UPLC–QTOF–MS and NMR analyses of graviola (Annona muricata) leaves

Ingrid Vieira Machado de Moraes Paulo Riceli Vasconcelos Ribeiro Flávio Luis Schmidt Kirley Marques Canuto Guilherme Julião Zocolo Edy Sousa de Brito Rensheng Luo Kristy M. Richards Kevin Tran Robert E. Smith About the authors

ABSTRACT

Graviola leaves (Annona muricata L., Annonaceae) are used by some people to try to treat or even cure cancer, even though over-consumption of the fruit, which contains the neurotoxins annonacin and squamocin has caused an atypical form of Parkinson's disease. In previous analyses, the fruits were extracted with methanol under ambient conditions before analyses. In the present study, UPLC–QTOF–MS and NMR were used to analyze freeze-dried graviola leaves that were extracted using dry methanol and ethanol at 100 ºC and 10 MPa (100 atm) pressure in a sealed container. Methanol solubilized 33% of the metabolites in the lyophilized leaves. Ethanol solubilized 41% of metabolites in the lyophilized leaves. The concentrations of total phenolic compounds were 100.3 ± 2.8 and 93.2 ± 2.0 mg gallic acid equivalents per g of sample, for the methanolic and ethanolic extracts, respectively. Moreover, the toxicophore (unsaturated γ-lactone) that is present in neurotoxic acetogenins was found in the lipophilic portion of this extract. The concentrations of the neurotoxins annonacin and squamocin were found by UPLC–QTOF–MS to be 305.6 ± 28.3 and 17.4 ± 0.89 µg/g-dw, respectively, in the dried leaves. Pressurized methanol solubilized more annonacin and squamocin than ethanol. On the other hand, a hot, aqueous infusion solubilized only 0.213% of the annonacin and too little of the squamocin to be detected. So, graviola leaves contain significant amounts of the neurotoxins annonacin and squamocin, as well as some potentially healthy phenolic compounds. Finally, the potential neurotoxicity of whole leaves in dietary supplements could be much higher than that of a tea (hot aqueous infusion) that is made from them.

Keywords:
Graviola; Annona muricata; UPLC–QTOF–MS; Parkinson's disease; NMR

Introduction

Graviola (Annona muricata L.) is a tropical fruit in the Annonaceae family that is grown in Asia, South America and many tropical islands (Lannuzel and Michel, 2009Lannuzel, A., Michel, P.P., 2009. Atypical parkinsonism in the French West Indies: the plant toxin annonacin as a potential etiological factor. In: Tseng, K.-Y. (Ed.), Cortico-Subcortical Dynamics in Parkinson’s Disease. Humana Press, New York, pp. 283–290.; Gajalakshmi et al., 2012Gajalakshmi, S., Vijayalakshmi, S., Devi, R.V., 2012. Phytochemical and pharmacological properties of Annona muricata: a review. Intl. J. Pharm. Pharmaceut. Sci. 4, 3-6.). The leaves can be used to make a tea (Port's et al., 2013Port's, P.S., Chisté, R.C., Godoy, H.T., Prado, M.A., 2013. The phenolic compounds and the antioxidant potential of infusion of herbs from the Brazilian Amazonian region. Food Res. Intl. 53, 875-883.; Hansra et al., 2014Hansra, D.M., Silva, O., Mehta, A., Ahn, E., 2014. Patient with metastatic breast cancer achieves stable disease for 5 years on graviola and Xeloda after progressing on multiple lines of therapy. Adv. Breast Cancer Res. 3, 84-87.) or consumed whole as dietary supplement in capsules that may have some health effects (Torres et al., 2012Torres, M.P., Rachagani, S., Purohit, V., Pandey, P., Joshi, S., Moore, E.D., Johansson, S.L., Singh, P.K., Ganti, A.K., Batra, S.K., 2012. Graviola: a novel promising natural-derived drug that inhibits tumorigenicity and metastasis of pancreatic cancer cells in vitro and in vivo through altering cell metabolism. Cancer Lett. 323, 29-40.). However, over-consumption of graviola and products made from it may have caused an atypical form of Parkinson's diseases on the French Caribbean island of Guadeloupe and the Pacific island of Guam (Caparros-Lefebvre and Elbaz, 1999Caparros-Lefebvre, D., Elbaz, A., 1999. Possible relation of atypical parkinsonism in the French West Indies with consumption of tropical plants: a case–control study. Lancet 354, 281-285.; Champy et al., 2005Champy, P., Melot, A., Guerineau, V., Gleye, C., Fall, D., Höglinger, G.U., Ruberg, M., Lannuzel, A., Laprevote, O., Laurens, A., Hocquemiller, R., 2005. Quantification of acetogenins in Annona muricata linked to atypical parkinsonism in Guadeloupe. Movement Disorders 20, 1629-1633.Texto; Badrie and Schauss, 2009Badrie, N., Schauss, A.G., 2009. Soursop (Annona muricata L.): composition, nutritional value, medicinal uses, and toxicology. In: Watson, R.R., Preddy, V.R. (Eds.), Bioactive Foods in Promoting Health. Fruits and Vegetables. Academic Press, Oxford, pp. 621–643.). This is due to the presence of neurotoxic acetogenins, such as annonacin (1) (C35H64O7, MW 596.88) and squamocin (2) (C37H66O7, MW 622.92). Like other neurotoxic acetogenins, they contain an unsaturated γ-lactone group that is the toxicophore (Smith et al., 2014Smith, R.E., Tran, K., Richards, K.M., 2014. Bioactive Annonaceous Acetogenins, Chapter 4, in Studies in Natural Products Chemistry, vol. 41. Elsevier, New York, pp. 93–117.). However, the neurotoxicity could be strongly dependent on the dose. That is, the first rule of toxicology is that the dose is the poison (Smith, 2014Smith, R.E., 2014. Medicinal Chemistry – Fusion of Traditional and Western Medicine, Second Edition. Bentham Science, Sharjah, U.A.E, pp. 316–326.). It was over-consumption, not consumption of graviola that caused Parkinson's disease. The dose of neurotoxic acetogenins that one consumes is very dependent on the form of graviola that is ingested. For example, one study used a tea made from 10 to 12 dry leaves that were boiled in 8 oz (about 237 ml) water for 5–7 min (Hansra et al., 2014Hansra, D.M., Silva, O., Mehta, A., Ahn, E., 2014. Patient with metastatic breast cancer achieves stable disease for 5 years on graviola and Xeloda after progressing on multiple lines of therapy. Adv. Breast Cancer Res. 3, 84-87.). Since acetogenins have very low solubilities in water (Höllerhage et al., 2009Höllerhage, M., Matusch, A., Champy, P., Lombes, A., Ruberg, M., Oertel, W.H., Höglinger, G.U., 2009. Natural lipophilic inhibitors of mitochondrial complex I are candidate toxins for sporadic neurodegenerative tau pathologies. Exp. Neurol. 220, 133-142.), this would have been a relatively low dose. The dose can be estimated from a previous study that found about 56 mg of annonacin per kg of leaves when about 2.5 g of leaves were boiled for 10 min in water (Champy et al., 2005Champy, P., Melot, A., Guerineau, V., Gleye, C., Fall, D., Höglinger, G.U., Ruberg, M., Lannuzel, A., Laprevote, O., Laurens, A., Hocquemiller, R., 2005. Quantification of acetogenins in Annona muricata linked to atypical parkinsonism in Guadeloupe. Movement Disorders 20, 1629-1633.Texto). On the other hand, much higher doses would be consumed if one were to ingest the entire leaves in a dietary supplement. That is, one group reported that the "therapeutic dose" of graviola leaves is 2–3 g, taken 3 or 4 times daily (Sun et al., 2014Sun, S., Liu, J., Kadouh, H., Sun, X., Zhou, K., 2014. Three new anti-proliferative Annonaceous acetogenins with mono-tetrahydrofuran ring from graviola fruit (Annona muricata). Bioorg. Med. Chem Lett. 24, 2773-2776.). If whole leaves are consumed, the expected dose would be higher than that in a tea. To estimate the concentration, 100 g of dried leaves were extracted with 1 l of methanol (Champy et al., 2005Champy, P., Melot, A., Guerineau, V., Gleye, C., Fall, D., Höglinger, G.U., Ruberg, M., Lannuzel, A., Laprevote, O., Laurens, A., Hocquemiller, R., 2005. Quantification of acetogenins in Annona muricata linked to atypical parkinsonism in Guadeloupe. Movement Disorders 20, 1629-1633.Texto). This recovered almost six times as much annonacin (about 300 mg/kg) as did the boiling water. However, this analysis was done with matrix assisted laser desorption and ionization coupled to time of flight mass spectrometry or MALDI–TOF MS (Champy et al., 2005Champy, P., Melot, A., Guerineau, V., Gleye, C., Fall, D., Höglinger, G.U., Ruberg, M., Lannuzel, A., Laprevote, O., Laurens, A., Hocquemiller, R., 2005. Quantification of acetogenins in Annona muricata linked to atypical parkinsonism in Guadeloupe. Movement Disorders 20, 1629-1633.Texto), which separates analytes based on their molecular weights. So, annonacin will not be separated from its isomers. This could cause an over-estimation of the concentration of annonacin, since its isomers will also be detected by the MALDI–TOF MS. It was observed, boiling methanol was not able to solubilize all the annonacin (1) or squamocin (2) from another fruit in the Annonaceae family, the North American pawpaw (Asimina triloba; Potts et al., 2012Potts, L.F., Luzzio, F.A., Smith, S.C., Hetman, M., Champy, P., Litvan, I., 2012. Annonacin in Asimina triloba fruit: Implication for neurotoxicity. NeuroToxicol. 33, 53-58.). That is, the method used to extract annonacin and squamocin can affect the results of the analysis. Boiling water will not solubilize all the annonacin or squamocin. Instead, pressurized liquid extraction using dry methanol at 100 ºC and 10 MPa (100 atm) pressure can solubilize over seven times as much annonacin and even some squamocin from the fruits (Potts et al., 2012Potts, L.F., Luzzio, F.A., Smith, S.C., Hetman, M., Champy, P., Litvan, I., 2012. Annonacin in Asimina triloba fruit: Implication for neurotoxicity. NeuroToxicol. 33, 53-58.). So, it is important to use hot, dry pressurized methanol to extract all the annonacin and squamocin. It is also important to use liquid chromatography coupled to high resolution mass spectrometry (LC-HRMS) to separate and detect isomers of annonacin and squamocin (Levine et al., 2015Levine, R.A., Richards, K.M., Tran, K., Luo, R., Thomas, A.L., Smith, R.E., 2015. Determination of neurotoxic acetogenins in pawpaw (Asimina triloba) fruit by LC-HRMS. J. Agric. Food Chem. 63, 1053-1056.).

LC-HRMS is especially useful when trying to quantify trace components, like acetogenins. The MS can be tuned so that it is as sensitive as possible for the analytes of interest (Smith, 2014Smith, R.E., 2014. Medicinal Chemistry – Fusion of Traditional and Western Medicine, Second Edition. Bentham Science, Sharjah, U.A.E, pp. 316–326.). However, the sensitivities for other analytes that are present in a sample are quite different. So, standards are required when quantifying analytes by MS. On the other hand, 1H NMR has the same sensitivity for all the hydrogens (1H isotope) in the sample (Smith, 2014Smith, R.E., 2014. Medicinal Chemistry – Fusion of Traditional and Western Medicine, Second Edition. Bentham Science, Sharjah, U.A.E, pp. 316–326.). So, it can be used without standards to determine the relative concentrations of different types of hydrogens in a sample (Smith, 2014Smith, R.E., 2014. Medicinal Chemistry – Fusion of Traditional and Western Medicine, Second Edition. Bentham Science, Sharjah, U.A.E, pp. 316–326.). Moreover, 13C NMR can be used to determine how many different kinds of carbons are in a sample and to identify the types of compounds that are present (Smith, 2014Smith, R.E., 2014. Medicinal Chemistry – Fusion of Traditional and Western Medicine, Second Edition. Bentham Science, Sharjah, U.A.E, pp. 316–326.). Therefore, NMR can be very useful in analyzing relatively unknown samples. NMR was used to analyze extracts of A. muricata for the presence of the α,β-unsaturated-γ-lactone toxicophore that is present in neurotoxic acetogenins (Machado et al., 2015Machado, A.R.T., Lage, G.A., Medeiros, F.S., Filho, J.D.S., Pimenta, L.P.S., 2015. Total α,β-unsaturated-γ-lactone acetogenins in Annona muricata by proton NMR spectroscopy. Appl. Mag. Res. 46, 153-160.).

Even though graviola leaves have been reported to contain about 300 mg/kg annonacin (Höllerhage et al., 2009Höllerhage, M., Matusch, A., Champy, P., Lombes, A., Ruberg, M., Oertel, W.H., Höglinger, G.U., 2009. Natural lipophilic inhibitors of mitochondrial complex I are candidate toxins for sporadic neurodegenerative tau pathologies. Exp. Neurol. 220, 133-142.), 32 mg/kg total phenolics and 5.6 mg/kg total flavonoids (Port's et al., 2013Port's, P.S., Chisté, R.C., Godoy, H.T., Prado, M.A., 2013. The phenolic compounds and the antioxidant potential of infusion of herbs from the Brazilian Amazonian region. Food Res. Intl. 53, 875-883.), little is known about its major chemical constituents. Even the concentration of total soluble substances is not well known. That is, the concentration depends on the method used to extract the soluble substances. For example, when dried leaves were soaked in distilled water for 48 h, about 5% of the solids dissolved (Florence et al., 2014Florence, N.T., Benoit, M.Z., Jonas, K., Alexandra, T., Desire, D.D.P., Pierre, K., Theophile, D., 2014. Antidiabetic and antioxidant effects of Annona muricata (Annonaceae), aqueous extract on streptozotocin-induced diabetic rats. J. Ethnopharmacol. 151, 784-790.). Another obtained only a 3.62% yield when leaves were extracted twice with distilled water at room temperature (Adewole and Ajewole, 2009Adewole, S.O., Ajewole, J.A.O., 2009. Protective effects of Annona muricata Linn. (Annonaceae) leaf aqueous extract on serum lipid profiles and oxidative stress in hepatocytes of streptozotocin-treated diabetic rats. Afr. J. Compl. Alt. Med. 6, 30-41.). When soaked for 48 h in 80% ethanol, about 10.55% of the compounds in the leaves dissolved (Foong and Hamid, 2012Foong, C.P., Hamid, R.A., 2012. Evaluation of anti-inflammatory activities of ethanolic extract of Annona muricata leaves. Rev. Bras. Farmacogn. 22, 1301-1307.; Hamizah et al., 2012Hamizah, S., Roslida, A.H., Fezah, O., Tan, K.L., Tor, Y.S., Tan, C.I., 2012. Chemopreventive potential of Annona Muricata L leaves on chemically-induced skin papillomagenesis in mice. Asian Pacific J. Cancer Prevention 13, 2533-2539.). A different study percolated dry leaves with 95% ethanol to get a 19.3% yield of soluble substances (Singleton et al., 1999Singleton, V.L., Orthofer, R., Lamuela-Raventos, R.M., 1999. Analysis of total phenols and others oxidation substrates and antioxidants by Folin–Ciocauteau reagent. Methods Enzymol. 299, 152.).

So, the aims of this work were characterizing the extracts obtained by this technique and see if methanol and ethanol present different results, as well as analyze annonacin and squamocin by UPLC–QTOF–MS.

Materials and methods

Chemicals and graviola leaves

Methanol (CH3OH), chloroform (CHCl3), acetonitrile (CH3CN) and ethanol were from Honeywell Burdick & Jackson (Muskegon, MI, USA). Deuterated chloroform (CDCl3), deuterated methanol (CD3OD) and deuterated water (D2O) were from SigmaAldrich (St. Louis, MO). Leucine-enkephalin was purchased from Waters (North Kingstown, RI, USA). Lyophilized graviola leaves were obtained by Ingrid de Moraes from a commercial cultivation located in the city of Trairi (Ceará, Brazil), latitude 3º22′15.98′′S and longitude 39º17′34.46′′W. Voucher specimens are kept at Embrapa in Fortaleza. The voucher number is 49002. After being harvested, the leaves were washed, sanitized with a sodium hypochlorite solution (100 µl l-1 or 100 ppm) and then lyophilized in a model LP 510 lyophilizer (Liobrás, São Carlos, SP - Brazil), at Embrapa Tropical Agroindustry located in Fortaleza, Ceará, Brazil.

Extractions

About 10 g of lyophilized leaves were mixed with enough of HydroMatrixTM (SigmaAldrich, St. Louis, MO) to fill the 100 ml stainless steel sample cell used in an Accelerated Solvent Extractor (ASE, ThermoFisher Scientific, Sunnyvale, CA). Then, CH3OH or ethanol was added while the temperature and pressure were increased to 100 ºC and 10.3 MPa (100 atm) over a 3 min time (static time). Next, the solvent was purged into a collection vessel. A total of four cycles were run to statically extract the sample, resulting in a total volume of about 160 ml. The solvent was evaporated off and the oily residues remaining after extraction with each solvent were weighed. A portion of the residue obtained from the methanol extraction was dissolved in 99.99% CD3OD and analyzed by NMR. Portions of the methanolic and ethanolic residues were analyzed for total phenolics, as described below. Then, portions of the residues obtained from the methanolic and ethanolic extraction were partitioned between CHCl3 and water. The CHCl3 phase was collected, the solvent evaporated off, the residue redissolved in CDCl3 and NMR spectra acquired. Finally, a 10 g portion of dried leaves were extracted by ultrasonication. It was done with 200 ml of 50% ethanol plus 50% water at 40 ºC for 10 min, using an Ultracelaner 1450 ultrasonicator from Unique (Indaia, SP, Brazil). It was done at a frequency of 20 kHz and with 800 W power. This extract was also analyzed for total phenolics.

Also, three portions of about 2.5 g and one portion of about 5.0 g of lyophilized leaves were added to 237 ml (one cup) of water at 90 ºC for 10 min, after which the solutions were filtered and analyzed by LC–MS/MS.

NMR analyses

1H and 13C{1H}-NMR spectra were obtained using an Agilent DD2 600 MHz NMR (Santa Clara, CA). A 30º pulse width and 1 s pulse delay were used for the 1H NMR, while a 30º pulse width and 2 s pulse delay were used for the 13C NMR spectra. Chemical shifts were referenced to either the CD3OD peaks at 3.35 and 4.78 (for 1H) and 49.3 ppm (for 13C) or the CDCl3 peaks at 7.27 and 77.23 ppm for 1H and 13C, respectively.

Analysis for total phenolic compounds

The concentrations of total phenolics in the methanolic and ethanolic extracts were determined using the Folin–Ciocalteu reagent and a gallic acid reference standard, as described previously (Singleton et al., 1999Singleton, V.L., Orthofer, R., Lamuela-Raventos, R.M., 1999. Analysis of total phenols and others oxidation substrates and antioxidants by Folin–Ciocauteau reagent. Methods Enzymol. 299, 152.). Results were expressed as mg gallic acid equivalents per g of sample, or mg GAE/g-dw.

UPLC–QTOF–MS analysis

First, about 10 mg of each of the residues remaining after evaporating the solvent off of the hot, pressurized extracts was dissolved in 5 ml of ethanol:H2O (1:1, v/v) and passed through a LC-18 Supelcoclean 6 ml (0.5 g) solid phase extraction (SPE) cartridge (Supelco, Bellefonte, PA, USA) that was preconditioned by washing it three times with CH3OH and then equilibrated with 3 ml of deionized water. The acetogenins were eluted with 6 ml of CH3CN. The CH3CN was evaporated off and the residue redissolved in 1 ml of CH3CN:H2O (15:7, v/v). The residues from each sample were analyzed by positive ion UPLC–MS analysis using a Xevo UPLC-QTOF from Waters (Milford, MA) and an Acquity UPLC BEH C18 column, 1.7 µm, 2.1 mm × 100 mm. A gradient elution was used with a mixture of H2O and CH3CN and 0.1% formic acid, flowing at 400 µl/min. That is, even though the concentrations of H2O and CH3CN varied, the concentration of formic acid did not. It was held constant at 0.1%. It started with 35:65 H2O/CH3CN (v/v) for the first 25 min, then increased linearly to 1:9 H2O/CH3CN (v/v) from 25 to 30 min. It was held at 1:9 H2O/CH3CN (v/v) from 30 to 35 min, then decreased linearly to 35:65 H2O/CH3CN (v/v) from 35 to 36 min and held at this from 36 to 40 min. The MS spectra were acquired in positive ion mode with an electrospray ionization (ESI) source. The MS parameters were as follows: desolvation temperature 350 ºC, capillary voltage 3.2 kV, sampling cone voltage 32 V, extraction cone voltage 1.0 V, source temperature 120 ºC, cone gas 2 l/h, desolvation gas flow 350 l/h, purge gas flow 350 l/h, collision energy 5.0 V. Centroid data were collected for each sample in a range 110–800 Da, and the m/z value of all acquired spectra was automatically corrected during acquisition based on lockmass by infusing 200 pg/l of leucine-enkephalin thorough Lockspray at a flow rate of 20 µl/min. The calibration was performed to achieve an acceptable accuracy error of ±5 ppm. The accurate mass and molecular formula denomination were acquired with the MassLynx 4.1 software (Waters MS Technologies).

Annonacin and squamocin standards from Pierre Champy at the University of Paris were used to identify and quantify the peaks due to annonacin (1) and squamocin (2). They eluted at 14.1 and 15.6 min, respectively. Ions with m/z = 597.47 and 623.48 Da were used to quantify annonacin and squamocin, respectively. Calibration curves were obtained using 10–200 ng/ml annonacin and squamocin, assuming that the standards were 100% pure. Samples were injected in triplicate with results reported as the average ± standard deviation.

Results and discussion

Extraction

Dry methanol at 100 ºC and 10 MPa (100 atm) pressure solubilized 33% of the leaves, while ethanol under the same conditions solubilized 41%. So, hot, pressurized methanol extracted much more material than boiling (or percolating) under ambient pressure, which solubilized 19.3% of the leaves in a previous study (Port's et al., 2013Port's, P.S., Chisté, R.C., Godoy, H.T., Prado, M.A., 2013. The phenolic compounds and the antioxidant potential of infusion of herbs from the Brazilian Amazonian region. Food Res. Intl. 53, 875-883.). The concentration of total phenolic compounds that were found using hot, pressurized methanol extraction was 100.3 ± 2.8 mg GAE/g. This compares to 93.2 ± 2.0 mg GAE/g in the hot, pressurized ethanol extract, 54.6 mg GAE/g in the ultrasonic extract using 50% ethanol and the value of 33 mg GAE/g that was found by others when leaves were extracted with boiling water. So, extraction with dry methanol at 100 ºC and 10 MPa (100 atm) pressure can extract more phenolic compounds than boiling water under ambient pressure.

NMR analysis

The 1H and 13C{1H}-NMR spectra of the methanolic extract of lyophilized graviola leaves are shown in Figs. 1 and 2. They both contain signals (chemical shifts) due to the CH3–(CH2)nCH2COOR of esters (1H: 0.66–1.0 and 1.1–1.4 ppm; 13C: 14.6, 16.3–39.3 and 175–177 ppm), the –HCOH and –H2COH of carbohydrates (1H: 3.0–5.4 ppm; 13C: 62.2–108.8 ppm) and the HC=C of aromatic hydrogens in phenolic compounds (1H: 6.4–8.1 ppm; 13C: 125.5–166.2 ppm). There are also signals due to –CH2–HC=CH– portion of unsaturated fatty acyls (1H: 6.4 ppm; 13C: 108.9–119.5 ppm). The relative contributions to the total peak area in the 1H spectrum were 24% alkyl (CH3–(CH2)n), 46% carbohydrate, and 4.2% phenyl. There were no peaks due to di- or triglycerides. The alkyl, carbohydrate and carbonyl carbons all produced chemical shifts that were consistent with the presence of esters of fatty acyls and sugars. These are more commonly called fatty acid glycosides. They have been found in the tropical fruit called noni (Morinda citrifolia) and have anticancer properties (Smith et al., 2014Smith, R.E., 2014. Medicinal Chemistry – Fusion of Traditional and Western Medicine, Second Edition. Bentham Science, Sharjah, U.A.E, pp. 316–326.; Akihisa et al., 2007Akihisa, T., Matsumoto, K., Tokuda, H., Yasukawa, K., Seine, K., Nakamoto, K., Kuninaga, H., Suzuki, T., Kimura, Y., 2007. Anti-inflammatory and potential cancer chemopreventive constituents of the fruits of Morinda citrifolia (Noni). J. Nat. Prod. 70, 754-757.; Kim, 2010Kim, H.-K., 2010. Identification of novel fatty acid glucosides from the tropical fruit Morinda citrifolia L. Phytochem. Lett. 3, 238-241.). It should also be noted that even though the methanolic extract was green due to the presence of chlorophyll, the chlorophyll was present at too low of a concentration to produce peaks in the NMR spectra. There may also be some signals due to terpenes that may be slightly soluble in methanol, even though they are usually extracted using hexane.

Fig. 1
1H-NMR spectrum of methanolic extract (100 ºC, 100 MPascal) of lyophilized graviola leaves, in CD3OD. Peaks: 1: –CH3; 2: –(CH2)n; 3: H2C–CH=CH–; 4: HCO of sugars 5: phenyls and/or phenolics.
Fig. 2
13C{1H}-NMR spectrum of methanolic extract (100 ºC, 100 MPascal) of lyophilized graviola leaves, in CD3OD.

About 37.5% of the methanolic extract partitioned into CHCl3. The 1H and 13C{1H}-NMR spectra of this lipophilic portion of the methanolic extract are shown in Figs. 3 and 4. There are chemical shifts at or about 7.0 and 7.2 ppm due to the unsaturated γ-lactone toxicophore that is in neurotoxic acetogenins (Luo et al., 2013Luo, R., Maia, J.G., de Moraes, M.R., Godoy, H.T., Sabaa-Srur, A.U.O., Tran, K., Richards, K.M., Monroe, D.M., Vocque, R.H., Smith, R.E., 2013. NMR analysis of potentially neurotoxic annonaceous fruits. Nat. Prod. J. 3, 230-241.Texto), and very small signals due to four different HC=C hydrogens in chlorophyll at 8.00, 8.55, 9.29 and 9.52 ppm.

Fig. 3
1H NMR spectrum of the portion of the methanolic extract (100 ºC, 10 MPa) of oven dried graviola leaves that partitioned into CHCl3, redissolved in CDCl3. Peaks: 1: –CH3; 2: –(CH2)n; 3: H2C–CH=CH–; 4: HCO of sugars; 5: HC=CH; 6: HC=C in the unsaturated γ-lactone toxicophore in acetogenins; 0.1% CHCl3 in 99.9% CDCl3 solvent.
Fig. 4
13C{1H}-NMR of the portion of the methanolic extract (100 ºC, 10 MPa) of oven dried graviola leaves that partitioned into CHCl3, redissolved in CDCl3.

In comparison, the 1H and 13C{1H}-NMR spectra of the portion of the ethanolic extract that partitioned into CDCl3 are shown in the Supplementary Material. The chemical shifts due to the unsaturated γ-lactone toxicophore are also present.

LC–MS analysis

Seven point calibration curves each for annonacin (1) and squamocin (2) showed adequate linear correlation between concentration and area (R2 > 0.99), the linear range of standard compounds was between 10 and 200 ng/ml for standard compounds. Standards and samples were injected in triplicate. The injections showed that the results are highly reproducible and repetitive. The LC–MS chromatograms of 200 ng/ml annonacin and squamocin standards and the methanolic extract of graviola leaves are shown in Fig. 5. The concentrations of the neurotoxins annonacin and squamocin were found to be 1.68 ± 0.08 and 0.023 ± 0.001 mg/g-dw, respectively in the dried leaves. Pressurized methanol solubilized more annonacin and squamocin than ethanol. That is, 305.6 ± 28.3 µg/g-dw annonacin was found in the methanolic extract and 226.0 ± 0.07 µg/g-dw was found in the ethanolic extract. On the other hand, 17.4 ± 0.89 µg/g-dw squamocin was found in the methanolic extract, and 4.75 ± 0.41 mg/g-dw squamocin was found in the ethanolic extract. On the other hand, a hot, aqueous infusion (tea) solubilized only 0.213% (0.65 ± 0.07 µg/g-dw) of the annonacin from 2.5 g of dry leaves and too little of the squamocin to be detected. When 5.0 g of leaves were used to prepare a tea, only 0.108% (0.33 µg/g-dw) of the leaves. So, tea bags containing about 5 g of dry graviola leaves will not solubilize much more annonacin than tea bags containing 2.5 g of leaves. Thus, whole graviola leaves contain significant amounts of the neurotoxins annonacin and squamocin, as well as phenolic compounds. Moreover, the potential neurotoxicities of whole leaves in dietary supplements could be much higher than that of a tea that is made from them.

Fig. 5
UPLC–QTOF–MS chromatogram of the hot, pressurized methanolic extract (100 ºC, 10 MPa) of lyophilized graviola leaves – total ion chromatogram of the sample (A); annonacin standard – 200 ng/ml (Rt = 14.1 min) (B); all peaks of ions [M+H]+ = 597.47 extracted ion chromatogram of the sample (C); squamocin standard – 200 ng/ml (Rt = 15.6 min) (D); all peaks of ions [M+H]+ = 623.48 extracted of the sample (E).

In conclusion much more material in dried graviola leaves can be solubilized using pressurized liquid extraction and either dry methanol or ethanol, with ethanol solubilizing a little bit more. The NMR spectra were dominated by peaks due to fatty acid glycosides, which may have anticancer and other health benefits (Kim et al., 1998Kim, G.-S., Alali, F., Rogers, L.L., Wu, F.-E., Sastrodihardjo, S., McLaughlin, J.L., 1998. Muricoreacin and murihexocin C, mono-tetrahydrofuran acetogenins from the leaves of Annona muricata. Phytochemistry 49, 565-571.; Akihisa et al., 2007Akihisa, T., Matsumoto, K., Tokuda, H., Yasukawa, K., Seine, K., Nakamoto, K., Kuninaga, H., Suzuki, T., Kimura, Y., 2007. Anti-inflammatory and potential cancer chemopreventive constituents of the fruits of Morinda citrifolia (Noni). J. Nat. Prod. 70, 754-757.; Smith, 2014Smith, R.E., 2014. Medicinal Chemistry – Fusion of Traditional and Western Medicine, Second Edition. Bentham Science, Sharjah, U.A.E, pp. 316–326.). They are also surfactants, so they will probably increase the solubilities of acetogenins in hot water. The NMR spectra also contained chemical shifts due to the unsaturated γ-lactone toxicophore that is in annonacin and squamocin (Machado et al., 2015Machado, A.R.T., Lage, G.A., Medeiros, F.S., Filho, J.D.S., Pimenta, L.P.S., 2015. Total α,β-unsaturated-γ-lactone acetogenins in Annona muricata by proton NMR spectroscopy. Appl. Mag. Res. 46, 153-160.). They are also surfactants, so they will probably increase the solubilities of acetogenins in hot water. Finally, the total phenolics in the hot, pressurized methanolic and ethanolic extracts were 83 and 64 mf GAE/g, respectively.

Moreover, methanol at 100 ºC and 10 MPa (100 atm) pressure solubilized 33% of the lyophilized graviola leaves. Ethanol solubilized 41% of the lyophilized leaves. The concentrations of total phenolic compounds were 83 and 64 mg GAE/g for the methanolic and ethanolic extracts, respectively. Moreover, the toxicophore (unsaturated γ-lactone) that is present in neurotoxic acetogenins was found in the lipophilic portion of these extracts. The concentrations of the neurotoxins annonacin (1) and squamocin (2) were found by UPLC–TOF–MS to be 305.6 ± 28.3 and 17.4 ± 0.89 mg/g, respectively in the dried leaves. Pressurized methanol solubilized more annonacin than ethanol, while pressurized ethanol solubilized more squamocin than methanol. On the other hand, a hot, aqueous infusion (tea) made from 2.5 g of dried leaves in 237 ml (one cup) of water solubilized only 0.213% (0.65 ± 0.07 µg/g-dw) of the annonacin and too little of the squamocin to be detected.

In summary, we conclude that hot, pressurized ethanol can solubilize more material than ethanol, the concentrations of total phenolic compounds were 100.3 ± 2.8 and 93.2 ± 2.0 mg gallic acid equivalents per g of sample for the methanolic and ethanolic extracts, respectively, the toxicophore (unsaturated γ-lactone) that is present in neurotoxic acetogenins was found in the lipophilic portion of these extracts, the concentrations of the neurotoxins annonacin and squamocin were found by UPLC–TOF–MS to be 305.6 ± 28.3 and 17.4 ± 0.89 mg/g, respectively in the dried leaves, pressurized methanol solubilized more annonacin and squamocin than ethanol, and a hot, aqueous infusion (tea) made from 2.5 g of dried leaves in 237 ml (one cup) of water solubilized only 0.213% (0.65 ± 0.07 µg/g-dw) of the annonacin and too little of the squamocin to be detected.

This work should not be taken as reflecting FDA policy or regulations.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bjp.2015.12.001.

References

  • Adewole, S.O., Ajewole, J.A.O., 2009. Protective effects of Annona muricata Linn. (Annonaceae) leaf aqueous extract on serum lipid profiles and oxidative stress in hepatocytes of streptozotocin-treated diabetic rats. Afr. J. Compl. Alt. Med. 6, 30-41.
  • Akihisa, T., Matsumoto, K., Tokuda, H., Yasukawa, K., Seine, K., Nakamoto, K., Kuninaga, H., Suzuki, T., Kimura, Y., 2007. Anti-inflammatory and potential cancer chemopreventive constituents of the fruits of Morinda citrifolia (Noni). J. Nat. Prod. 70, 754-757.
  • Badrie, N., Schauss, A.G., 2009. Soursop (Annona muricata L.): composition, nutritional value, medicinal uses, and toxicology. In: Watson, R.R., Preddy, V.R. (Eds.), Bioactive Foods in Promoting Health. Fruits and Vegetables. Academic Press, Oxford, pp. 621–643.
  • Caparros-Lefebvre, D., Elbaz, A., 1999. Possible relation of atypical parkinsonism in the French West Indies with consumption of tropical plants: a case–control study. Lancet 354, 281-285.
  • Champy, P., Melot, A., Guerineau, V., Gleye, C., Fall, D., Höglinger, G.U., Ruberg, M., Lannuzel, A., Laprevote, O., Laurens, A., Hocquemiller, R., 2005. Quantification of acetogenins in Annona muricata linked to atypical parkinsonism in Guadeloupe. Movement Disorders 20, 1629-1633.Texto
  • Florence, N.T., Benoit, M.Z., Jonas, K., Alexandra, T., Desire, D.D.P., Pierre, K., Theophile, D., 2014. Antidiabetic and antioxidant effects of Annona muricata (Annonaceae), aqueous extract on streptozotocin-induced diabetic rats. J. Ethnopharmacol. 151, 784-790.
  • Foong, C.P., Hamid, R.A., 2012. Evaluation of anti-inflammatory activities of ethanolic extract of Annona muricata leaves. Rev. Bras. Farmacogn. 22, 1301-1307.
  • Gajalakshmi, S., Vijayalakshmi, S., Devi, R.V., 2012. Phytochemical and pharmacological properties of Annona muricata: a review. Intl. J. Pharm. Pharmaceut. Sci. 4, 3-6.
  • Hamizah, S., Roslida, A.H., Fezah, O., Tan, K.L., Tor, Y.S., Tan, C.I., 2012. Chemopreventive potential of Annona Muricata L leaves on chemically-induced skin papillomagenesis in mice. Asian Pacific J. Cancer Prevention 13, 2533-2539.
  • Hansra, D.M., Silva, O., Mehta, A., Ahn, E., 2014. Patient with metastatic breast cancer achieves stable disease for 5 years on graviola and Xeloda after progressing on multiple lines of therapy. Adv. Breast Cancer Res. 3, 84-87.
  • Höllerhage, M., Matusch, A., Champy, P., Lombes, A., Ruberg, M., Oertel, W.H., Höglinger, G.U., 2009. Natural lipophilic inhibitors of mitochondrial complex I are candidate toxins for sporadic neurodegenerative tau pathologies. Exp. Neurol. 220, 133-142.
  • Kim, H.-K., 2010. Identification of novel fatty acid glucosides from the tropical fruit Morinda citrifolia L. Phytochem. Lett. 3, 238-241.
  • Kim, G.-S., Alali, F., Rogers, L.L., Wu, F.-E., Sastrodihardjo, S., McLaughlin, J.L., 1998. Muricoreacin and murihexocin C, mono-tetrahydrofuran acetogenins from the leaves of Annona muricata Phytochemistry 49, 565-571.
  • Luo, R., Maia, J.G., de Moraes, M.R., Godoy, H.T., Sabaa-Srur, A.U.O., Tran, K., Richards, K.M., Monroe, D.M., Vocque, R.H., Smith, R.E., 2013. NMR analysis of potentially neurotoxic annonaceous fruits. Nat. Prod. J. 3, 230-241.Texto
  • Lannuzel, A., Michel, P.P., 2009. Atypical parkinsonism in the French West Indies: the plant toxin annonacin as a potential etiological factor. In: Tseng, K.-Y. (Ed.), Cortico-Subcortical Dynamics in Parkinson’s Disease. Humana Press, New York, pp. 283–290.
  • Levine, R.A., Richards, K.M., Tran, K., Luo, R., Thomas, A.L., Smith, R.E., 2015. Determination of neurotoxic acetogenins in pawpaw (Asimina triloba) fruit by LC-HRMS. J. Agric. Food Chem. 63, 1053-1056.
  • Machado, A.R.T., Lage, G.A., Medeiros, F.S., Filho, J.D.S., Pimenta, L.P.S., 2015. Total α,β-unsaturated-γ-lactone acetogenins in Annona muricata by proton NMR spectroscopy. Appl. Mag. Res. 46, 153-160.
  • Port's, P.S., Chisté, R.C., Godoy, H.T., Prado, M.A., 2013. The phenolic compounds and the antioxidant potential of infusion of herbs from the Brazilian Amazonian region. Food Res. Intl. 53, 875-883.
  • Potts, L.F., Luzzio, F.A., Smith, S.C., Hetman, M., Champy, P., Litvan, I., 2012. Annonacin in Asimina triloba fruit: Implication for neurotoxicity. NeuroToxicol. 33, 53-58.
  • Singleton, V.L., Orthofer, R., Lamuela-Raventos, R.M., 1999. Analysis of total phenols and others oxidation substrates and antioxidants by Folin–Ciocauteau reagent. Methods Enzymol. 299, 152.
  • Smith, R.E., 2014. Medicinal Chemistry – Fusion of Traditional and Western Medicine, Second Edition. Bentham Science, Sharjah, U.A.E, pp. 316–326.
  • Smith, R.E., Tran, K., Richards, K.M., 2014. Bioactive Annonaceous Acetogenins, Chapter 4, in Studies in Natural Products Chemistry, vol. 41. Elsevier, New York, pp. 93–117.
  • Sun, S., Liu, J., Kadouh, H., Sun, X., Zhou, K., 2014. Three new anti-proliferative Annonaceous acetogenins with mono-tetrahydrofuran ring from graviola fruit (Annona muricata). Bioorg. Med. Chem Lett. 24, 2773-2776.
  • Torres, M.P., Rachagani, S., Purohit, V., Pandey, P., Joshi, S., Moore, E.D., Johansson, S.L., Singh, P.K., Ganti, A.K., Batra, S.K., 2012. Graviola: a novel promising natural-derived drug that inhibits tumorigenicity and metastasis of pancreatic cancer cells in vitro and in vivo through altering cell metabolism. Cancer Lett. 323, 29-40.

Publication Dates

  • Publication in this collection
    Mar-Apr 2016

History

  • Received
    06 Apr 2015
  • Accepted
    07 Dec 2015
Sociedade Brasileira de Farmacognosia Universidade Federal do Paraná, Laboratório de Farmacognosia, Rua Pref. Lothario Meissner, 632 - Jd. Botânico, 80210-170, Curitiba, PR, Brasil, Tel/FAX (41) 3360-4062 - Curitiba - PR - Brazil
E-mail: revista@sbfgnosia.org.br