and Hucl (1999)} method with modifications. The infusion was measured and adjusted to pH 1 with 4N HCl. The sample was read in a Multiskan GO at 535 nm against the blank. Total anthocyanin content per sample (mg ml^-1) was calculated as cyanidin 3-glucoside:

where C is the concentration of total anthocyanin (mg ml^-1), A is the absorbance reading, ε is the molar absorptivity (cyanidin 3-glucoside=25,965 cm^-1), vol is the total volume of infusion, and MW is molecular weight of cyanidin 3-glucoside=449.

Antioxidant capacity (ABTS and DPPH)

The method of 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid diammonium salt (ABTS) to evaluate the antioxidant capacity was carried out according to ^{Re et al. (1999)}. The radical ABTS+ was prepared dissolving 3.8 mg of the reactive ABTS (Sigma-Aldrich®, USA) in 1 ml of 2.45 mM potassium persulphate (Fermont), and the mixture was kept in the dark at 4°C for 12 h before use. Then 0.15 ml of the ABTS+ solution was diluted in 14 ml of phosphate buffer at pH 7.4 to obtain an absorbance of0.7+0.020 at 734 nm using a spectrophotometer (Hach DR3900, USA). The phosphate buffer was prepared by adding 8 g NaCl, 0.2 g KCl and 1.44 g KH₂PO₄ in 1 L of distilled water. First, the absorbance at initial time (t₀) was measured in 990 of the diluted radical ABTS+ solution, and 10 of the soursop infusion was immediately added. The mix was kept in the dark at room temperature for 6 min; after that time the absorbance was measured at 734 nm (t₆). The antioxidant capacity was calculated by comparing the absorbance values of the infusion to a Trolox (Sigma-Aldrich®, USA) standard curve (Trolox equivalent=0.0253(Abs(_t0-t6)) + 0.3864, R²=0.9878). The curve was obtained using known concentrations of Trolox (3000-500 mM). The result was expressed as the equivalent Trolox per milliliter (mmol TE ml^-1).

The DPPH antioxidant activity of infusions was determined using ^{Brand-Williams et al. (1995)} with some modifications using 2,2-diphenyl-1-picrylhydrazyl (DPPH) (Sigma-Aldrich^®, USA) as a free radical. A 200 DPPH solution in methanol (Reasol) was prepared. The solution was prepared on the same day of the analysis and the mixture was kept in the dark. A 180 μl aliquot of the DPPH radical solution was mixed with 20 μl of the infusion in wells of a 96-well plate and kept in the dark for 30 min. Absorbance was read at 515 nm in a Multiskan GO. The antioxidant activity was calculated using a Trolox calibration curve (Trolox equivalent= 0.0117(Abs_(t0-t30)) - 0.0286, R²=0.9748). The antioxidant capacity value was expressed as Trolox equivalents per milliliter (mmol TE ml^-1).

FTIR-ATR spectroscopy analysis

Two g of milled leaves from each region were mixed with 40 ml of absolute methanol; the mix was sonicated for 10 min, and was allowed to rest for 24 h, after which it was filtered. This procedure was repeated twice, and the filtrates were mixed, then evaporated using a rotary evaporator (Yamato Scientific, RE201) until a dark brown syrup was obtained. A small amount of ME syrup was analyzed using FTIR (PerkinElmer, Dynascan® brand spectrum 100 model) equipped with an attenuated total reflectance interferometer (ATR) at a temperature of 25±2°C. The spectra were obtained from 16 points with a resolution of 4 cm^-1 in a region of4000-500 cm^-1 (^{Ramírez-Hernández et al., 2020}).

Preparation of leaf extract for cytotoxic activity assay

The procedure of ^{Coria-Téllez et al. (2019)} was used with some modifications; specifically, 630 mg of lyophilized and crushed soursop leaves were mixed with 15 ml of boiling water and the mixture was ground for 10 min, then allowed to steep until reaching room temperature. The mixture was centrifuged at 10,000 rpm for 10 min, and the supernatant was filtered through a 0.45 membrane.

Cytotoxic activity assay

The tetrazolium salt (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) (MTT) assay was used to measure the cytotoxic activity of the extracts against MCF7, and HT-29 cells line according to the procedure described by ^{Coria-Téllez et al. (2019)}. MCF7 cells were cultured in Dulbecco's Modified Eagle Medium (DMEM) (Sigma Aldrich); and the HT-29 cells were cultured in McCoy's Medium (Gibco), both containing 10% (v/v) fetal bovine serum, 1% penicillin-streptomycin antibiotic solution, and 5 ml trypsin. The cells were cultured in a CO₂ incubator with 5% CO₂ at 37°C. The cells were seeded at a density of 1x10⁴ cells per well, incubated for 24 h in darkness and exposed to various concentrations of aqueous concentrated extract (21, 10.5, and 5.25 mg ml^-1). The cells were also treated with 50 of dimethylsulfoxide (DMSO) (negative control). At the end of the treatment, the supernatant was removed in each well and replaced with 150 of MTT (3.33 mg ml^-1) followed by incubation at the same conditions for 4 h. The reaction was stopped by adding 100 DMSO to each well. The absorbance of each solution was read in a microplate spectrophotometer at 595 nm.

Statistical analysis

All experiments were conducted in triplicate and data were presented as the means ± standard deviations. Statistical analysis was performed using Minitab 17 for Windows (Minitab Statistical Software, USA). Significant differences between the samples were tested using analysis of variance (ANOVA) followed by Tukey's test comparison (P<0.05).

Total soluble phenols (TSP)

TSP of samples ranged from 0.182 to 0.246 mg GAE ml^-1 (Tab. 1). TSP in the extracts from Nayarit region correlated with extraction time, the longer infusion time the higher the phenolic content, but not for the Chiapas region. In leaves recollected in Chiapas, at 10 min of extraction the phenolic content decreased; and it increased at 15 min of extraction, even though the infusion from Chiapas reposed for 15 min and had the highest TSP (0.246±0.009). ^{Hardoko et al. (2015)} analyzed soursop leaf samples at different time extractions and reported the highest concentration of TSP at 45 min of extraction (0.321 mg GAE ml^-1), while the lowest concentrations were at 15 min of extraction (0.200 mg GAE ml ^-1). In accordance with the results obtained in this study, ^{Pérez-Burillo et al. (2018)} reported for white tea that increasing the extraction time in the infusions also improved the phenolic extraction (about 10 min).

TABLE 1 Phenolic contents and antioxidant activity of infusions from soursop leaves harvested in two Mexican states. 

Total soluble phenols (TSP); Flavonoid content (FC); Anthocyanin content (AC); Condensed tannins (CT); Antioxidant capacity (ABTS and DPPH). Mean is the average of the three repose times for each region for each phenolic and antioxidant activity. For each column, different lowercase letters indicate significant differences (P<0.05) between times measured by Tukey's multiple range test. For each column, different capital letters indicate significant differences (P<0.05) measured by Tukey's multiple range test.

The increase in TSP with respect to extraction time is related to the polarity and hydrophilicity of the bioactive compounds (^{Kelebek, 2016}). In addition, the water temperature used in the infusions (100°C) influences the extraction efficiency, because the heating process opens the vacuoles of the cells (^{Hardoko et al., 2015}) facilitating the migrations and solubility of polyphenols due to diffusion coefficients of the components to be extracted (^{Saklar et al., 2015}).

Flavonoid content (FC)

For Chiapas, the FC of soursop leaf infusion increased with time, reaching the highest content at 15 min, while Nayarit had the highest concentration at 10 min (0.178 mg CE ml^-1). After this time the content decreased, but it was not significant. Several authors found the same effect for long infusion times. For example, ^{Saklar et al. (2015)} report for green tea elevated catechin concentrations at the beginning of the infusion using water at 75°C and after this time the concentration increased. However, using water at 95°C the catechin content did not change significantly during steep time. Therefore, increasing the steeping time of the infusion beyond 20 min would not result in a greater extraction of flavonoids (^{Saklar et al., 2015}). ^{Ramalho et al. (2013)} report that the catechin flavonoid content decreased after 10 and 15 min for Indian and British brands of black teas, but the Brazilian brand tea continued increasing up to 30 min. The levels of flavonoids in infusions are liable to change depending on the sample preparation conditions (^{Jin et al., 2019}). For example, long infusion times suggest that the flavonoids as catechins can suffer epimerization (the flavonoids are converted to their corresponding non-epi isomers) (^{Saklar et al., 2015}); and the high temperatures used in the extraction process could induce the flavonoid oxidation resulting in damage to the flavonoid compounds (^{Hardoko et al., 2015}).

Anthocyanin content (AC)

The AC in Chiapas infusions increased with the infusion time (Tab. 1), meanwhile, Nayarit infusions did not show the same behavior because at 10 min of infusion the AC decreased and increased at 15 min. The AC of infusions for the two regions ranged from 0.148 to 0.330 mg C3GE ml^-1, and Nayarit showed the highest content of AC at the three-repose time. A wide range of AC values has been reported for different infusion types, ranging from 0.126 to 1.645 mg C3GE ml^-1 for Hibiscus sabdariffa infusion (^{Salmerón-Ruiz et al., 2019}), 0.119 mg C3GE ml^-1 for Hibiscus sabdariffa mixed with bilberry tea that contained berry fruits of aronia, black currant, rose hip, raspberry, bilberry, and blueberry (^{Bratu et al., 2018}). In this sense, the variables that have the greatest impact on anthocyanin extraction are the number of leaves, volume of water, water temperature, and repose time used in the infusion preparation. Therefore, fewer leaves and a lower proportion of water allow for greater extraction efficiency, but it may take longer (^{Talib et al., 2020}). In addition, the use of high temperature water increases the extraction process that improves both anthocyanin solubility and diffusion coefficients (^{Paraíso et al., 2021}). However, high temperatures may not be suitable for anthocyanins, since they are susceptible to thermal degradations and could cause them to break down and lose their antioxidant activity and, consequently, their bioactivity (^{Salmerón-Ruiz et al, 2019}).

Condensed tannins (CT)

The CT content was increased according to the infusion time (Tab. 1) that ranged from 0.131 mg to 0.219 mg CE ml^-1. Previous research reports the concentrations ranging from 0.199 to 0.519 mg CE ml^-1 (^{Hardoko et al., 2015}; ^{Hardoko et al., 2018}). The variations in the concentrations reported in the literature can be attributed to the maturity and treatment of leaves and the temperature and time used for drying (^{Jimenez-Garcia et al., 2020}). The time, water temperature, and type of extraction are also related to extraction conditions. This may explain the prevention of tannin oxidation by increasing the concentration of these compounds in the extract (^{Castiglioni et al., 2015}; ^{Silva-Ramírez et al., 2020}).

Antioxidant activity

The results of ABTS assays show that the infusions of Chiapas have the higher antioxidant activity (1.47 mmol TE ml^-1) (Tab. 1). The infusion time of 5 min has the lowest antioxidant activity with a significant increase only after 15 min from Chiapas. This behavior is similar to that observed for TSP, AC, and CT. The ABTS results obtained in this study are lower than those obtained by ^{Kelebek (2016)} for black tea (1.96 mmol TE ml^-1). ^{Almajano et al. (2008)} report that the antioxidant activity of peppermint infusion and red tea was equal to 0.315 and 0.825 mmol TE ml^-1, respectively. This indicates that the antioxidant capacity of the soursop infusion is similar to that of the commercial teas.

The DPPH assays found that Nayarit leaves infused for 15 min produce the highest DPPH values (0.492 mmol TE ml^-1). There were significant statistical differences (P<0.05) in antioxidant activity depending on infusion time for Nayarit, but not for Chiapas. The behavior of antioxidant activity with respect to time of extraction for Chiapas and Nayarit was similar to TSP and CT behavior. ^{Jin et al. (2019)} point out the influence of temperature and time of extraction on the antioxidant activity through DPPH method, and they report a maximum at 10 min and 95°C, while after 30 min the activity decreased; however, at 65°C after 10 min the value increased significantly.

It is important to note that no one method determines all the antioxidant compounds of a sample. For this reason, in this study, two methodologies were used to evaluate the antioxidant activity of the infusions obtained. The difference in the results between the ABTS and DPPH was explained by the different compounds with antioxidant activity in the samples. These compounds have different reaction mechanisms and kinetics, due to their chemical structure and functional groups. However, both methods show a decrease in absorbance due to the reduction of radicals. On the one hand, the cationic form of the ABTS radical is reduced by the interaction with hydrogen or electron donor species, while the DPPH radical is reduced in the presence of hydrogen donor antioxidants (^{Brand-Williams et al., 1995}; ^{Re et al., 1999}; ^{Grijalva-Verdugo et al., 2018}; ^{Guzmán-Maldonado et al., 2020}).

The high antioxidant activity of the infusions is due to the presence of phenolic compounds and secondary metabolites, such as alkaloids, vitamins, terpenoids, saponins, and essential oils produced for the high metabolic activity of the leaves; and all these compounds in the infusions of soursop leaves play an important role over their antioxidant activity (^{Kelebek et al., 2016}; ^{Menezes et al., 2019}; ^{Roduan et al., 2019}; ^{Balderrama-Carmona et al., 2020}; ^{Mannino et al., 2020}; ^{Silva-Ramírez et al., 2020}).

The reports about the antioxidant activity shows variations between the different authors, because the results are reported using different units (mg L^-1 or μmol L^-1). However, the use of these practices makes it difficult to understand the real contribution in terms of consumption by portion size (^{Urías et al., 2020}). In this research, the consumption of a portion (240 ml) of soursop leaf infusion contributes to antioxidant activity of 229 mM TE per portion. The same activity is proportioned by drinking the same portion of apple juice. Meanwhile, the same portion of red wine gives 1.8 times more antioxidant activity than soursop leaf infusion (^{Park et al., 2018}).

Methanolic extraction yields

Table 2 shows the polyphenolic contents and antioxidant activity of methanolic extract of soursop leaves. The results of methanolic extracts did not show significant differences between regions in the variables tested. However, the infusions showed significant differences (Tab. 1) for TSP, FC, AC, CT and DPPH, where a higher content of FC, AC, CT and DPPH in the infusions was also observed with Nayarit leaves, while TSP was higher for infusions with Chiapas leaves. The ABTS assays of the infusions did not show significant differences between Chiapas and Nayarit. Chiapas and Nayarit have different climatic and geographical conditions (temperatures, air humidity, and soil composition), inducing differences in the biochemical and physiological responses of the plants (^{Syed-Najmuddin et al., 2017}; ^{Ovando-Domínguez et al., 2019}).

TABLE 2 Phenolic contents and antioxidant activity of methanolic extracts from soursop leaves harvested in two Mexican states. 

Total soluble phenols (TSP); Flavonoid content (FC); Anthocyanin content (AC); Condensed tannins (CT); Antioxidant capacity (ABTS and DPPH). The results are the mean ± standard deviations of n=3. For each column, different letters indicate significant differences at P<0.05 as measured by Tukey's multiple range test.

The infusion extraction yields for Chiapas at 15 min showed values of 91.7%, 70%, and 64.5% for ME, TSP, FC and CT, respectively. The yield for the Nayarit region showed results of 34.7% and 23.6% for AC and DPPH, respectively. ^{Coz-Bolaños et al. (2018)} report yields of 62.4% and 40% for TSP and FC for moringa infusions. ^{Balderrama-Carmona et al. (2020)} point out yields of 71.97% and 12.57% of antioxidant activity of ABTS and DPPH in soursop leaf infusions, compared to the yield of the extractions with acidified alcohol. Also, Nam et al. (2007) report yields of 20% in extractions for TSP in soursop leaves using water, in comparison to the methanol extractions. An accurate comparison could be difficult between the results previously reported by different authors. Due to a lack of uniformity in the conditions of infusion preparation and the properties of the leaves such as leaf age, leaf size and harvesting season, it is difficult to compare accurately the results obtained in the present work with other studies (^{Saklar et al., 2015}).

FTIR-ATR analysis of A. muricata extracts

The FTIR spectrometry is a physico-chemical analytical technique that reveals the functional groups of the components separated based on their peak ratio. Figure 1 provides the functional groups present in A. muricata leaves extract.

FIGURE 1 FTIR spectra in the region of 4000-850 cm-1 of A. muricata leaves extracts from Chiapas and Nayarit. 

In the FTIR, the broad and strong band between 3700 and 3000 cm^-1 is due to overlapping stretches of vibrations of functional - OH groups, showing the presence of phenolic compounds and the methanol residual in the extracts (^{Daud et al., 2016}). The peak at 2,924 cm^-1 is due to the carboxylic acid group (Ibrahim et al., 2021). The peaks observed at 2,853 cm^-1 are assigned to asymmetric stretching of the -CH₃ groups of the lactone rings, meanwhile the 1,740 cm^-1 band is attributed to the -C=0 stretch vibrations of the v-lactone ring of acetogenins (^{Hidalgo et al., 2019}). The absorptions at 1,656,1,513, and 1,450 cm^-1 correspond to the C=C stretch vibration characteristics of the flavonoid bonds (^{Grijalva-Verdugo et al., 2018}). The bands between 1,376-1,320 cm^-1 are assigned to the -CH₃ groups of the alkanes and alkenes (Ibrahim et al., 2021). The peaks observed at 1,286,1,248, and 1,205 cm^-1 are attributed to the functional group -OH out of plane torsion of carboxylic acids, C-O stretching vibration and C-O-C asymmetric stretching, respectively, belonging to the pyran ring structure of tannins (^{Daud et al., 2016}; ^{Grijalva-Verdugo et al., 2018}). The signals observed at 1,161, 1,071, and 1,027 cm^-1 correspond to the 0=C-0 ester functional group attributed to coumarins (Ibrahim et al., 2021). Strong absorptions are observed in the region 828-719 cm^-1, assigned to the C-O stretching vibrations of a and β pyranose compounds present in anthocyanins (^{Grijalva-Verdugo et al., 2018}).

The FTIR spectra of the infusion from the two regions showed slight differences in the intensity and width of the bands and peaks; the differences maybe due to the variation in the number of functional groups that produce particular vibrational signals of molecular bonds corresponding to that specific wavenumber. The resulting spectrum is a fingerprint of the A. muricata leaf extract (^{Farooq & Sehgal, 2019}).

Cytotoxic activity assays

Figure 2 shows the results of cytotoxic activity assays of aqueous extract from Nayarit and Chiapas regions against MCF7 and HT-29 cells. The high cytotoxic activity was found for MCF7 cells with aqueous extract from Chiapas. Also, the results for HT-29 cells did not show statistical differences.

FIGURE 2 Cytotoxic effect of the aqueous extracts from Nayarit and Chiapas against MCF7 and HT-29 cells. Vertical bars correspond to the standard deviation. 

^{Coria-Téllez et al. (2019)} report that aqueous extracts of A. muricata from Nayarit reached a viability of 50% against HeLa cells with a dose extract of 2.42 mg ml^-1. ^{Hadisaputri et al. (2021)} mention a viability cell of 20% against MCF7 cell line using an aqueous fraction of A. muricata leaves at 250 μg ml^-1; the results show that the survival rate of the cells decreased with increased concentration. In the same way, ^{Syed Najmuddin et al. (2016)} evaluated the cytotoxic activity of water extracts of leaves of A. muricata against breast cells lines, MDA-MB231, MCF7, MCF-10A and 4T1 and reports a varied anti-breast cancer activity, around 50% for MCF-10A, 40% for MCF7, 30% for the cell lines MDA-MB231 and 4T1. The cytotoxic activity has been attributed to acetogenins in the leaves of A. muricata, from which more than 45 have been identified (^{Hadisaputri et al., 2021}). These secondary metabolites are responsible for the cytotoxicity activity because they inhibit the complex I of the mitochondrial respiratory chain, reducing the production of ATP, bringing on cellular apoptosis (^{Aguilar-Hernández et al., 2020}; ^{Grba et al., 2022}).