Sage Journals: Discover world-class research

Abstract

Dietary supplements containing plant materials of Annonaceae species (Annona muricata L., A. squamosa L., A. mucosa JACQ., A. squamosa × cherimola Mabb.) were extracted by hot, pressurized ethyl acetate and analyzed for their effect in vitro on Lund human mesencephalic neurons. Cell viability was measured by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay, and cell death was determined by lactate dehydrogenase levels. Three supplements strongly decreased the cell viability at extract concentrations of 1 µg/mL, of which 1 decreased cell viability at 0.1 µg/µL. Also, strong neuronal toxicities of these supplements were found. Cell death was observed at concentrations of 10 µg/mL. The degree of toxicity was comparable to the ones found in Annonaceous fruit extracts. Two fruit pulps of Annonaceae (A. muricata and A. squamosa) showed a reduction in cell viability at lower concentrations. The fruit pulp extract of A. muricata revealed the strongest neurotoxic effect, with 67% cell death at a concentration of 1 µg/mL. A high reduction in cell viability coupled with pronounced cell death was found at 0.1 µg/mL for an Annonaceous seed extract. These results demonstrate that the intake of dietary supplements containing plant material from Annonaceae may be hazardous to health in terms of neurotoxicity.

Keywords

annonaceae graviola pawpaw dietary supplements neurotoxicity

Introduction

Graviola (Annona muricata L.), ata (A. squamosa L.), atemoya (A. cherimola × squamosa Mabb.), biribá (Annona mucosa JACQ), and the North American pawpaw (Asimina triloba (L.) Dunal) belong to Annonaceae, a large plant family that contains about 2,400 species in 108 genera.¹ Annonaceous fruits are served in cafeterias in a worker’s nutrition program in Brazil² and have been recommended for oral rehydration therapy in developing countries such as Nigeria due to its moisture content of >70% and its large variety of electrolytes.³ Both graviola and pawpaw have uses in folk remedies and traditional medicine. Different parts of them have been reported to have potent anticancer properties.^4,5 Capsules claiming to contain extracts of graviola and pawpaw are freely available as dietary supplements in the United States. They are promoted to increase overall health and to provide selective anticancer effects.

Epidemiological studies, however, linked the consumption of Annonaceae to a high prevalence of atypical parkinsonism, in Guadeloupe,^6
–8 in parts of the Afro-Caribbean and Indian population in London⁹ and New Caledonia.¹⁰ In several patients who desisted in their consumption of Annonaceae fruits, the progression of atypical parkinsonism ceased, while in 1 patient the symptoms even disappeared after a change in diet.⁶ Previous studies found that graviola (A. muricata) and the North American pawpaw (Asimina triloba) contain annonacin.^11
–13 A more recent study showed that one sample of lyophilized pawpaw fruits contained not just 7.72 mg/g annonacin but also 0.162 mg/g squamocin.¹⁴ Moreover, all of the dietary supplements and fruits analyzed in the current study were found to contain annonacin and/or squamocin when analyzed by LC-MS/MS.¹⁵ NMR analysis also found that they all contained the α,β-unsaturated γ-lactone toxicophore that is present in neurotoxic acetogenins.¹⁶

Annonacin, extracted from graviola, was shown to be toxic for embryonic rat primary mesencephalic¹⁷ and striatal neurons.^18,19 North American pawpaw extracts were toxic in primary rat cortical neurons.¹³ Moreover, annonacin induces neurodegeneration in rats²⁰ and increases tau pathology in tau transgenic mice.²¹

Annonacin belongs to the group of Annonaceous acetogenins. These are highly lipophilic compounds and potent inhibitors of the mitochondrial nicotinamide adenine dinucleotide : ubiquinone oxidoreductase, the complex I of the respiratory chain.^19,22 Similar to annonacin, other acetogenins cause tau pathology and cell death in primary striatal cultures of embryonic rats in correlation with their potential to inhibit complex I.²²

Besides the strong evidence that complex I inhibition is responsible for the neurotoxic effect of Annonaceous acetogenins in vitro, also other mechanisms for cytotoxicity are discussed, including modulating histone H3 phosphorylation²³ or complexation with calcium ions.²⁴ Moreover, besides aceto-genins, Annonaceous plants contain other toxic substances, such as specific alkaloids. However, these have been shown to be 1000 times less toxic to neuronal cells than acetogenins.¹¹ It is unknown whether freely available dietary supplements claiming to contain graviola or pawpaw exert neurotoxicity.

The aim of this study was to determine the biological effects of Annonaceae containing dietary supplements on Lund human mesencephalic neurons (LUHMES) in terms of cell viability and neurotoxicity and to compare these data to results of fruit pulp and seed extracts of different Annonaceous species.

Materials and Methods

Test Materials and Chemicals

Four dietary supplements were obtained over the Internet. Two of them (S1 and S2) were labeled as containing dried graviola leaves and stems. Another was labeled as containing pawpaw twigs (S3). The fourth (S4) was labeled as containing dried graviola leaves. Graviola fruit pulps (A. muricata) were collected in São Paulo (Brazil). Atemoya (A. squamosa × cherimola) fruit pulp and seeds were collected in Belém (Pará, Brazil) and Pilar (São Paulo, Brazil). Ata (A. squamosa) and biribá (A. mucosa) fruit pulps were from Belém (Pará, Brazil). The pulp and seeds were removed from the fruits, lyophilized, and stored frozen until ready for extraction. High-purity ethyl acetate was obtained from Honeywell Burdick & Jackson (Muskegon, Michigan).

Extraction of Samples

About 10 g of each sample was mixed with HydroMatrix (Sigma Aldrich, St Louis, Missouri) to reach the 100 mL mark in the sample cell used in an accelerated solvent extractor (Dionex ASE 300; ThermoFisher Scientific, Sunnyvale, California). Then, 40 mL ethyl acetate was added. The temperature and pressure were increased to 100°C and 10.3 MPa (100 atm) over 3 minutes (static time). Next, the solvent was flushed out into a collection vessel. This was repeated 3 more times to obtain a total volume of 160 mL. The solvent was evaporated off, and the remaining residue was weighed. Stock solutions of portions of the residues obtained from the various extracts were prepared in dimethyl sulfoxide (DMSO; Sigma-Aldrich, St. Louis, Missouri). The drug–extract ratio (DER) for each extract was calculated. The DER states how many parts of starting material are needed to produce 1 part of extract. It helps to compare different extracts and can be used to draw conclusions on the enrichment of extract ingredients.

Cell Culture

The LUHMES cells were used for the study of cell viability and toxicity and were derived from human mesencephalic tissue. They were immortalized by cloning in a v-myc vector under tetracycline control and can be differentiated to postmitotic dopaminergic neurons.^25
–27 The cells were proliferated on flasks (Nunclon Δ surface; NUNC A/S, Roskilde, Denmark) coated with poly-L-lysine (PLL; Sigma-Aldrich) 0.1 mg/mL at +4°C overnight in growth medium (Dulbecco modified Eagle medium [DMEM]/F12 [Sigma-Aldrich] with 1% N-2 supplement [Life Technologies, Carlsbad, California], and 0.04 µg/mL basic fibroblast growth factor [R&D Systems, Minneapolis, Minnesota]). The cells were always kept at 37°C. For experiments, cells were seeded in growth medium on multiwell plates (Nunclon DELTA surface; NUNC A/S) coated with 0.1 mg/mL PLL at 4°C overnight and washed 3 times with phosphate-buffered saline (PBS; Life Technologies) followed by coating with 5 µg/mL fibronectin (Sigma-Aldrich) for at least 6 hours in the incubator (37°C, 5% CO₂). The wells were then washed once with PBS and air dried before plating at a density of 55 000/cm². The cells were kept at 37°C, 5% CO₂, and 100% humidity. Twenty-four hours after plating, medium was changed to differentiation medium (DMEM/F12 with 1% N-2 supplement, 1 µg/mL tetracycline, 0.49 mg/mL dibutyryl cyclic AMP [Sigma-Aldrich] and 2 ng/mL GDNF [R&D Systems]). The cells were differentiated for 8 days with changes of 50% of the differentiation medium every second day.

Treatment of Cells

After 8 days of differentiation, the medium was changed to low glucose differentiation medium composed of DMEM with 250 µmol/L of glucose with the same supplements as described above. This was done in accordance with previous studies investigating the toxic effects of acetogenins or other complex I inhibitors in primary cell cultures^18,19 and LUHMES cells.²⁸

After medium change, the cells were kept at 37°C, 5% CO₂, for at least 1 hour. Thereafter, the cells were treated with ethyl acetate extracts in different concentrations dissolved in DMSO (AppliChem, Darmstadt, Germany) and diluted into low glucose differentiation medium. The concentration of DMSO never exceeded 0.1%. Negative control cells were treated with the same amount of low glucose differentiation medium containing the same amount of DMSO. Experiments were repeated 3 times with 4 different wells treated per condition in each experiment.

Measurement of Toxicity

The 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was performed 48 hours after incubation with different samples as previously described.^28,29 MTT of 5 mg/mL (Sigma-Aldrich) was dissolved in PBS. This solution was added 1:10 to the medium of each well of the cell culture plate to a concentration of 0.5 mg/mL. After 1 hour of incubation at 37°C, the MTT solution and medium were removed and the plate was frozen at −80°C for 24 hours. Thereafter, the plate was thawed and 300 µL of DMSO were added to each well. After shaking for 5 minutes, 100 µL from each well was transferred to a transparent U-shaped-bottom 96-well plate, and the absorption was measured at a wavelength of λ = 590 nm using a FLUOstar Omega plate reader (BMG Labtech, Ortenberg, Germany) in absorbance mode. Values were normalized to negative control cells.

As a second assay, the lactate dehydrogenase (LDH) released from dying cells into the culture medium was measured 48 hours after treatment of the cells with the extracts using the CytoTox-ONE Homogeneous Membrane Integrity Assay (Promega, Fitchburg, Wisconsin) according to the manufacturer’s instructions with a FLUOstar Omega plate reader (BMG Labtech) in fluorescence mode with an excitation of λ = 544 nm and an emission of λ = 590 nm. Data were normalized to data from cells lysed by treating them with the lysis buffer of the CytoTox-ONE Homogeneous Membrane Integrity Assay for 30 minutes as positive controls (maximum LDH-release).

Statistics

One-way analysis of variance (ANOVA) followed by Tukey post hoc test was used to determine significance using Prism 6 for Windows (GraphPad Sofware Inc, La Lolla, California).

Results

Pressurized Liquid Extraction

Using an accelerated solvent extractor, we carried out a pressurized liquid extraction with ethyl acetate and were able to receive extraction yields between 1.9% and 23% (Table 1). Ethyl acetate preferably solubilizes lipophilic substances in the plant material and is known to extract neurotoxic ingredients of Annonaceae as annonacin.¹³ Expectedly, the DERs that we determined for the fruit pulp ethyl acetate extracts are very high (Table 1), indicating that the fruit pulps only contain a low amount of lipophilic substances. By contrast, the extracts of the leaves, stems, twigs, and seed material had a relatively low DER, showing that a larger amount of the contained substances are more lipophilic.

Table 1.

Overview of Produced Ethyl Acetate Extracts.

Extracted sample	Drug–extract ratio^a	Extraction yield^b
Dietary supplements
1. Leaves and stems from Graviola	9.1: 1	11.0%
2. Leaves and stems from Graviola	13.9: 1	7.2%
3. Twigs from North American Pawpaw	6.3: 1	16.0%
4. Leaves from Graviola	4.6: 1	22.0%
Annonaceae pulp extracts
Graviola (Annona muricata L.)	52.6: 1	1.9%
Atemoya (A. squamosa × cherimola Mabb.)	33.3: 1	3.0%
Ata (A. squamosa L.)	33.1: 1	3.0%
Biribá (A. mucosa Jacq.)	54.1: 1	1.9%
Annonaceae seed extract
Atemoya (A. squamosa × cherimola Mabb.)	4.8: 1	23.0%

^aThe drug–extract ratio states the ratio between the quantity of herbal raw material used for extraction and the herbal extract obtained.

^bExtraction yield is the amount of extract related to the starting material (plant material) [m/m].

Neurotoxicity of Dietary Supplements

In order to test the toxicity of extracts of 4 different dietary supplements claiming to contain different parts of Annonaceous plants, graviola (A. muricata L.) and the North American pawpaw (Asimina triloba (L.) Dunal), we measured cell viability with the MTT assay and cell death by measurement of LDH release into the cell culture medium. Extracts of the fruits were analyzed previously by NMR and found to contain the methylated γ-butyrolactone toxicophore that is present in acetogenins.³⁰ For the purpose of analysis, dry methanol at 100°C and 10 MPa (100 atm) pressure in a sealed container was found to be best for solubilizing annonacin and squamocin.¹⁴ Subsequent LC-MS/MS analyses of the fruits, atemoya seeds, and dietary supplements that were extracted this way found the concentrations of annonacin and squamocin in µg/g dry weight (µg/g-dw) that are shown in Table 2.¹⁵ The LC-MS/MS analyses used a C₁₈ column and gradient elution to separate the analytes. The gradient elution started with 4:1 CH₃OH–H₂O (v/v) containing 0.1% formic acid for the first 25 minutes. This was changed linearly to 0.1% formic in 49:1 CH₃OH–H₂O (v/v) from 25 to 30 minutes and held at this mixture from 30 to 35 minutes, followed by a step change back to 4:1 CH₃OH–H₂O (v/v) that was held from 35 to 45 minutes. Annonacin and squamocin were detected and quantified as H⁺, Na⁺, and formate (HCOO⁻) ions in an Orbitrap Exactive HRMS (ThermoFisher) with heated electrospray ionization that was able to switch rapidly from positive to negative ionization. So, for annonacin, masses in the range of 597.47071 to 597.47549, 619.45251 to 619.45747, and 641.46397 to 641.46911 were monitored for [M+H]⁺, [M+Na]⁺, and [M+HCOO]⁻, respectively. For squamocin, masses in the range of 623.48628 to 623.49126, 645.46808 to 645.47324, and 667.47985 to 667.48519 Daltons were monitored for [M+H]⁺, [M+Na]⁺, and [M+HCOO]⁻, respectively. Annonacin and squamocin standards (kindly provided by Pierre Champy, University of Paris) were used to identify and quantify the peaks due to annonacin and squamocin that eluted at 19.74 and 22.51 minutes, respectively. These are the only 2 acetogenins that were available to us as purified standards. However, many more neurotoxic acetogenins have been seen in graviola and other fruits in the Annonaceae family.^11,12,19,22 So, although ata and biribá fruit pulps had much less annonacin and squamocin than graviola or atemoya, they had similar neurotoxicities. This could be due to other, not yet quantified acetogenins. Finally, the apparently low neurotoxicity of dietary supplement S4 should not be taken as an indication that fresh graviola leaves are not very toxic. In a separate study, freshly lyophilized graviola leaves were found to have 305.6 ± 28.3 µg/g-dw annonacin and 17.4 ± 0.89 µg/g-dw squamocin.³¹ 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.³¹

Table 2.

Concentrations of Annonacin and Squamocin Found in Samples Evaluated in This Study.¹⁵

	Annonacin, µg/g-dw	Squamocin, µg/g-dw
Dietary supplements
1. Leaves and stems from Graviola	337	34
2. Leaves and stems from Graviola	231	37
3. Twigs from North American Pawpaw	135	67
4. Leaves from Graviola	nd	35
Annonaceae pulp extracts
Graviola (Annona muricata L.)	768	4.5
Atemoya (A. squamosa × cherimola Mabb.)	3.8	68
Ata (A. squamosa L.)	2.2	0.7
Biribá (A. mucosa Jacq.)	1.8	4.2
Annonaceae seed extract
Atemoya (A. squamosa × cherimola Mabb.)	454	14200

Abbreviation: nd, none detected.

The treatment of LUHMES cells with extracts from 3 of the 4 dietary supplements led to a reduced cell viability. Already at a concentration of 0.1 µg/mL, the extract of supplement S1 reduced the MTT level to 57.5% ± 19.8% of that of negative control. At concentrations of 1.0 and 10 µg/mL, the MTT level was further reduced to 12.3% ± 2.1% and 9.0% ± 0.7% (P < .0001 vs negative control, Figure 1B [S1]). With 1 µg/mL of the extract from supplement S2, the MTT level was reduced to 48.4% ± 4.6% (P < .0001 vs negative control), while a concentration of 10 µg/mL further reduced the MTT level to 13.4% ± 2.4% (P < .0001 vs negative control; Figure 1B [S2]). While 1 µg/mL of the extract from supplement S3 did not decrease the MTT signal, a concentration of 10 µg/mL reduced the MTT level to 37.1% ± 10.9% (P < .0001 vs negative control; Figure 1B [S3]). Concentrations of up to 10 µg/mL of supplement S4 did not decrease the cell viability (Figure 1B [S4]), indicating that this supplement was not toxic in the tested range. The 3 supplements (S1, S2, and S3) that reduced cell viability all led to cell death as measured by LDH release into the cell culture medium. In correspondence to the data from the MTT assay, the extract from supplement S1 was the most toxic, leading to a significant LDH increase in 37.5% ± 2.5% at a concentration of 1 µg/mL, compared to the LDH level in negative controls of 21.1% ± 0.3% (P = .003). Ten µg/mL of this extract led to an increase in the LDH level to 78.9% ± 4.3% (P < .0001 vs negative control), indicating almost maximal cell death (Figure 1B [S1]). While treatment with 1 µg/mL of the extract from supplement S2 was not toxic, we observed a significant increase in the LDH level with 10 µg/mL to 87.0% ± 2.6% compared to 23.5% ± 1.0% in negative controls (P < .0001; Figure 1B [S2]). We also did not observe significant cell death with 1 µg/mL of the extract from supplement S3, while 10 µg/mL of this extract led to an increase in the LDH level to 62.7% ± 12.3% compared to 27.1% ± 2.5% in negative controls (P = .001, Figure 1B [S3]). In correspondence to the data from the MTT assay, the extract from supplement S4 did not induce cell death in the tested concentration range (Figure 1B [S4]).

Figure 1.

Toxicity of Annonaceous ethyl acetate extract. Human mesencephalic (LUHMES) cells were treated for 48 hours with different concentrations of ethyl acetate extracts from 4 Annonaceae-containing dietary supplements (S1–S4) and from Annonaceous fruit pulp (a: Annona muricata, b: Annona squamosa, c: Annona mucosa, d: Annona cherimola × squamosa) and seeds (e: Annona cherimola × squamosa). Cell viability was measured by relative MTT levels (A) and cell death measured by LDH release (B) after treatment with 4 dietary supplements and corresponding data after treatment with Annonaceous pulp and seeds extracts (MTT: C, LDH: D). Data are normalized to the MTT signal measured in DMSO-treated control cells. Data from the LDH release assay are normalized to lysed cells (maximal cell death) as positive controls. For details and statistics, see text. LDH indicates lactate dehydrogenase; DMSO, dimethyl sulfoxide; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide.

Neurotoxicity of Annonaceous Fruit Pulps or Seeds

We measured the MTT levels in LUHMES cells and the release of LDH into the cell culture medium after treatment with extracts of fruit pulps in the same way that we tested the extracts from the 4 dietary supplements. We tested fruit pulp extracts of A. muricata L., A. squamosa L., A. mucosa Jacq., A. cherimola × squamosa Mabb and a seed extract from A. cherimola × squamosa Mabb. The pulp extract of A. muricata already reduced the MTT levels in LUHMES cells to 10.7% ± 1.5% (P < .0001 vs negative control) at a low concentration of 1 µg/mL, (Figure 1C [a]). In a similar way, the extract from pulp of A. squamosa reduced the MTT levels to 14.4% ± 5.4% at a concentration of 1 µg/mL (P = .0009 vs negative control) and to 4.6% ± 2.5% at 10 µg/mL (P = .0003 vs negative control; Figue 1C [b]). The extract from A. mucosa pulp did not reduce the MTT levels in LUHMES cells at a concentration of 1 µg/mL. However, a concentration of 10 µg/mL did reduce MTT to 6.3% ± 1.0% (P < .0001 vs negative control; Figure 1C [c]). In a similar way, treatment with 10 µg/mL of the extract from A. cherimola × squamosa pulp led to a reduction in the MTT levels to 12.7% ± 3.7% (P < .0001 vs negative control; Figure 1C [d]). In addition to the pulp of Annonaceous fruits, we also tested an extract from seeds of A. cherimola × squamosa. Interestingly, this extract showed a much higher toxicity than the extracts from the pulp. A concentration as low as 0.01 µg/mL slightly reduced the MTT level to 86.8% ± 5.6% (P < .05 vs negative control), while 0.1 µg/mL of this extract reduced the MTT level to 4.0% ± 0.8% (P < .0001 vs negative control; Figure 1C [e]). Likewise, the extracts from Annonaceous fruit pulps and seeds led to cell death as measured by LDH release into the cell culture medium. One µg/mL of the extract from A. muricata pulp increased the LDH level to 67.0% ± 5.6% compared to 17.8% ± 1.2% in the negative control (P < .0001). Ten µg/mL of this extract increased the LDH level to 78.1% ± 3.9% (P < .0001 vs negative control; Figure 1D [a]). The treatment with the extract of A. squamosa pulp led to cell death at 10 µg/mL, increasing the LDH level at this concentration to 68.7% ± 2.1% compared to 15.9% ± 0.8% in negative controls (P < .0001; Figure 1D [b]). In a similar manner, treatment with 10 µg/mL of the pulp extract from A. mucosa increased the LDH level to 75.5% ± 1.5% compared to 21.1% ± 1.5% of the negative control (P < .0001; Figure 1D [c]), and treatment with 10 µg/mL of the extract from A. cherimola × squamosa pulp increased the LDH level to 52.8% ± 3.9% compared to 18.0% ± 1.1% of the negative control (P = .007, Figure 1D [d]). In the same manner as observed in the MTT assay, the extract from A. cherimola × squamosa seeds induced cell death at even smaller concentrations than the pulp extracts. Treatment of LUHMES cells with 0.1 µg/mL led to an increase in the LDH level to 60.8% ± 4.3% compared to 11.3% ± 2.4% in negative controls (P = .0001; Figure 1D [e]).

Discussion

The consumption of herbal tea and fruits from the Annonaceae family (A. muricata and A. squamosa) was linked to atypical parkinsonism on Guadeloupe.^6
–8 Previous studies showed that acetogenins, which are long-chained lipophilic inhibitors of the mitochondrial complex I and found in Annonaceous species, are highly neurotoxic in vitro ^18,19 and in vivo.²¹ Nevertheless, different parts of Annonaceae are sold as dietary supplements in the United States and are promoted to have anticancer effects and to increase general health. In the present study, we demonstrated that extracts from these freely available dietary supplements are toxic to neuronal cells by an extent comparable to the neurotoxicity of extracts from fruit pulp and seeds of different Annonaceous species (A. muricata L., A. squamosa L., A. mucosa Jacq., A. cherimola × squamosa Mabb.).

We used ethyl acetate for the extraction procedure to solubilize lipophilic ingredients of Annonaceous plant material including neurotoxic acetogenins.¹³ The amount of raw material used to obtain a certain amount of extract was lower in the dietary supplements (DERs 4.6:1 to 13.9:1) than in the pulp of 4 different Annonaceous species (DER 33.1:1 to 52.6:1). The reason for that was likely that fruit pulp contains more hydrophilic material than the dietary supplements that contained leaves, stems, and twigs of the plants. However, our data suggest that the content of lipophilic ingredients from Annonaceous species, including potentially toxic lipophilic acetogenins, is higher in the plant material used for the dietary supplements than for pulp fruit extracts. Dietary supplements S1, S3, and S4 had an extraction yield similar to Atemoya seeds. This implies that they contain similar amounts of lipophilic components.

The toxicities of the extracts from dietary supplements and Annonaceous fruit extracts and seeds were measured with 2 different assays, MTT and LDH release. The MTT assay is used as a cell viability assay that is affected by changes in mitochondrial oxidative phosphorylation as well as other cellular pathways, including glycolysis.³² However, it does not necessarily indicate cell death. On the other hand, LDH is only released into the cell medium by dying cells, when cell membranes become leaky. Therefore, it is a good tool to measure cell death.

Of the 4 dietary supplements, 2 (S1 and S2) led to a gross loss of cell viability, as measured by MTT assay, at a concentration of 1 µg/mL. Supplement S1 led to a reduction in the signal by 87.7% ± 2.1% and S2 by 51.7% ± 4.6% (Figure 1A). Moreover, supplement S1 already showed significant cell death at this concentration (Figure 1B). With the higher concentration of 10 µg/mL, 3 supplements showed marked reduction in cell viability and cell death. This observation is in line with previous observations made with several acetogenins that concentrations needed to impair the cellular energy metabolism were slightly lower than concentrations that lead to cell death.¹⁹ This suggests that impairment of energy metabolism can be compensated for a while before ultimately resulting in cell death. The concentrations of the extracts from the dietary supplements which led to reduced cell viability and cell death were in the same range as the concentrations of the extracts from pulp of A. squamosa and A. cherimola × squamosa, which also induced cell death at 10 µg/mL. On the other hand, the extract from pulp of A. muricata was slightly more toxic, leading to marked cell death at 1 µg/mL. Interestingly, the extract from seeds of A. cherimola × squamosa seeds already led to marked decrease in cell viability and induced cell death at 0.1 µg/mL. This suggests that the seeds contain even more toxic ingredients than the fruit pulp. Also, dietary supplements S1 and S2 (containing leaves plus stem from A. muricata) were toxic at the tested concentrations, while dietary supplement S4 that contains leaves but not stems from A. muricata was not toxic at the toxic concentrations. This suggests that the stems contained more toxic compounds than the leaves.

In advertising texts, the ingredients of the Annonaceous dietary supplements are described as generally healthy, reasoned due to their long tradition in herbal medicine. Some advertisements even concede that the acetogenins, which are present in Annonaceous fruits, impair the ATP production in mitochondria. It is suggested that this might be the case for specific cell types, especially cancer cells. This is supported by several reports showing that extracts from Annonaceae are toxic to cancer cells.^33

–36 In line with that, it was recently reported that A. muricata extracts inhibited the growth of a promyelocytic leukemia cell line in vitro with an IC₅₀ between 6 and 49 µg/mL.³⁷ The data may not directly be comparable to the data we obtained here. A different solvent was used to extract the samples, and different cell culture media were used to grow tumor cell lines and neuronal cells. Still, it is worth noting that the concentrations of A. muricata extract that led to growth inhibition in the promyelocytic leukemia cell line were much higher than the concentrations of A. muricata extract that led to reduced cell viability and cell death in LUHMES neurons here.

The effects of acetogenins on the cell viability of tumor cells, which are essentially dependent on high levels of energy, are undisputed. However, there is currently no evidence that the cytotoxic effect of acetogenins is selective or specific for cancer cells. In contrast, the data presented here and by previous reports show that extracts from Annonaceae or purified acetogenins are especially toxic to neuronal cells. The toxic effects of annonacin on rat primary mesencephalic¹⁷ and striatal cells^18,19 as well as on human neuronal cells²⁸ have been shown previously. Moreover, in vitro experiments revealed that treatment with Annonaceae acetogenins increased tau phosphorylation^18,19 and led to an isoform shift from 3R-tau to 4R-tau,²⁸ both contributing to the pathophysiology of neurodegenerative tauopathies.

We treated the cells with ethyl acetate extracts from dietary supplements, fruit pulp, or seeds of Annonaceae. These extracts contain a complex mixture of lipophilic substances, and the toxic effect that we observed can therefore not exclusively be attributed to acetogenins. A previous report showed that crude ethyl acetate extracts of A. muricata pulp contain up to 10% annonacin,¹³ but it is very likely that the other 90% contain other potentially toxic compounds, for example, other acetogenins or specific alkaloids that contributed to the observed effect. By the intake of the dietary supplements, one ingests all toxic ingredients, regardless of the specific composition. Because of this, we investigated full extracts and not solely isolated compounds in the present study.

In summary, our data clearly demonstrate that dietary supplements containing twigs or leaves from Annonaceous species are toxic to neuronal cells in the same manner as fruit pulp extracts from Annonaceae that are known to contain huge amounts of neurotoxic acetogenins. This finding should be taken serious because these dietary supplements are advertised as being health enhancing and helpful against malignant diseases. Moreover, they are designated for chronical intake. The consumption of non-evidence–based dietary supplements of Annonaceae could contribute to the occurrence of neurodegenerative disorders in the same way as the consumption of Annonaceous fruits in Guadeloupe.

Footnotes

Acknowledgments

We acknowledge support by the German Research Foundation (DFG, HO2402/6-2), the CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico do Brasil), FAPERJ (Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro), and FAPESPA (Fundação de Amparo à Pesquisa do Estado do Pará).

Authors’ Note

This work should not be taken as reflecting FDA policy or regulations. Matthias Höllerhage, and Thomas W. Rösler equally contributed to this work.

Author Contribution

Höllerhage, Matthias, Höglinger, Guenter, and Rösler, Thomas contributed to conception and design, contributed to acquisition, analysis, and interpretation. Berjas, Magda contributed to conception and design and analysis. Luo, Rensheng contributed to conception, acquisition, and interpretation. Tran, Kevin contributed to conception, acquisition, analysis, and interpretation. Richards, Kristy contributed to conception and analysis; Sabaa-Srur, Armando, Maia, JGM, De Moraes, MR, and Godoy, HT contributed to conception and design and acquisition. Smith, RE contributed to conception and design, acquisition, analysis, and interpretation. All authors drafted manuscript, critically revised manuscript, gave final approval, and agreed to be accountable for all aspects of work ensuring integrity and accuracy.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Chatrou

Pirie

Erkens

RHJ

. A new subfamilial and tribal classification of the pantropical flowering plant family Annonaceae informed by molecular phylogenetics. Bot J Linn Soc. 2012;169(1):5–40.

Lustosa

Barroso

Vareda

Carvalho

. Comparative analysis of fruit and vegetable consumption in a food service unit: is it possible to influence the choices? Nat Prod J. 2012;2(3):161–163.

Enweani

Obroku

Enahoro

Omoifo

. The biochemical analysis of soursop (Annona muricata L.) and sweetsop (A. squamosa L.) and their potential use as oral rehydration therapy. J Food Agric Environ. 2004;2(1):39–43.

Chen

. Antitumor activity of annonaceous acetogenins in HepS and S180 xenografts bearing mice. Bioorg Med Chem Lett. 2012;22(8):2717–2719.

Hamizah

Roslida

Fezah

. Chemopreventive potential of Annona muricata L leaves on chemically-induced skin papillomagenesis in mice. Asian Pac J Cancer Prev. 2012;13(6):2533–2539.

Caparros-Lefebvre

Elbaz

. Possible relation of atypical parkinsonism in the French West Indies with consumption of tropical plants: a case-control study. Caribbean Parkinsonism Study Group: a case-control study. Caribbean Parkinsonism Study Group. Lancet. 1999;354:281–286.

Caparros-Lefebvre

Lees

. Atypical unclassifiable parkinsonism on Guadeloupe: an environmental toxic hypothesis. Mov Disord. 2005;20(suppl 12);S114–S118.

Lannuzel

Höglinger

Verhaeghe

. Atypical parkinsonism in Guadeloupe: a common risk factor for two closely related phenotypes? Brain 2007;130(pt 3):816–827.

Chaudhuri

Brooks

. Atypical parkinsonism in Afro-Caribbean and Indian origin immigrants to the UK. Mov Disord. 2000;15(1):18–23.

10.

Angibaud

Gaultier

Rascol

. Atypical parkinsonism and Annonaceae consumption in New Caledonia. Mov Disord. 2004; 19: 603–4.

11.

Champy

Melot

Guérineau

. Quantification of acetogenins in Annona muricata linked to atypical parkinsonism in guadeloupe. Mov Disord. 2005;20(12):1629–1633.

12.

Zhou

Lewis

. Quantitative evaluation of annonaceous acetogenins in monthly samples of paw paw (Asimina triloba) twigs by liquid chromatography electrospray ionization tandem mass spectrometry. Phytochem Analysis. 1999;10(10):32–38.

13.

Potts

Luzzio

Smith

Hetman

Champy

Litvan

. Annonacin in Asimina triloba fruit: implication for neurotoxicity. Neurotoxicology 2012;33(1):53–58.

14.

Levine

Richards

Tran

Luo

Thomas

Smith

. Determination of neurotoxic acetogenins in pawpaw (Asimina triloba) fruit by LC-HRMS. J Agr Food Chem. 2015;63(4):1053–1056.

15.

Smith

Levine

Tran

. LC-MS and NMR Analysis of Neurotoxic Fruits in the Annonaceae Family. Paper presented at: 2014 Midwest Regional Meeting American Chemical Society; 2014. https://www.researchgate.net/publication/268516208_LC-MS_and_NMR_Analysis_of_Neurotoxic_Fruits_in_the_Annonaceae_Family.

16.

Luo

Maia

de Moraes

. NMR analysis of potentially neurotoxic annonaceous fruits. Nat Prod J. 2013;3:230–241.

17.

Lannuzel

Michel

Höglinger

. The mitochondrial complex I inhibitor annonacin is toxic to mesencephalic dopaminergic neurons by impairment of energy metabolism. Neuroscience. 2003;121(2):287–296.

18.

Escobar-Khondiker

Höllerhage

Muriel

. Annonacin, a natural mitochondrial complex I inhibitor, causes tau pathology in cultured neurons. J Neurosci. 2007;27(29):7827–7837.

19.

Höllerhage

Matusch

Champy

. Natural lipophilic inhibitors of mitochondrial complex I are candidate toxins for sporadic neurodegenerative tau pathologies. Exp Neurol. 2009;220(1):133–142.

20.

Champy

Höglinger

Féger

. Annonacin, a lipophilic inhibitor of mitochondrial complex I, induces nigral and striatal neurodegeneration in rats: possible relevance for atypical parkinsonism in Guadeloupe: possible relevance for atypical parkinsonism in Guadeloupe. J Neurochem. 2004;88(1):63–69.

21.

Yamada

Respondek

Müssner

. Annonacin, a natural lipophilic mitochondrial complex I inhibitor, increases phosphorylation of tau in the brain of FTDP-17 transgenic mice. Exp Neurol. 2014;253:113–125.

22.

Bermejo

Figadere

Zafra-Polo

Barrachina

Estornell

Cortes

. Acetogenins from Annonaceae: recent progress in isolation, synthesis and mechanisms of action. Nat Prod Rep. 2005;22(2):269–303.

23.

Lee

Lin

Chang

. Squamocin modulates histone H3 phosphorylation levels and induces G1 phase arrest and apoptosis in cancer cells. BMC Cancer. 2011;11:58.

24.

Liaw

Liao

Chen

. The calcium-chelating capability of tetrahydrofuranic moieties modulates the cytotoxicity of annonaceous acetogenins. Angew Chem Int Ed Engl. 2011;50(34):7885–7891.

25.

Höllerhage

Goebel

de Andrade

. Trifluoperazine rescues human dopaminergic cells from wild-type α-synuclein-induced toxicity. Neurobiol Aging. 2014;35(7):1700–1711.

26.

Lotharius

Barg

Wiekop

Lundberg

Raymon

Brundin

. Effect of mutant alpha-synuclein on dopamine homeostasis in a new human mesencephalic cell line. J Biol Chem. 2002;277(41):38884–38894.

27.

Lotharius

Falsig

van Beek

. Progressive degeneration of human mesencephalic neuron-derived cells triggered by dopamine-dependent oxidative stress is dependent on the mixed-lineage kinase pathway. J Neurosci. 2005;25(27):6329–6342.

28.

Bruch

de Andrade

Höglinger

. Mitochondrial complex 1 inhibition increases 4-repeat isoform tau by SRSF2 upregulation. PLoS One. 2014;9(11):e113070.

29.

Hansen

Nielsen

Bergm

. Re-examination and further development of a precise and rapid dye method for measuring cell growth/cell kill. J Immunol Methods. 1989;119(2):203–210.

30.

Luo

Maia

de Moraes

. NMR analysis of potentially neurotoxic annonaceous fruits. Nat Prod J. 2013;3:230–241.

31.

de Moraes

IVM

Ribeiro

PRV

Schmidt

. UPLC-QTOF-MS and NMR analyses of graviola (Annona muricata) leaves. Braz J Pharmacog. Submitted 2015.

32.

Takahashi

Abe

Gotoh

Fukuuchi

. Substrate-dependence of reduction of MTT: a tetrazolium dye differs in cultured astroglia and neurons. Neurochem Int. 2002;40(5):441–448.

33.

McLaughlin

. Paw paw and cancer: annonaceous acetogenins from discovery to commercial products. J Nat Prod. 2008;71(7):1311–1321.

34.

Paul

Gnanam

Jayadeepa

Arul

. Anticancer activity on Graviola, an exciting medicinal plant extract vs various cancer cell lines and a detailed computational study on its potent anti- cancerous leads. Curr Top Med Chem. 2013;13(14):1666–1673.

35.

Wang

Rizwani

Guo

. Annona squamosa Linn: cytotoxic activity found in leaf extract against human tumor cell lines. Pak J Pharm Sci. 2014;27(5 spec no);1559–1563.

36.

Moghadamtousi

Karimian

Rouhollahi

. Annona muricata leaves induce G₁ cell cycle arrest and apoptosis through mitochondria-mediated pathway in human HCT-116 and HT-29 colon cancer cells. J Ethnopharmacol. 2014;156:277–289.

37.

Pieme

Kumar

Dongmo

. Antiproliferative activity and induction of apoptosis by Annona muricata (Annonaceae) extract on human cancer cells. BMC Complement Altern Med. 2014;14:516.

Neurotoxicity of Dietary Supplements from Annonaceae Species

Abstract

Keywords

Introduction

Materials and Methods

Test Materials and Chemicals

Extraction of Samples

Cell Culture

Treatment of Cells

Measurement of Toxicity

Statistics

Results

Pressurized Liquid Extraction

Neurotoxicity of Dietary Supplements

Neurotoxicity of Annonaceous Fruit Pulps or Seeds

Discussion

Footnotes

Acknowledgments

Authors’ Note

Author Contribution

Declaration of Conflicting Interests

Funding

References