**An Investigation of Potential Sources of Nutraceuticals from the Niger Delta Areas, Nigeria for Attenuating Oxidative Stress**

**Lucky Legbosi Nwidu 1,2,\*, Philip Cheriose Nzien Alikwe 3, Ekramy Elmorsy 2,4 and Wayne Grant Carter <sup>2</sup>**


Received: 18 December 2018; Accepted: 15 January 2019; Published: 20 January 2019

**Abstract: Background:** Diets rich in fruits, vegetables, and medicinal plants possess antioxidants potentially capable of mitigating cellular oxidative stress. This study investigated the antioxidant, anti-acetylcholinesterase (AChE), and total phenolic and flavonoids contents (TPC/TFC) of dietary sources traditionally used for memory enhancing in Niger Delta, Nigeria. **Methods:** *Dacroydes edulis* methanolic seed extract (DEMSE), *Cola lepidota* methanolic seed extract (CLMSE), *Terminalia catappa* methanolic seed extract (TeCMSE), *Tricosanthes cucumerina* methanolic seed extract (TrCMSE), *Tetrapleura tetraptera* methanolic seed extract (TTMSE), and defatted *Moringa oleifera* methanolic seed extract (DMOMSE); *Dennettia tripetala* methanolic fruit extract (DTMFE), *Artocarpus communis* methanolic fruit extract (ACMFE), *Gnetum africana* methanolic leaf extract (GAMLE), *Musa paradisiaca* methanolic stembark extract (MPMSE), and *Mangifera indica* methanolic stembark extract (MIMSE) were evaluated for free radical scavenging antioxidant ability using 2,2-Diphenyl-1-picrylhydrazyl (DPPH), reducing power capacity (reduction of ferric iron to ferrous iron), AChE inhibitory potential by Ellman assay, and then TPC/TFC contents determined by estimating milli-equivalents of Gallic acid and Quercetin per gram, respectively. **Results:** The radical scavenging percentages were as follows: MIMSE (58%), MPMSE (50%), TrCMSE (42%), GAMLE (40%), CLMSE (40%), DMOMSE (38%), and DEMFE (37%) relative to β-tocopherol (98%). The highest iron reducing (antioxidant) capacity was by TrCMSE (52%), MIMSE (40%) and GAMLE (38%). Extracts of MIMSE, TrCMSE, DTMFE, TTMSE, and CLMSE exhibited concentration-dependent AChE inhibitory activity (*p* < 0.05–0.001). At a concentration of 200 μg/mL, the AChE inhibitory activity and IC50 (μg/mL) exhibited by the most potent extracts were: MIMSE (≈50%/111.9), TrCMSE (≈47%/201.2), DTMFE (≈32%/529.9), TTMSE (≈26%/495.4), and CLMSE (≈25%/438.4). The highest TPC were from MIMSE (156.2), TrCMSE (132.65), GAMLE (123.26), and CLMSE (119.63) in mg gallic acid equivalents/g, and for TFC were: MISME (87.35), GAMLE (73.26), ACMFE (69.54), CLMSE (68.35), and TCMSE2 (64.34) mg quercetin equivalents/gram. **Conclusions**: The results suggest that certain inedible and edible foodstuffs, most notably MIMSE, MPMSE, TrCMSE, GAMLE, and CLMSE may be beneficial to ameliorate the potentially damaging effects of redox stress.

**Keywords:** acetylcholinesterase inhibitor; antioxidants; memory enhancers; nutraceuticals

#### **1. Introduction**

Oxidative stress is associated with a number of diseases and arises as a consequence of an imbalance between reactive oxygen species (ROS), reactive nitrogen species (RNS), and their dissipation via enzymatic and non-enzymatic mechanisms [1]. Many factors including UV-irradiation, industrial emissions, tobacco smoke, licit or illicit drug usage, heavy metal exposure, inorganic and organic contaminants, and xenobiotics are among the potential exogenous sources of ROS generation in regions such as the Niger Delta, Nigeria. Endogenous sources of ROS include those generated via normal cellular metabolism and via pathological means [2]. The chronic exposure to exogenous factors and/or activation of endogenous means can provoke oxidative insults that may activate stress response pathways including inflammation, cytokine secretion, and apoptosis that might contribute to wide range of pathophysiological events [1,3].

Nutraceuticals are used as dietary supplements, but also as auxiliaries for the perceived prevention and/or treatment of a variety of diseases and disorders. The role of consumption of nutraceuticals and their protective effects in mammals and humans against diseases such as neurodegeneration that includes an elevation of oxidative stress have been reviewed [4,5]. Natural antioxidants in foods may exhibit protective antioxidant effects and thereby aid in the reduction of premature mortality [6,7]. One such group of antioxidants is flavonoids, compounds that are ubiquitous in many edible plants [8]. Collectively, there are an extensive number of low molecular weight organic compounds, including polyphenols and flavonoids that have been termed secondary metabolites or phytochemicals that can be specific to each plant. These arrays of secondary metabolites can exhibit a wide spectrum of pharmacological effects including provision of cellular antioxidant activities capable of scavenging damaging free radicals. This endowed useful activity provides potential functional benefits to humans beyond basic nutrition, and could be exploited as commercial sources of nutraceutical formulations [9]. These potential health benefits provide an impetus for subjecting plant extracts and fractions for scrutiny to elucidate and quantify their respective antioxidant and other health benefiting abilities.

In the Niger Delta region of Nigeria residents may benefit from the wide patronage and chronic consumption of numerous endogenous edible seeds, fruits, nuts, pods, green leafy vegetables, herbs, spices, and crops. These are commonly consumed in either raw or in cooked forms in various cuisines. Many of these foodstuffs have yet to be evaluated for their ability to mitigate oxidative stress. Additionally, there is an association of oxidative stress and a cholinergic deficit with neurodegenerative diseases such as Alzheimer's disease and Parkinson's diseases [10–14]. Hence, the intake of appropriate foodstuffs able to combat this cellular damage and loss of neuronal functionality may limit the development or indeed propagation of neurodegenerative disease [15–18].

Therefore, this study investigated extracts of *Dacroydes edulis* methanolic seed extract (DEMSE), *Cola lepidota* methanolic seed extract (CLMSE), *Terminalia catappa* methanolic seed extract (TeCMSE), *Tricosanthes cucumerina* methanolic seed extract (TrCMSE), *Tetrapleura tetraptera* methanolic seed extract (TTMSE), defatted *Moringa oleifera* methanolic seed extract (DMOMSE); *Dennettia tripetala* methanolic fruit extract (DTMFE), *Artocarpus communis* methanolic fruit extract (ACMFE), *Gnetum africana* methanolic leaf extract (GAMLE), *Musa paradisiaca* methanolic stembark extract (MPMSE), and *Mangifera indica* methanolic stembark extract (MIMSE) for in vitro antioxidant and anti-acetylcholinesterase effects and associated polyphenolic and flavonoid contents.

*Dacryodes edulis* G. Don Lam (Burseraceae) is an edible pear native to the tropics. In the Niger Delta, the fruit is boiled or softened by exposure to heat and used to eat Zea mays (maize) or guinea corn. The pulp may also be boiled or roasted to form a kind of butter [19,20]. *D. edulis* leaf, fruit, and resin extracts have numerous pharmacological activities including antioxidant [21], anti-microbial [22], and anti-carcinogenic [23] properties.

*Cola lepidota* (Sterculiaceae) is popularly known as monkey cola. The plant is indigenous to tropical Africa and has its center of greatest diversity in West Africa [24]. The native peoples of southern Nigeria and the Cameron relish the fruits as a source of foodstuffs. Seeds of the monkey cola species are not edible, unlike the seeds of kola nut (*C. nitida*). *C. lepidota* is used in traditional medicine with functions that include its use as a stimulant, and to suppress sleep and for pulmonary problems and cancer-related ailments [25,26], with seed and fruit pulp extracts also displaying antioxidant activity [27]. Phytochemical analysis of the plant included detection of flavonoids [28].

*Trichosanthes cucumerina* Linn (Cucurbitaceae) is an annual, dioecious climber, widely distributed in Asian countries [29]. *T. cucumerina* fruit is consumed as a vegetable by rural dwellers, especially in the Western part of Africa. It is commonly called snake gourd, viper gourd, snake tomato, or long tomato [30]. The fruit is used as a cathartic, the seeds used for stomach disorders, anti-febrile and anti-helmintic activities and cardioprotective activities have also been reported [31].

*Dennettia tripetala* G. Baker (Annonaceae) is a fruit used as a spice and condiment in West Africa [32]. *D. tripetala* fruit is used in ethnomedicine to treat cold, fever, typhoid, cough, worm infestation, vomiting, stomach upset, and as an appetite enhancer [33]. Strong anti-nociceptive effects comparable to opioid agonists and non-steroidal anti-inflammatory drugs have been demonstrated [33]. Tannins, terpenoids, and other phytochemicals of *D. tripetala* are reported to be responsible for wide range of bioactivities [34].

*Artocarpus comminis* (Moraceae) is a flowering tree from the mulberry family. It is locally called breadfruit tree because of the "bread-like texture" of its edible fruits. *A. comminis* grown in the Niger Delta region is extensively used as both a food and traditional medicine. Studies have shown that *A. communis* possesses several bioactivities, such as antioxidant [35], anti-cancer [36,37], and anti-inflammatory activities [38,39]. Biologically active phytochemicals within *A. communis* include flavonoids, chalcones, and stilbenes [40].

*Terminalia catappa* Linn. (Combretaceae) is native to Southeast Asia. It is widely planted throughout the tropics and the Niger Delta region of Nigeria. *T. catappa* nut kernel can be eaten raw [41]. The ethnopharmacological properties of this plant have yet to be fully evaluated.

*Moringa oleifera* (Moringaceae), commonly known as horseradish tree or drumstick tree, is widely cultivated in Africa and other regions including South East Asia, and is considered a multi-purpose plant [42]. There are a broad number of bioactive agents present within the seeds, and for which seed extracts have been reported to exhibit neuroprotective effects [43–45].

*Tetrapleura tetraptera* (Mimosaceae), locally referred to as "Arindan" in Yoruba, is a flowering plant of the pea family native to West Africa. The dried fruit is used as a seasoning spice in the Southern part of Nigeria [46,47] and in the Niger Delta areas, the pod, fruit, and seeds are used as spices. A neuroprotective effect of *T. tetraptera* has been reported in scopolamine-induced amnesic rats [48].

*Gnetum africanum* (Gnetaceae) leaf, also known as wild spinach, is utilized as a food [49]. The vegetable is locally known under different nomenclature: sorgo (in Ogoni in Rivers state, Nigeria), afang (Ibibios in South-South, Nigeria), okazi (Igbo in South-East Nigeria), okok or eru (Cameroun), and fumbwa (Democratic Republic of Congo). Leaves are eaten as a vegetable raw or cooked and revered for their nutritional and therapeutic properties. The vegetable is domesticated for its economic potential, and useful component of dietary fibre, essential amino acids, vitamins, and minerals [50].

*Musa paradisiaca* (Musaceae), a banana plant, has a number of reported medicinal uses, for example banana seed mucilage has been applied as a treatment for catarrh and diarrhea [51]. There is also an extensive list of reported pharmacological activities including hepatoprotective effects [51].

*Mangifera indica* (Anacardiaceae) aqueous stem bark extract has been utilized as a remedy for diarrhea, fever, gastritis, and ulcers [52]. It has a number of biological activities, including the ability to act as an anti-cancer, and anti-bacterial agent [53].

Collectively, the plants described above were among the commonly edible and non-edible components of daily diets of Niger Deltans, hence an investigation of their relative antioxidant and anti-cholinesterase activities were undertaken.

#### **2. Materials and Methods**

#### *2.1. Collection and Identification of Plant Materials*

The seeds of *Dacroydes edulis, Cola lepidota, Tricosanthes cucumerina, Terminalia catappa, Tetrapleura tetraptera*, and defatted *Moringa oleifera* seed; fruits of *Dennettia tripetala, Artocarpus communis*, green leafy vegetable of *Gnetum africana*; stembark of *Musa paradisiaca* and *Mangifera indica* were collected from Niger Delta University Agricultural Extension farm, Amassoma, Yenegoa, Bayelsa State, Nigeria in March 2015 and authenticated by Mr. Philip Cheriose Nzien Alikwe, an Agriculturalist from the Department of Animal Science, Niger Delta University, Wilberforce Island, Bayelsa State, Nigeria. The voucher numbers (NDUH/P/71-81) were deposited in the University's herbarium, Niger Delta University, Nigeria. All of the samples were shade or air-dried for seven days and powdered using an electrical blender.

#### *2.2. Preparation of Hydromethanolic Extracts*

The seeds of *Dacroydes edulis*, *Cola lepidota, Tricosanthes cucumerina, Terminalia catappa, Tetrapleura tetraptera*, and defatted *Moringa oleifera* seed; fruits of *Dennettia tripetala*, *Artocarpus communis* and leafy vegetable of *Gnetum africanum*; stembark of *Musa paradisiaca* and *Mangifera indica* were powdered, and 10 g of each were macerated in 100 mL of 50% methanol for 72 h with vigorous hand agitation for one minute three times daily. Double layered gauze was used for filtration to obtain filtrates that were then reduced in volume at 40 ◦C on a water bath (Model TT6, Techmel and Techmel, Asaba, Nigeria) to obtain dried extracts. The extracts were weighed and the yield recovered (as a percentage) and recorded (Table 1). Extracts were stored in a refrigerator at 4 ◦C in an airtight container until used for experimental studies.


**Table 1.** Percentage yield, DPPH radical scavenging activity, AChE inhibitory potency, and total phenolic and flavonoid content for the evaluated food sources.

TeCMSE, *Terminalia catappa* methanolic seed extract; TrCMSE, *Tricosanthes cucumerina* methanolic seed extract; TTMSE, *Tetrapleura tetraptera* methanolic seed extract; DTMFE, *Dennettia tripetala* methanolic fruit extract; ACMFE, *Artocarpus communis* methanolic fruit extract; DEMSE, *Dacroydes edulis* methanolic seed extract; GAMLE, *Gnetum africanum* methanolic leaf extract; CLMSE, *Cola lepidota* methanolic seed extract; DMOMSE, Defatted *Moringa oleifera* methanolic seed extract; MIMSE, *Mangifera indica* methanolic stembark extract; MPMSE, *Musa Parasidisiaca* methanolic stem-bark extract. GAE: gallic acid equivalents; QUER E: quercetin equivalents. Extract was evaluated at least in triplicate across concentration range, and an approximate IC50 calculated.

#### *2.3. Chemicals*

Acetylthiocholine iodide (ATCI), L-ascorbic acid, bovine serum albumin (BSA), 2, 2-Diphenyl-1 picrylhydrazyl (DPPH), 5,5-dithiobis [2-nitrobenzoic acid] (DTNB), Folin-Ciocalteu reagent, physostigmine, and β-tocopherol were all purchased from Sigma Aldrich (Poole, UK), as were all the other chemicals used unless stated otherwise.

#### *2.4. Animals*

Rat brain homogenates from male F344 strain rats (200–230 g) were utilized as source of mammalian AChE, as described in an earlier report [54]. Rats were maintained at a controlled temperature of 21 ± 1 ◦C and a cycle of 16 h light/8 h dark with food intake daily and water ad libitum. Approval for the use of animals was obtained from the University of Nottingham Local Ethical Review Committee (study reference CHE 10, project licence approval code: PPL: 40/2624, approval date 13 June 2005 and the study was executed in line with the Animals Scientific Procedures Act (UK) 1986.

#### *2.5. 2,2-Diphenyl-1-Picrylhydrazyl (DPPH) Radical Scavenging Effect*

Spectrophotometric assays that utilized DPPH radical scavenging were used to quantify antioxidant activity. DPPH has been utilized extensively as a stable organic radical to evaluate scavenging activities of plethora of natural compounds such as flavonoids, polyphenols, and crude plant extracts and fractions. Antioxidants scavenge DPPH radicals by donating an electron to form reduced DPPH changing the colour of the solution from purple to yellow, the level of which can be quantified by spectrophotometry. DPPH radical scavenging assays were performed according to the method of Nwidu et al. [54]. Stock solutions of the plant extracts (5 mg/mL) were diluted to final concentrations of 200, 100, 50, 25, 12.5, and 6.25 μg/mL in ethanol. Then, 160 μL of 0.1 mM DPPH in ethanol solution was added to 20 μL solutions of the extracts as well as a standard and 20 μL H2O. For the standard, *β*-tocopherol was prepared at concentrations of 1.56, 0.78, 0.39, 0.195, and 0.0975 mg/mL. Assays were performed at 37 ◦C for 40 min in the dark, and thereafter the absorbance was read at 517 nm, as described in a previous report [54]. All reactions were performed in triplicates, from which an average was generated.

#### *2.6. Reducing Power Capacity Assay*

The reducing capacity (antioxidant ability) of the plant extract was also estimated based on its ability to reduce ferric ions (Fe3+) to ferrous ions (Fe2+). The plant extracts were assayed over the concentration ranged of 6.25–50 μg/mL. Four μL of 5 mg/mL of each plant extract was mixed with 400 μL of phosphate buffer (0.2 M dibasic sodium phosphate and 0.2 M monobasic sodium phosphate buffer adjusted to pH 7.4), 250 μL of 1% potassium ferricyanide was then added and the mixture incubated at 50 ◦C for 20 min. Then, 250 μL of 10% trichloroacetic acid was added, and after mixing, the solution was centrifuged at 3000 rpm for 10 min. One hundred μL from the supernatant was mixed with an equal volume of water, followed by 20 μL of freshly prepared ferric chloride solution. After mixing, the absorbance was measured at 700 nm in a microtiter plate reader, as previously reported [54]. Ascorbic acid was the reference substrate and the following concentration range was employed (0.3, 0.6, 0.9, 1.2, 1.5, and 3 mg/mL). All reactions were performed in triplicates, from which an average reading was generated.

#### *2.7. Acetylcholinesterase Inhibition Assay*

The assay for AChE inhibition was based upon the method of Ellman et al. [55], but modified for a 96-well microtiter plate format, as reported in Nwidu et al. [54]. In a microtiter plate, 40 μL of plant extract (at concentrations of 200, 20, 2, 0.2 and 0.02 μg/mL) was mixed with 35 μL of 50 mM Tris-HCl (pH 8.0) containing 0.1% BSA, 50 μL of 3 mM DTNB, and 50 μL of AChE. The AChE used was either from electric eel at 1 mg/mL (Sigma, Poole, UK) or that present within rat brain homogenate (prepared at 10% (*w*/*v*), according to the procedure of Carter et al. [56,57], which had been diluted 1:10 in 10 mM Tris-HCl pH 8.0 for assays. Plates were incubated at 37 ◦C for 5 min before the cholinesterase reaction initiated by the addition of 25 μL of 15 mM ATCI substrate, resulting in the production of 5-thio-2-nitrobenzoate anion that was read at 412 nm every 5 s for 10 min using a Spectramax microplate reader (Thermo Fisher, Stafford, UK). Eserine was employed at 0.02 μg/mL as a positive control for AChE inhibition. At this concentration (or above), eserine inhibits AChE to ≈100% [55]. All reactions were performed in triplicates, from which an average reading was generated.

#### *2.8. Determination of Total Phenolic Content*

A Folin-Ciocalteu Reagent (FCR) spectrophotometric method was used to quantify the total phenolic content in plant extracts, as described previously [54]. Twenty μL of each concentration of the plant extracts (ranging from 1–100 μg/mL) was added to 90 μL of water followed by addition of 30 μL of FCR, and samples were vigorously shaken within a microtiter plate reader. Within 30 s and a total assay time of eight minutes, 60 μL of 7.5% Na2CO3 solution was added to each microtiter well and then plates were incubated at 40 ◦C on a shaking incubator. The absorbance of the mixture was read after 40 min at 760 nm, as detailed in a previous report [56]. Gallic acid was used as the positive control substance. All reactions were performed in triplicates, from which an average reading was generated.

#### *2.9. Determination of Total Flavonoid Content*

Total flavonoid contents of the plant extracts were determined according to the method described by Nwidu et al. [54] using quercetin as a reference compound. Twenty microliter of plant extracts (5 mg/mL) were dissolved in ethanol and then mixed with 200 μL of 10% aluminum chloride solution and 1 M potassium acetate solution in microtiter plate wells. Samples were incubated for 30 min at room temperature, after which the absorbance of the solution was measured at 415 nm, as reported earlier [54]. Quercetin was used as the reference compound. All reactions were performed in triplicates, from which an average reading was generated.

#### *2.10. Statistical Analysis*

Results are expressed as the mean ± SD. IC50 values for each extract or fraction were calculated using non-linear regression analysis. A Spearman rank-order correlation coefficient was used to assess the relationship between total phenolic content, total flavonoid content, antioxidant content, and inhibition of AChE activity. Statistical analyses were performed using GraphPad Prism (Version 5.3) for Windows (GraphPad Software, Inc., San Diego, CA, USA, www.graphpad.com). A *p* value of <0.05 for results was considered to be statistically significant.

#### **3. Results**

#### *3.1. DPPH Radical Scavenging Activity*

Aqueous methanolic extracts of the inedible stem-bark, edible fruits, seeds and leaf extracts displayed DPPH radical scavenging activities in a concentrations-dependent manner as shown in Figure 1. From these analyses IC50 values for radical scavenging were calculated with results displayed in Table 1. When assessed at a concentration of 1000 μg/mL, the majority of aqueous methanolic extracts displayed significant (*p* < 0.05–0.001) DPPH radical scavenging effects, that ranged from ≈38–58% of that observed with Vitamin E (set at 100%). For the tested extracts, MIMSE (IC50 = 321 μg/mL) and MPMSE (IC50 = 106 μg/mL) demonstrated the highest percent inhibitions of 58% and 50%, respectively, and the latter extract was also the most potent (lowest IC50 value). Collectively, the descending order of DPPH radical scavenging activity was: MIMSE > MPMSE > TrCMSE > GAMLE > CLMSE > DMOMSE > DEMSE > DTMSE > ACMFE > TeCMSE > TTMSE with radical scavenging percentages of 58%, 50%, 42%, 40%, 40%, 38%, 37%, 21%, 20%, 19%, and 18%, respectively. The descending orders of potency of the extracts as radical scavengers as determined via IC50 values

(μg/mL) were: MPMSE > DTMFE > DEMSE > DMOMSE > TTMSE > TeCMSE > MIMSE > CLMSE > GAMLE > TrCMSE > ACMFE (Table 1).

**Figure 1.** DPPH radical scavenging activity of plant extracts. Plant antioxidant activity was measured via percentage inhibition of radical scavenging of DPPH. Results are expressed as means ± SEM for three separate experiments at each concentration. TeCMSE, *Terminalia catappa* methanolic seed extract; TTMSE, *Tetrapleura tetraptera* methanolic seed extract; TrCMSE, *Tricosanthes cucumerina* methanolic seed extract; DTMFE, *Dennettia tripetala* methanolic fruit extract; ACMFE, *Artocarpus communis* methanolic fruit extract; MPMSE, *Musa parasidisiaca* methanolic stem-bark extract; DMOMSE, Defatted *Moringa oleifera* methanolic seed extract; DEMSE, *Dacroydes edulis* methanolic seed extract; GAMLE, *Gnetum africanum* methanolic leaf extract; CLMSE, *Cola lepidota* methanolic seed extract; MIMSE, *Mangifera indica* methanolic stem-bark extract; Vit E, Vitamin E (β-Tocopherol). Results are expressed as means ± SEM for three separate experiments at each concentration. For marked significance from controls, a: *p* < 0.05, b: *p* < 0.01, c: *p* < 0.001.

#### *3.2. Reducing (Antioxidant) Capacity*

An evaluation of the reducing capacity of the aqueous methanolic extracts from the edible and non-edible foods showed that these also displayed antioxidant abilities in a concentration-dependent manner (Figure 2).

**Figure 2.** Reducing capacity of plant extracts. Plant reducing power was assessed via the ability to reduce ferric (Fe3+) to ferrous (Fe2+) iron. The percentage increase of reductive capacity with increasing plant extract concentration was determined. Vitamin C was used as a positive control. Results are expressed as means ± SEM for three separate experiments at each concentration. TeCMSE, *Terminalia catappa* methanolic seed extract; TTMSE, *Tetrapleura tetraptera* methanolic seed extract; TrCMSE, *Tricosanthes cucumerina* methanolic seed extract; DTMFE, *Dennettia tripetala* methanolic fruit extract; ACMFE, *Artocarpus communis* methanolic fruit extract; MPMSE, *Musa parasidisiaca* methanolic stem-bark extract; DMOMSE, Defatted *Moringa oleifera* methanolic seed extract; DEMSE, *Dacroydes edulis* methanolic seed extract; GAMLE, *Gnetum africanum* methanolic leaf extract; CLMSE, *Cola lepidota* methanolic seed extract; MIMSE, *Mangifera indica* methanolic stem-bark extract. Results are expressed as means ± SEM for three separate experiments at each concentration. For marked significance from controls, a: *p* < 0.05, b: *p* < 0.01, c: *p* < 0.001.

At a concentration of 50 μg/mL, all of the extracts demonstrated significant antioxidant effects except for TeCMSE and DTMFE when compared with Vitamin C (ascorbic acid). The highest antioxidant capacity was demonstrated by TCMSE2 (52%), MIMSE (40%) and then GAMLE (38%) relative to Vitamin C at 100%. The order of descending reducing capacity for the extracts was: TrCMSE > MIMSE > GAMLE > DEMSE > DMOMSE > ACMFE > CLMSE > MPMSE > DTMFE > TeCMSE.

#### *3.3. Acetylcholinesterase Inhibitory Activity*

Methanolic aqueous extracts of the inedible stem-bark, edible fruits, seeds, and leaf extracts of the evaluated plants displayed concentrations-dependent AChE inhibition, as shown in Figure 3.

**Figure 3.** AChE inhibitory activity of plant extracts. Plant inhibition of AChE was measured using a modified Ellman assay, with percentage inhibition of AChE calculated relative to eserine. Results are expressed as means ± SEM for three separate experiments at each concentration. MPMSE, *Musa parasidisiaca* methanolic stem-bark extract; DMOMSE, Defatted *Moringa oleifera* methanolic seed extract; DEMSE, *Dacroydes edulis* methanolic seed extract; GAMLE, *Gnetum africanum* methanolic leaf extract; CLMSE, *Cola lepidota* methanolic seed extract; MIMSE, *Mangifera indica* methanolic stem-bark extract; TeCMSE, *Terminalia catappa* methanolic seed extract; TTMSE, *Tetrapleura tetraptera* methanolic seed extract; TrCMSE, *Tricosanthes cucumerina* methanolic seed extract; DTMFE, *Dennettia tripetala* methanolic fruit extract; ACMFE, *Artocarpus communis* methanolic fruit extract. Results are expressed as means ± SEM for three separate experiments at each concentration. For marked significance from controls, a: *p* < 0.05, b: *p* < 0.01, c: *p* < 0.001.

Across the investigated concentrations the level of AChE inhibition was used to generate IC50 concentrations (Table 1).

At the higher concentrations assayed, all evaluated extracts exhibited significant (*p* < 0.001) concentration dependent AChE inhibitory activity, with percentage AChE inhibitions at 200 μg/mL ranging from ≈17–50%. The descending order of AChE inhibitory activity for the extracts was MISME > TrCMSE > DEMSE > PPMS >TTMS > CLMSE > DMOMSE > MPMSE > GAMLE > ACMFE > TeCMSE. The descending order of potency, as determined by IC50 values, were: MIMSE > TrCMSE > GAMLE > CLMSE > DEMSE > ACMFE > MPMSE > DTMSE > DMOMSE > TeCMSE > TTMSE.

#### *3.4. Total Phenolic Content and Total Flavonoid Content*

Total phenolic and total flavonoid contents were determined for each of the extracts and these have been included in Table 1. The relatively higher TPC levels (above 100 mg GAE/g) were observed for MISME, TrCMSE, GAMLE, CLMSE, and ACMFE at 156.2, 132.65, 123.26, 119.63, and 102.45 mg GAE/g, respectively. The relatively higher TFC levels (above 50 mg QUER/g) were recorded with MIMSE, GAMLE, ACMFE, CLMSE, TrCMSE, and DEMSE at 87.35, 73.26, 69.54, 68.35, 64.34, and 53.35 mg QUER/g, respectively.

#### *3.5. Correlation between AChE Inhibition, Antioxidant Ability, and Total Phenolic and Flavonoid Contents*

To consider if there was a relationship between AChE inhibition potency or DPPH radical scavenging potency and total phenolic or flavonoid content, Spearman rank correlations were calculated. The correlation coefficients (*R*-values) and significance of association (*p*-values) are shown in Table 2. The ability of extracts to inhibit AChE (measured as increasing IC50 values i.e., reduced potency) was significantly inversely correlated with increasing phenolic or increasing flavonoid content. Hence, extracts that displayed relatively high AChE inhibitory activity also retained relatively high phenolic or flavonoid content. By comparison, there was a positive but non-significant correlation between AChE inhibitory potency and either DPPH radical potency, or total phenolic or flavonoid content (Table 2).


**Table 2.** Correlation variables for AChE IC50, DPPH radical scavenging IC50 and total phenolic and flavonoid content of the evaluated food sources.

#### **4. Discussion**

Medicinal plants, spices, fruits, seeds, or vegetables provide an array of chemical entities with therapeutic potential. For example, medicinal plants may provide antioxidants in the form of flavonoids or polyphenols that are valuable assets for protection against oxidative stress and associated diseases. The public and scientific interest regarding the utilization of natural antioxidants continues to grow due to their potential or indeed perceived health-promoting effects.

Our analyses have shown *M. parasidisiaca* (106 μg/mL), *D. tripetala* (136 μg/mL), defatted *M. oleifera* (138 μg/mL), *T. tetraptera* (205 μg/mL), *T. catappa* (302 μg/mL) and *M. indica* (321 μg/mL) have highly active and significant DPPH (IC50) radical scavenging abilities. Other independent studies have also reported antioxidant properties of *M. parasidisiaca* [58–60], *D. tripetala* [61] *M. oleifera* [62,63], *T. tetraptera* [64,65], *T. catappa* [66], *M. indica* [67,68], and *T. cucumenina* [69,70]. Additionally, the protective effect of a natural extract from the stem-bark of *M. indica* was able to counter age-associated oxidative stress in elderly humans, indicative of its potential to act as a nutraceutical in vivo [68]. The antioxidant effects of the fruit of *Artrocarpus communis* [71] and the leaves of *Dacroydes edulis* [72] have also been reported. Interestingly, fruit (*Dacroydes edulis)* and vegetable (*Gnetum africanum*) intake

and an imbalance of oxidant/antioxidant status was reported to be associated with the development of diabetic retinopathy [72], with a recommendation that a diet rich in antioxidant supplements and tight glycemic control could postpone the onset of diabetic retinopathy [72].

The assessment of diets in a number of epidemiological studies and via quantitative evaluation have suggested that adherence to a Mediterranean-style diet and diets rich in fruits and vegetables may have protective benefits against age-related cognitive decline and neurodegenerative diseases [73–82]. This led us to consider the acetylcholinesterase inhibitory activity of these plants, since cholinesterase inhibitors are the mainstay of treatment for mild to moderate AD. An assessment of the potencies of a broad number of plant anticholinesterases inhibitors has been undertaken [83,84], with IC50 values ranging from 0.3 to 100.4 μg/mL. Hence, the plant extracts analyzed herein only displayed mild or moderate anti-AChE activities, with *M. indica* the most potent (IC50 of 111.9 μg/mL). Nevertheless, although only relatively weak cholinesterase inhibitors per se, chronic consumption of these foodstuffs might still provide provision of chemical entities able to ameliorate development or propagation of neurodegenerative disease. Indeed, an aqueous decoction of mango (*Mangifera indica* L.) stem bark has been developed on an industrial scale to be used as a nutritional supplement, cosmetic, and as a nutraceutical with neuroprotective effects [68,85,86]. Furthermore, neuroprotective effects of *Moringa oleifera* seed extract [52,53] and likewise *T. tetraptera* have also been demonstrated [48].

Our study also quantified the levels of phenolics and flavonoids, as these phytochemicals are widely distributed in the plant kingdom and possess antioxidant and anti-inflammatory activities [87]. Many of the active extracts investigated possessed high polyphenols and flavonoids content (Table 1) comparable to gallic acid and quercetin, respectively. Certain dietary phytochemicals, such as polyphenols have been reported to possess potential protection of cognitive function during aging [88] and may serve as natural neuroprotective agents [89–91]. In addition to their action as neuroprotective agents, flavonoids may also be efficacious candidates as potential pharmaceuticals or nutraceuticals for the treatment of AD [92]. Antioxidant activities of green tea phytochemicals and nutraceuticals such as curcumin, catechins, licopene, resveratrol, piperine, and anthocyanins, have been reported using in vitro and in vivo models [93,94].

Of interest, there was a significant inverse correlation between the potency of AChE inhibition (IC50 values) and total phenolic or flavonoid contents. This suggests that the agent(s) responsible for the AChE inhibitory activity are resident within the phenolic and flavonoid compounds. By contrast, there was no significant correlation between the AChE inhibition potency and that for DPPH radical scavenging, suggesting that the agent(s) that provide AChE inhibition is different from that for radical scavenging. Likewise, there was no correlation between AChE inhibitory potency and antioxidant activity, or between antioxidant activity and TPC or TFC, hence the chemical agent(s) that provided antioxidant protection were not AChE inhibitors, or likely to be abundant polyphenols or flavonoids.

A clear limitation of our study is that we have only assessed in vitro properties of these plant parts and their respective polyphenol and flavonoid content. We are unable to directly comment on how much of these foodstuffs are typically eaten, and indeed this will vary extensively between peoples and their food preparation methods. However, irrespective of these limitations, it is provocative to propose that a suitable diet rich in certain phytochemicals may provide beneficial counter-measures against oxidative stress-induced damage and its impact upon disease pathogenesis and propagation.

**Author Contributions:** L.L.N. was involved with conception and design of the study, performed experiments, collection and assemblage of the data, drafting of the article and final approval of the article, provided administrative, technical and logistic support. P.C.N.A. provided the plant materials used for this investigation. E.E. performed experiments, analyzed and interpreted the data, and provided statistical expertise. W.G.C. participated in critical editing and revision of the article for important intellectual content; provision of laboratory space and funding.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors are grateful to the University of Nottingham for the International Visiting Fellowship grants awarded to W.G.C. to support L.L.N. and E.E.

**Conflicts of Interest:** The authors confirm that there are no conflicts of interest connected with the publication of this manuscript. The funding sponsor had no role in the design of this study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results in this journal.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Antioxidant and Anti-Inflammatory Activities of Fractions from** *Bidens engleri* **O.E. Schulz (Asteraceae) and** *Boerhavia erecta* **L. (Nyctaginaceae)**

**Moussa COMPAORE 1,2,\*, Sahabi BAKASSO 3, Roland Nâg Tiero MEDA <sup>4</sup> and Odile Germaine NACOULMA <sup>1</sup>**


Received: 25 May 2018; Accepted: 7 June 2018; Published: 12 June 2018

**Abstract: Background:** According to recent studies, reactive oxygen is the leader of human metabolic disease development. The use of natural antioxidants is the best way to stop or prevent this problem. Therefore, the aim of this study was to evaluate the antioxidant and anti-inflammatory activities and to determine the polyphenolic contents of the *Bidens engleri* and *Boerhavia erecta* fractions. **Methods:** Plant fractions were obtained using Soxhlet procedures with hexane, dichloromethane, acetonitrile, ethyl acetate, methanol, and butanol solvent, successively. The different fractions were compared according to their antioxidant, anti-inflammatory activities, total phenolic, and total flavonoid contents. The phenolic contribution to the biological activity was evaluated. **Result:** The *Bidens engleri* and *Boerhavia erecta* fractions showed the highest antioxidant abilities, notably the polar fractions, which inhibited significantly the radical 2,2-diphenyl-1-picrylhydrazyl (DPPH) and 2,2-*O*-azinobis(3-ethylbenzoline-6-sulphonate) (ABTS). The butanol fraction from *Bidens engleri* and methanol fraction from *Boerhavia erecta* have presented the best iron (III) reduction power with 211.68 and 198.55 mgAAE/g, respectively. Butanol and acetonitrile were the best solvents for extracting phenolic compounds from *Bidens engleri* and *Boerhavia erecta*, respectively. In contrast, dichloromethane was the best solvent for extracting a flavonoid from two plants with anti-COX-2 and anti-LOX-15 active compounds. The phenolic compound contributed significantly to antioxidant activity (r > 0.80). **Conclusion:** The *Bidens engleri* and *Boerhavia erecta* fractions possessed a potential antioxidant for fighting oxidative stress and helping to prevent diabetes, hypertension, and cardiovascular diseases. The uses of this plant could be promoted in Burkina Faso.

**Keywords:** antioxidants; cyclooxygenase; lipoxygenase; phenolic; flavonoid; traditional medicine

#### **1. Introduction**

Oxidative stress is an inevitable consequence of life in an oxygen-rich atmosphere. The environment is filled with a lot of reactive oxygen species (ROS) and reactive nitrogen species (RNS). In recent data, it was demonstrated that ROS and RNS played an important role in human disease development [1]. In Burkina Faso, some recent studies have shown some alarming data concerning the prevalence of metabolic diseases [2]. The oxidative stress plays a direct or indirect role in the pathophysiology of diseases, such as cancer, diabetes, and cardiovascular diseases [3,4]. However,

the intake of natural antioxidants has been reported to reduce the risk of cancer, cardiovascular diseases, diabetes, and other diseases that are associated with aging [5,6].

Plants, fruits, vegetables, and medicinal herbs possess a wide variety of free radical scavenging biomolecules, such as phenolic compounds, flavonoids, vitamins, terpenoids, and some other endogenous phytometabolites, which are rich in antioxidant capacity [7–9]. *Bidens engleri* and *Boerhavia erecta* were some well-known medicinal plants in the Central Plateau, because they were used in the treatment of diabetes mellitus, hypertension, and old wounds [10]. According to Nacoulma's investigations, the traditional healers and herbalists used *Bidens engleri* and *Boerhavia erecta* in combination, for treating diabetes in Burkina Faso [10]. The same information was found in Cote d'Ivoire by ethnobotanical investigations [11].

The previous data demonstrated the antioxidant and anti-diabetic activities of *Boerhavia erecta* in India were associated with some polyphenolic compounds, such as phenolics and flavonoids [12,13]. Some antioxidant compounds, such as (+)-catechin (−)-epicatechin, quercetin, isorhamnetin, rutin, narcissin, isoquercitrin, and isorhamnetin 3-*O*-*β*-D-glucopyranoside, as well as other metabolites, were isolated from *B. erecta* leaves extract [14,15]. According to the medicinal importance of *B. engleri* and *B. erecta* in Burkina Faso, the present study aimed to highlight the potential of this plant by determining the antioxidant and anti-inflammatory activities, and the polyphenolic content of six organic fractions for identifying the type of metabolites that were responsible for biological activity.

#### **2. Materials and Methods**

#### *2.1. Plant Material*

Whole plants of *Bidens engleri* and *Boerhavia erecta* were taken from the Gampela region, which was situated in the mid-east of Kadiogo (central region), during the rainy season (August–September 2012). The sample was dried in the laboratory under ventilation. The sample was certified by Professor Jeanne MILLOGO, a botanist from the Laboratory of Plant Biology and Ecology (University of Ouagadougou). The herbaria were saved in the University Herbarium with numbers MC\_501 and MC\_502 for *Bidens engleri* and *Boerhavia erecta*, respectively.

#### *2.2. Reagents and Solvents*

The Folin–Ciocalteu reagent, sodium phosphate mono- and di-basics, sodium tetraborate, potassium persulfate, aluminum trichloride, trolox, 2,2-diphenyl-1-picrylhydrazyl (DPPH), 2,2-*O*-azinobis(3-ethylbenzoline-6-sulphonate) (ABTS), gallic acid, and trichloro acetic acid (TCA) were purchased from Sigma-Aldrich (Berlin, Germany). The sodium carbonate, potassium hexacyanoferrate, ascorbic acid, and ferric chloride were from Prolabo (Paris, France). The colorimetric COX (ovine) inhibitor screening assay kit, 15-lipoxygenase (soybean P1), linoleic, and arachidonic acids were purchased from Sigma-Aldrich, (New York, NY, USA).

#### *2.3. Extraction Procedures*

Of the sample, 20 g were successively extracted using hexane, dichloromethane, acetonitrile, ethyl acetate, methanol, and butanol in a Soxhlet system. The solvent was removed in a rotary evaporator system.

#### *2.4. Antioxidant Effects Evaluation*

#### 2.4.1. Radical DPPH Inhibition Determination

The fractions' capacities to inhibit radical DPPH were evaluated according to the method that was presented by Compaoré et al. [16]. In a 96 micro-well plate, 200 μL of DPPH (20 mg/L) and 100 μL of fraction were incubated in the dark for 10 min, and the absorbencies were read at 517 nm using a spectrophotometer (BioTek Instruments, New York, NY, USA). Quercetin was used to generate a standard curve (*<sup>y</sup>* <sup>=</sup> −27.94 + 8.15, *<sup>r</sup>*<sup>2</sup> = 0.99, *<sup>p</sup>* < 0.0001). The results were expressed in milligram Quercetin equivalent per gram (mgQE/g).

#### 2.4.2. Trolox Equivalent Antioxidant Capacity Assay

The method that was described by Compaoré et al. was used to evaluate the sample scavenging ABTS ability [16]. To 200 μL of diluted ABTS solution, 50 μL of fraction or trolox was added, with incubation in the dark for 5 min. The absorbance was read at 734 nm, with a microplate reader (BioTek Instruments, New York, NY, USA). Trolox was used to generate the standard curve (*y* = −72.38*x* + 54.57, *r*<sup>2</sup> = 0.99, *p* < 0.001) and the results were expressed in millimole Trolox equivalent per gram (mMTE/g).

#### 2.4.3. Ferric (Fe III) Reducing Antioxidant Power (FRAP) Assay

The reducing power of the extracts was determined according to the method that was presented by Compaoré et al. [16]. The data were transformed to mg of ascorbic acid per gram of fraction (mgAAE/g), because the standard curve was obtained with ascorbic acid (*y* = 105.9*x*, *r*<sup>2</sup> = 0.99, *p* < 0.0001). The iron (III) reducing activity of each sample was obtained from two of the three independent determinations.

#### *2.5. Anti-Inflammatory Tests*

#### 2.5.1. COX-1 and COX-2 Inhibition Assay

The inhibition of COXs was performed using a commercially available colorimetric COX (ovine) inhibitor screening assay kit (Cayman Chemical Company, New York, NY, USA). All of the inhibitors were dissolved in an appropriate solvent. The COX activity was evaluated using *N*,*N*,*N*'*N*'-tetramethyl-*p*-phenylenediamine (TMPD) as a co-substrate, with arachidonic acid. The TMPD oxidation was monitored spectrophotometrically at 590 nm (BioTek Instruments, New York, NY, USA). The inhibition percentage that was induced by 100 μg/mL of the sample was calculated.

#### 2.5.2. Lipoxygenase 15 Inhibition Assay

The assay was performed according to the previous procedure that was presented by Compaoré et al. [17]. The incubation mixture consisted of the sample solution (100 μg/mL) in an appropriate solvent and 200 μL of the enzyme solution (167 U/mL) in a boric acid buffer (0.2 M, pH 9). After the incubation at room temperature for 5 min, the reaction was started by adding 250 μL of linoleic acid solution (250 mM in buffer). The conversion of linoleic acid to 13-hydroperoxylinoleic acid was recorded by measuring the samples' absorbencies at 234 nm, during 3 min, and against the appropriate blank solutions, without extracts. The inhibition percentage was calculated.

#### *2.6. Polyphenolic Amount Quantification*

#### 2.6.1. Phenolic Content Determination

The total phenolic content was evaluated using a Folin–Ciocalteu colometric assay, as described by Compaoré et al. [16]. The sample was mixed with Folin-Ciocalteu Reagent (0.2N). After incubation in the dark, 100 μL of sodium carbonate was added. The absorbance (760 nm) was measured after a second incubation in the dark (2 h), using the Biotek equipment (BioTek Instruments, New York, NY, USA). Gallic acid was used to produce the standard curve (*<sup>y</sup>* = 201*<sup>x</sup>* − 21.22, *<sup>r</sup>*<sup>2</sup> = 0.99, *<sup>p</sup>* < 0.0001) and the results were expressed in mg gallic acid, equivalent per gram (mgGAE/g) of extract.

#### 2.6.2. Total Flavonoid Content Evaluation

The total flavonoid content was determined according to the previous method that was described by Compaoré et al. [16]. Then, 100 μm of sample and 100 μL of AlCl3 (2%) were mixed in 96 micro-wells and were incubated for 10 min. The absorbance was measured at 415 nm with a microplate reader (BioTek Instruments, New York, NY, USA). Quercetin was used to generate the standard curve (*<sup>y</sup>* = 39.8*<sup>x</sup>* − 3.5, r2 = 0.99, *<sup>p</sup>* < 0.0001) and the results were expressed at mg quercetin equivalent per gram (mgQE/g) of sample.

#### *2.7. Statistical Analyses*

Microsoft Excel was used to calculate the average and standard deviation of the repeated tests (*n* = 2 × 3). GraphPad Prism 6.01 (San Diego, CA, USA, 2012) and Xlstat Pro 7.5 (Paris, France, 2005) were used to produce the standard curve and to measure the statistical significant results, respectively (*p* < 5%).

#### **3. Results and Discussion**

#### *3.1. Antioxidant Activities*

The use of medicinal plants in Burkina Faso has been a current activity of the population [18]. However, the main role of the researchers was to promote the medicinal uses. The plant extracts' antioxidant potential is shown in Table 1. The radical DPPH scavenging effect was decreased from 63.94 mgQE/g to 2.25 mgQE/g, and the radical ABTS scavenging power was decreased from 22.86 mMTE/g to 7.16 mMTE/g. The hexane fractions presented radical ABTS scavenging activities contrary to the anti-DPPH radical effect. The butanol fraction from *Bidens engleri* demonstrated the best antiradical possibility, similar to the acetonitrile from *Boerhavia erecta*. The ability of the fractions to reduce iron (III) were increased from 10.20 mgAAE/g to 211.68 mgAAE/g. In general, the *B. erecta* sample presented some antioxidant activity that was superior to the *B. engleri* samples. This data demonstrated the importance of these plant samples in stress oxidative management. In the previous data, it was demonstrated that *B. erecta* possessed some antioxidant activity that was supported by the flavonoid compounds [14,19]. The anti-DPPH, anti-ABTS, and iron (III) reduction abilities were evaluated [13,20]. However, it was the first antioxidant activity data from *Bidens engleri*, according to our bibliographic survey. According to previous antioxidant activities of similar fractions from *Commifora africana* (A. Rich.) Engl. (Burseraceae) and *Loeseneriella africana* (Willd.) (Celastraceae), which were from the same region, the present plants possessed a lowest antioxidant power [16].


**Table 1.** Yield and antioxidant activity of fractions.

Data in each column were statistically different letter (a–j) (*p* < 0.05) except data with same letters. The data were obtained in two independent triplate tests *(n* = 2 × 3). Aterisk (\*) indicated data that were obtained by one procedure extraction. DPPH: 2,2-diphenyl-1-picrylhydrazyl, ABTS—2,2-*O*-azinobis(3-ethylbenzoline-6-sulphonate); FRAP—ferric (Fe III) reducing antioxidant power. mgQE/g: milligram quercetin equivalent per gram, mMTE/g: millimole Trolox equivalent per gram, mgAAE/g: milligram ascorbic acid equivalent per gram.

#### *3.2. Anti-Inflammatory Activity*

Table 2 presents the data concerning the inhibition of prostaglandin production from COXs and LOX-15 regular activities. COX-2 was more sensitive than COX-1 and LOX-15, which were not sensitive to the *B. erecta* fractions. The percentage inhibition of COX-2 was from 23.65% to 64.72% at 100 μg/mL, as the final concentration of the fraction. The LOX-15 inhibition percentage was increased from 36.76 to 64.90%. The maximal inhibition of COX-1 was obtained with ethyl acetate (42.51%) from *B. engleri*. Interestingly, the dichloromethane fractions from *B. engleri* and *B. erecta* were the active fractions for COX-2 (64.72 ± 2.13%) and LOX-15 (62.55 ± 5.09%), respectively, according to the enzyme activity classification scale [21]. According to the previous data, *B. erecta* possessed some anti-inflammatory activity in vivo [22], but in the present study, the *B. erecta* fractions could not significantly inhibit the prostaglandin production from the COXs and LOX-15 activity. It was suggested that the enzyme inhibition was not the method of action of this anti-inflammatory effect. This was the first study of the evaluation of the COXs and LOX-15 inhibition power of two plants. These enzyme inhibition activities of the *B. engleri* fractions were very little compared with the *Commifora africana* and *Loeseneriella africana* fractions inhibition effect [16]. In contrast, the butanol and dichloromethane fractions from *B. engleri* presented some interesting inhibition activity of LOX-15, compared with the *Bauhinia rufescens* extract, Lam. (Caesalpiniaceae) [17].



Data in each column were statistically different letter (*p* < 0.05) except data with same letters (a–j). The data were obtained in triplate tests *(n* = 3).

#### *3.3. Total Phenolic and Total Flavonoid Contents*

As the metabolites were the main contributor to the antioxidant and anti-inflammatory powers, the phenolic and flavonoid contents were evaluated [16,23]. The yield of extraction is shown in Table 1. The methanol was the best solvent for extracting some metabolites from two plants, with a yield that was superior to 100 mg/g. Figure 1 shows the amount of flavonoid and phenolic in all of the fractions from *B. erecta* and *B. engleri*. The phenolic content was decreased from 425.12 to 5.92 mgGAE/g, and the flavonoid amount was increased from 2.62 to 30.38 mgQE/g. A notable variable distribution of polyphenolic compounds was found in concordance with the solvent polarities. Notably, *B. engleri* contained some non-polar flavonoids in the major compound that were extracted in dichloromethane, in contrast to *B. erecta*, which presented some polar compounds that were extractible by acetonitrile, ethyl acetate, and methanol. In previous phytochemical investigations, the flavonoid and phenolic contents were evaluated in the extracts from *B. erecta*. It was found that the ethanol and phosphate buffer were able to extract the flavonoid and phenolic compounds [13,14]. The flavonoid and phenolic individual compounds, with a radical scavenging activity and iron (III) reduction ability, were previously detected in the *B. erecta* extracts [24–26]. It was quercetin and isorhamnetin

and their glycosides, rutin, narcissin, isoquercitrin, and isorhamnetin 3-*O*-*β*-D-glucopyranoside, as well as the two flavan-3-ols, [(+)-catechin] and [(−)-epicatechin], that are well known antioxidant phytometabolites [24–26]. These compounds showed anti-COX and anti-LOX properties [27,28].

The correlation analysis showed that phenolic contributed significantly to the radical scavenging and iron (III) reduction. The contribution to the anti-DPPH, anti-ABTS, and iron (III) reduction were 0.91, 0.86, and 0.99, respectively (*p* < 0.0001). Similar findings were shown in a previous study [16,29,30]. Additionally, it was found in this study that there was an insignificant correlation between the COX-2 and phenolic compound, contrary to a previous study [16].

**Figure 1.** Polyphenolic content of fractions. A—total flavonoid contents; B—total phenolic contents. The data were obtained in two independent triplate tests (*n* = 2 × 3). The data in each histogram were statistically different (*p* < 0.05), except for the data with the same letters (a–h). BuOHF: butanol Fraction, MeOHF: methanol fraction, EAF: ethyl acetate fraction, ACNF: acetonitrile fraction, DCMF: dichloromethane fraction, HF: hexane fraction, mgQE/g: milligram quercetin equivalent per gram, mgGAE milligram gallic acid equivalent per gram.

#### **4. Conclusions**

This study highlighted the antioxidant, anti-inflammatory, and the phytochemical potential of six fractions from *B. engleri* and *B. erecta*, well-known medicinal plants of Burkina Faso. Their utilization could be supported partially by antiradical scavenging and iron (III) reduction. According to the interesting biological activity of *B. engleri*, the next step would be to isolate the anti-radical flavonoid from butanol, ethyl acetate, and acetonitrile fractions, as well as the anti-COX-2 and anti-LOX-15 compounds from the dichloromethane in vitro model.

**Author Contributions:** M.C. corresponding author, sampling, conception of protocol, redaction of paper, and data analysis; R.N.T.M. conception of protocol, redaction, and correction; S.B. redaction and correction of the article; and O.G.N. laboratory headmaster, validation of plant list, and correction.

**Funding:** This research was funded by the International Foundation for Sciences grant number AF/20286.

**Acknowledgments:** Supported by International Foundation for Sciences (grant No. AF/20286). We also thanked Professor Jeanne MILLOGO for the botanical authentication of two plants.

**Conflicts of Interest:** The authors declare no conflicts of interest.

#### **References**


© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Effect of Hochuekkito (Buzhongyiqitang) on Nasal Cavity Colonization of Methicillin-Resistant** *Staphylococcus aureus* **in Murine Model**

#### **Masaaki Minami 1,\*, Toru Konishi <sup>2</sup> and Toshiaki Makino <sup>2</sup>**


Received: 29 June 2018; Accepted: 25 July 2018; Published: 1 August 2018

**Abstract: Background:** Methicillin-resistant *Staphylococcus aureus* (MRSA) infections are largely preceded by colonization with MRSA. Hochuekkito is the formula composing 10 herbal medicines in traditional Kampo medicine to treat infirmity and to stimulate immune functions. We evaluated the efficacy of hochuekkito extract (HET) against MRSA colonization using a nasal infection murine model. **Methods:** We evaluated the effects of HET as follows: (1) the growth inhibition by measuring turbidity of bacterial culture in vitro, (2) the nasal colonization of MRSA by measuring bacterial counts, and (3) the splenocyte proliferation in mice orally treated with HET by the 3H-thymidine uptake assay. **Results:** HET significant inhibited the growth of MRSA. The colony forming unit (CFU) in the nasal fluid of HET-treated mice was significantly lower than that of HET-untreated mice. When each single crude drug—Astragali radix, Bupleuri radix, Zingiberis rhizoma, and Cimicifugae rhizome—was removed from hochuekkito formula, the effect of the formula significantly weakened. The uptake of 3H-thymidine into murine splenocytes treated with HET was significantly higher than that from untreated mice. The effects of the modified formula described above were also significantly weaker than those of the original formula. **Conclusions:** Hochuekkito is effective for the treatment of MRSA nasal colonization in the murine model. We suggest HET as the therapeutic candidate for effective therapy on nasal cavity colonization of MRSA in humans.

**Keywords:** MRSA; Hochuekkito; Japanese traditional Kampo medicine; murine colonization model

#### **1. Introduction**

*Staphylococcus aureus* infection, such as surgical site infection, is a common hospital-associated infectious disease. It causes the extension of hospital stays and increases the costs of health-care [1]. The increasing rates of clinical isolates of *S. aureus* worldwide are methicillin-resistant [2]. The attributable mortality of *S. aureus* septicemia infection is about 20% for methicillin-sensitive strains and about 30% for methicillin-resistant *S. aureus* (MRSA) [3]. The development of new effective medication is desired for the improvement of morbidity and mortality regarding *S. aureus* infection.

Nasal colonization is an important risk factor for *S. aureus* infection. It is associated with up to 13-fold increased risk of *S. aureus* infection [4]. A study of nosocomial *S. aureus* bacteremia demonstrated nasal colonization on admission in most cases [5]. Nasal colonization is the predecessor to infection because the infecting strain was identical to the isolated colonizing strain before infection in four-fifths of *S. aureus* septicemia cases [6]. Decolonization therapy reduces the risk of healthcare-associated *S. aureus* infection in high-risk settings such as surgery, supporting the hypothesis that colonization leads to infection [7].

Traditional Chinese medicine (TCM) is one of the most popular alternative, complementary therapies worldwide [8]. In Japan, Kampo medicine, which is the traditional medicine developed from ancient Chinese medicine, is recognized as an effective alternative medicine against several diseases [9,10]. Hochuekkito (Buzhongyiqitang) is a formula in both traditional Japanese Kampo medicine and Chinese medicine. This formula comprises 10 crude drugs shown in Table 1. Hochuekkito extract (HET) has been used to treat severe infirmity such as weakness and loss of appetite of the elderly [11]. As HET is a popular alternative medicine in Japan, limited scientific evidence is available on the use of HET for the treatment of MRSA colonization [12,13]. Thus, the clarification of the precise mechanism of Hochuekkito efficacy against MRSA colonization has been desired.

In the present study, we evaluated the efficacy of HET against MRSA colonization using a nasal infection murine model. Furthermore, we also evaluated the efficacy of the constitutive crude drug of HET against MRSA and immunological activity of murine splenocytes from HET-treated mice.


**Table 1.** Composition of hochuekkito.

Daily doses of crude drugs in hochuekkito are in Japanese Pharmacopoeia 17th Edition [14].

#### **2. Materials and Methods**

#### *2.1. Bacterial Strains and Culture Condition*

MRSA (ATCC\_BAA-1556 (FPR3757)) (American Type Culture Collection, Rockville, MD, USA) was used in this study. After overnight pre-incubation on TSAII sheep blood agar (Nihon Becton Dickinson, Tokyo, Japan), a fresh colony of bacteria was cultured for 16 h at 37 ◦C. The bacteria were harvested by centrifugation and re-suspended in sterile Luria–Bertani (LB) medium (Becton Dickinson, Franklin Lakes, NJ, USA). Bacterial density was determined by measuring the absorbance at 600 nm (A600). The bacterial suspension was then diluted with LB to 106 CFU (colony forming unit)/mL using a standard growth curve to relate measured A600 to bacterial concentration. The bacteria were cultured at 37 ◦C and A600 was measured at every 2 h.

#### *2.2. Crude Drugs and Exteact Preparation*

Astragali radix (lot number, 6C30M), 4.0 g of Ginseng radix (5D25), 4.0 g of Atractylodes rhizome (3J07M), 3.0 g of Angelicae radix (5G06M), 2.0 g of Zizyphi fructus (5G07M), 2.0 g of Aurantii nobilis pericarpium (6B16M), 2.0 g of Bupleuri radix (6C15M), 1.5 g of Glycyrrhizae radix (6B22), 1.0 g of Cimicifugae rhizome (0F28M), and 0.5 g of Zingiberis rhizome (5G07M). These cut crude drugs were purchased from Daiko Shoyaku (Nagoya, Japan) and standardized by Japanese Pharmacopoeia 17th Edition [14]. Voucher specimens of each single crude drug were deposited in the Department of Pharmacognosy, Graduate School of Pharmaceutical Sciences, Nagoya City University. The mixture of the above crude drugs was boiled in 20-times weight of water for 30 min, and filtered. The decoction was lyophilized to yield powdered extract (HET, the ratio of the extract yielded was 36%). A fingerprint pattern of this HET was created as follows. HET (50 mg) was suspended with MeOH (1 mL) and sonicated for 30 min. The supernatant (30 μL) was injected to HPLC with the following conditions: system, Shimadzu LC-10A*VP* (Kyoto, Japan); column, TSK-GEL ODS-80TS (4.6 × 250 mm, Tosoh, Tokyo, Japan); mobile phase, 0.05 M AcOH–AcONH4 buffer (pH 3.6)/CH3CN 90:10 (0 min)–45:55 (40 min), linear gradient; flow rate, 1.0 mL/min; column temperature, 40 ◦C; and detection, 200–400 nm by a photodiode array detector. Some peaks were identified by the retention times and UV spectra of the standard compounds. The fingerprint chromatogram of HET extract is shown in Figure 1. HET was suspended in distilled water to prepare the stock solution at a concentration of 0.1 g/mL, and kept at −20 ◦C until use. From the 10 crude drugs of hochuekkito formula, each single crude drug was removed to make 10 kinds of modified hochuekkito formula containing 9 crude drugs. The extracts of the modified formula were prepared in the same way.

**Figure 1.** HPLC fingerprint of hochuekkito extract (HET). Compounds were identified by comparison of the retention times of the UV spectra with those of standard compounds.

#### *2.3. Murine Model of Bacterial Nasal Infection*

This study was approved by the Animal Experiment Committee of Graduated School of Medical Sciences of Nagoya City University in accordance with the guidelines of the Japanese Council on Animal Care. Ethical approval code: H28M-05. Date of approval: 2 March 2016. Mice were purchased from Japan SLC (Hamamatsu, Japan). The ability of the colonized effect of MRSA in mice after nasal inoculation was assessed using a previous procedure [15]. In brief, bacteria were harvested after 16 h of growth on TSAII sheep blood agar, and were mixed in 1 mL of phosphate buffered saline (PBS, pH 7.2, 0.15 M), then centrifuged at 2000× *<sup>g</sup>* for 2 min. The pellets were diluted in 100 <sup>μ</sup>L PBS to <sup>1</sup> × <sup>10</sup><sup>7</sup> CFU, and then inoculated into both nostrils of inbred six-week-old female Balb/c mice using a micropipette. The number of CFU inoculated was verified for each experiment by plating the bacteria on TSAII sheep blood agar and counting CFU. Mice were observed daily. In the HET-treated group, mice were administered with HET (0.85, 1.7, or 3.4 g/kg body weight/day body weight, which were equivalent to 5, 10, and 20 times the dosage of humans, respectively) on days −1, 0, 1, 2, and 3 after the bacterial inoculation (Figure 2). The mice in the control group were given PBS without infection.

**Figure 2.** Protocols for murine experiments of methicillin-resistant *Staphylococcus aureus* (MRSA) colonized model. In the infected group, 1 <sup>×</sup> <sup>10</sup><sup>7</sup> colony forming unit (CFU) bacteria were injected into both nostrils of mice using a 29 gauge needle at day 0. In the hochuekkito extract (HET)-treated group, mice were administrated with HET *p.o.*

#### *2.4. Nasal Lavage Cultures*

The procedure of nasal cultures was described elsewhere [16]. In brief, the mice were sacrificed by CO2 inhalation. After that, the external noses, oral cavity, and head were disinfected with a moist alcohol swab and allowed to dry. Nasal lavage was performed with 200 μL of PBS. The recovered fluid was then serially diluted, and 10 μL of each dilution was plated onto TSAII sheep blood agar plates. The plates were incubated for 24 h, and then colonies of bacteria were counted. The results were quantified as the number of CFU/mL.

#### *2.5. Determination of Splenocyte Proliferative Response*

The oral administration protocol for this assay was done in almost the same manner for bacterial nasal infection, except for no-infection with MRSA. After the mice were sacrificed by CO2 inhalation, the spleen was removed aseptically, and splenocytes were filtered and cultured in RPMI 1640 (Wako Pure Chemical Industry, Osaka, Japan) containing 5% fetal calf serum (FBS, Sigma-Aldrich, St. Louis, MO, USA). At 20 h prior to the culmination of the splenocyte culture, 3H-thymidine (2.0 Ci/mmol; PerkinElmer, Waltham, MA, USA) was added into the medium, and the cells were further incubated for 4 h. Then, the cells were adsorbed on 0.45-μm membrane filters, washed with distilled water, and then dried. The filters were transferred to vials filled with liquid scintillator cocktail (Ultima Gold, Perkin Elmer, Inc., Waltham, MA, USA), and the radioactivity was measured by using a liquid scintillation counter (LSC-6100, Hitachi Aloka Medical, Tokyo, Japan). The results are given as disintegrations per minute (DPM).

#### *2.6. Statistical Analysis*

All statistical analyses were conducted using Tukey/Bonferroni's multiple comparison test for differences among multiple groups (EZR version 1.36, http://www.jichi.ac.jp/saitama-sct/SaitamaHP. files/statmedEN.html). Values less than 0.01 indicated statistical significance.

#### **3. Results**

#### *3.1. Bacterial Growth Inhibitory Effect*

First of all, we tried to evaluate whether or not HET could inhibit the growth of MRSA. MRSA was grown in LB medium with or without HET, and the inhibitory ability of bacterial growth was assessed. As expected, HET (10 mg/mL) significantly inhibited the growth of MRSA (*p* < 0.01). We confirmed that this inhibitory ability was in dose- and time-dependent manners (Figure 3).

**Figure 3.** Bacterial growth inhibitory effect of hochuekkito extract (HET). MRSA was cultured on Luria–Bertani (LB) medium with or without HET for 10 h. The bacterial growth was evaluated by measuring absorbance at 600 nm. Open circle, closed circle, open square, and closed square exhibited HET 0, 0.1, 1, and 10 mg/mL, respectively. Data shown represent the mean ± S.D. (*n* = 6). \*\* *p* < 0.01 by Tukey/Bonferroni's multiple comparison test.

#### *3.2. Murine Nasal Infection Model*

Next, we tried to assess whether HET would provide in vivo effects against MRSA. Four days after nostril infection of MRSA, we evaluated the bacterial colony counts in murine nose. The CFUs of HET-treated murine nasal lavage were lower than those of HET-untreated mice in dose-dependent manners, and the group treated with 3.4 g/kg/day exhibited statistical significance (*p* < 0.01) (Figure 4). In order to find the active components in the hochuekkito formula, we prepared the extracts of the modified formulas, which contain nine crude drugs. The extracts of modified hochuekkito formulas—that is, Astragali radix, Bupleuri radix, Zingiberis rhizoma, or Cimicifugae rhizome—exhibited significantly lower activities than HET, respectively (*p* < 0.01) (Figure 5).

**Figure 4.** The colonies of MRSA in hochuekkito extract (HET)-treated and untreated murine nasal lavage. The nasal fluids were inoculated on TSAII sheep blood agar and incubated for 24 h. Comparisons of colony count between HET-treated and untreated mice were performed. Data represent the mean ± S.D. (*n* = 6). \*\* *p* < 0.01 by Tukey/Bonferroni's multiple comparison test.

**Figure 5.** The colonies of MRSA of nasal lavage collected from mice treated with the extracts of modified hochuekkito formulas. The nasal fluids were inoculated on TSAII sheep blood agar and incubated for 24 h. a: untreated, b: hochuekkito extract (HET)-treated, c: aurantii nobilis pericarpium-removed HET, d: zizyphi fructus-removed HET, e: angelicae radix-removed HET, f: zingiberis rhizome-removed HET, g: Atractylodes rhizome-removed HET, h: Ginseng radix-removed HET, i: astragali radix-removed HET, j: bupleuri radix-removed HET, k: cimicifugae rhizome-removed HET, l: glycyrrhizae radix-removed HET, respectively. Dosage of HET was 3.4 g/kg/day, and those of the extracts of other modified hochueekito formulas were equivalent to this dosage. Data represent the mean ± S.D. (*n* = 6). \*\* *p* < 0.01 by Tukey/Bonferroni's multiple comparison test.

#### *3.3. Splenocyte Proliferative Activity in HET-Treated Mice*

We also studied the activity of splenocyte in mice treated with HET, because splenocytes play major roles in murine bacterial infection models. To determine whether or not the activity of the splenocytes collected from HET-treated mice was elevated, we performed 3H-thymidine uptake analysis. As shown in Figure 6, the uptake of 3H-thymidine into splenocytes collected from mice orally treated with HET was significantly (*p* < 0.01) higher than that from untreated mice in dose-dependent manners. The extracts of modified hochuekkito formulas—that is, Astragali radix, Bupleuri radix, Zingiberis rhizoma, or Cimicifugae rhizome—exhibited significantly lower 3H-thymidine uptake compared with HET (*p* < 0.01) (Figure 7).

**Figure 6.** 3H-thymidine-uptake assay in hochuekkito extract (HET)-treated and untreated murine splenocyte. Six-week-old female Balb/c mice were administrated with HET for four days, and the splenocyte were collected. Data represent the mean ± S.D. (*n* = 6). \*\* *p* < 0.01 by Tukey's/Bonferroni multiple comparison test. DPM—disintegrations per minute.

**Figure 7.** 3H-thymidine-uptake assay of the splenocytes collected from mice treated with the extracts of modified hochuekkito formulas. Six-week-old female Balb/c mice were administrated with hochuekkito extract (HET) for four days, and splenocytes were collected. Symbols of a–l and the dosages of the samples were as same as those shown in Figure 5. Data represent the mean ± S.D. (*n* = 6). \*\* *p* < 0.01 by Tukey/Bonferroni's multiple comparison test.

#### **4. Discussion**

In this study, we tried to clarify that HET would be effective for the eradication in the MRSA-colonized murine model. Our results showed that the turbidity of the bacteria increases over time at the start and that the turbidity decreases compared with the control when HET is added, so this is the growth suppressing effect. After MRSA nasal infection, HET-treated mice showed a reduction of MRSA colonization in murine nose and the upregulation of murine splenocyte activity. Furthermore, we demonstrated that four crude drug components of HET—Astragali radix, Bupleuri radix, Zingiberis rhizoma, and Cimicifugae rhizome—affected the eradication of MRSA. Our results suggest that HET can play a crucial part in protection against MRSA colonization in the mouse model.

Several studies of HET on microbial infections have been investigated. In a small-scale clinical trial about MRSA infection, eradication of MRSA was successful when HET was administered to five MRSA carriers' patients [12]. Another study showed that when HET was administered to 34 asymptomatic patients from which MRSA was isolated from urine, MRSA was not isolated from urine in 12 patients, and 10 patients decreased the bacterial volume to less than 1/100 [13]. Other human clinical trials in lung *Mycobacterium avium* complex patients with HET for six months resulted in weight gain and increased serum albumin value without a tendency for infectious disease to exacerbate on chest radiograph [17]. Even in healthy elderly humans, natural killer (NK) cell activity increased at 30 days and 120 days after administration of HET. The serum interferone (IFN)-γ activity also increased [18]. A clinical large-scale trial to confirm the effect of HET on MRSA carriage of human nasal cavity is desired from the investigation of the effect of HET on these bacterial colonizations and chronic infections for humans and our experimental results.

Non-human experimental studies also revealed the efficacy of HET against bacterial infection. When *Listeria monocytogenes* was infected intraperitoneally in mice, HET showed an increase in polynuclear leukocytes and macrophages in the spleen. HET also confirmed renewal of phagocytic capacity of *L. monocytogenes* in intraperitoneal macrophages [19]. In a mouse infected with *L. monocytogenes*, HET showed a decrease in bacterial quantities in Peyer's patches, lymph nodes, and liver. HET showed increased phagocytosis of liver macrophages against bacteria. It also showed an increase in IFN-γ producing cells in intraepithelial lymphocytes [20]. When HET was administrated in *L. monocytogenes* infected infant mice, the amount of *L. monocytogenes* in the liver and spleen decreased. Activation of IFN-γ producing CD4 T cells enhanced IFN-γ activity. The ability of macrophages to present antigen by MHC class II expression is enhanced [21]. In vitro experiments inhibited the growth

of *Helicobacter pylori* at a concentration of HET 2.5 mg/mL. In addition, the amounts of bacteria in the stomach were decreased in mice by oral administration of HET in an in vivo experiment. Furthermore, the expression of IFN-γ in the gastric mucosa was elevated [22]. As our bacterial growth study showed that HET suppressed the MRSA in a dose-dependent manner, HET may have a bacterial inhibitory effect regardless of bacterial species. By infecting mice with *Brucella abortus*, causing a chronic fatigue syndrome, the combined effect of HET and IFN-γ increased the activity of thymic NK cells [23]. HET treatment increased the expression of human monocyte-like THP-1 cells on the cell surface of toll-like receptor (TLR) 4, resulting in an increase in receptors responsive to gram-negative bacteria. From this result, it is also considered to activate the protective effect against pathogenic bacteria [24].

Several reports about viral infection also showed that HET was effective for respiratory viral infection via immunomodulation system such as cytokines. As the nasal cavity belongs to respiratory organs, the anti-infective effect against respiratory infections may give some hint to the eradication of nasal colonization. HET administration resulted in improvement of survival rate and survival time with the mouse influenza virus infection. We also found suppression of viral load in bronchoalveolar lavage fluid (BALF) [25]. An increase in lung interleukin (IL)-1β and tumor necrosis factor-α was observed in combination with HET and osetamivir for influenza A virus-infected mice. In addition, hyperactivity of mouse alveolar macrophages was also observed [26]. In the mouse influenza virus infection model, virus titres decreased in BALF with HET administration. HET also stimulated not only the release of type 1 IFN in the lung, but also the anti-inflammatory response derived from granulocyte macrophage colony-stimulating factor. Furthermore, the defensin expression of the antimicrobial peptide was also increased [27]. When mice were infected with rhinovirus, it was thought that HET inhibited intracellular migration of rhinovirus by decreased expression of intercellular adhesion molecule-1 of airway epithelial cells. It also inhibited IL-1β, IL-6, and IL-8 secretion from respiratory epithelial cells. In our results, glycyrrhizae radix that contains glycyrrhizin was not prominent in the effect of HET, but it is reported that glycyrrhizin reduced viral antibody titter [28].

Splenocytes are the major immunomodulation system against bacterial infection [29]. Several crude drug are known to promote immunostimulation of spleen cells. Atractylodes rhizome extract promotes T cell activity by expressing CD28 of T cells in spleen [30]. It also promotes secretion of IL-2, 6, 10, and T cell differentiation via phosphorylation of extracellular signal-regulated kinases [31]. In addition, it reduces the IFN-γ secretion from T cells in helper T (Th) 1 cells and promotes the IL-4 secretion in Th2 cells [32]. Zingiberis rhizoma extract stimulates CD8+ T cells of splenocytes [33,34]. In addition, it is involved in the TLR2/NF-κB pathway by suppressing the expression of TLR2/NF-κB p65 in lung tissue with mouse pneumococcal infection [35]. Bupleuri radix extract reduces the Th1 subunit and increases the Th2 subunit of peripheral blood [36]. It also has B cell mitogenic activity in spleen cells [37]. Moreover, it also has antimicrobial and antiviral action [38]. Our findings also suggested that these constitutional crude drugs of HET were involved in the activation of spleen cells.

Furthermore, three kinds of active crude drugs, Bupleuri radix, Cimicifugae rhizoma, and Zingiberis rhizome, in hochuekkito belong to superfices-syndrome relieving drugs in Kampo medicinal theory, which means that it excludes the *evils* of the inner surface of the body. This traditional medicinal theory may explain that these three crude drugs have the effect of nasal infection of bacteria.

#### **5. Conclusions**

In summary, HET is significantly effective for the treatment of nasal cavity colonization of MRSA in the murine model. We suggest HET as the therapeutic candidate for effective therapy on nasal cavity colonization of MRSA in humans.

**Author Contributions:** M.M., T.K. and T.M. conceived and designed the experiments; M.M. and T.K. performed the experiments and analyzed the data; M.M. and T.M. wrote the paper.

**Funding:** This work was supported by Grants-in-Aid for Scientific Research (JSPS KAKENHI) grant number JP16K09251, and the Research Foundation for Oriental Medicine.

**Acknowledgments:** We thank Masashi Ishihara, and Miwako Fujimura for excellent support through this investigation.

**Conflicts of Interest:** All authors have no conflicts of interest.

#### **References**


© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Optimisation of the Microwave-Assisted Ethanol Extraction of Saponins from Gac (***Momordica cochinchinensis* **Spreng.) Seeds**

#### **Anh V. Le 1,2,\*, Sophie E. Parks 1,3, Minh H. Nguyen 1,4 and Paul D. Roach <sup>1</sup>**


Received: 6 June 2018; Accepted: 1 July 2018; Published: 3 July 2018

**Abstract: Background**: Gac (*Momordica cochinchinensis* Spreng.) seeds contain saponins that are reportedly medicinal. It was hypothesised that the extraction of saponins from powdered Gac seed kernels could be optimised using microwave-assisted extraction (MAE) with ethanol as the extraction solvent. The aim was to determine an appropriate ethanol concentration, ratio of solvent to seed powder and microwave power and time for extraction. Whether or not defatting the Gac seed powder had an impact on the extraction of saponins, was also determined. **Methods**: Ethanol concentrations ranged from 60–100% were used to compare total saponins content (TSC) extracted from full-fat and defatted Gac seeds. Ratios of solvent to Gac seeds ranged from 10 to 100 mL g−<sup>1</sup> and microwave conditions ranged from 1–4 cycles at power levels ranged from 360–720 W, were examined successively to evaluate their efficiency in extracting saponins from full-fat Gac seeds. **Results**: A four-fold higher of TSC was obtained in extracts from full-fat Gac seed powder than from defatted powder (100 vs. 26 mg aescin equivalents (AE) per gram of Gac seeds). The optimal parameters for the extraction of saponins were a ratio of 30 mL of 100% absolute ethanol per g of full-fat Gac seed powder with the microwave set at 360 W for three irradiation cycles of 10 s power ON and 15 s power OFF per cycle. **Conclusions**: Gac seed saponins could be efficiently extracted using MAE. Full-fat powder of the seed kernels is recommended to be used for a better yield of saponins. The optimised MAE conditions are recommended for the extraction of enriched saponins from Gac seeds for potential application in the nutraceutical and pharmaceutical industries.

**Keywords:** Gac seeds; *Momordica cochinchinensis*; saponins; microwave-assisted extraction; optimization

#### **1. Introduction**

*Momordica cochinchinensis* Spreng. is a perennial climber, which belongs to the Cucurbitaceae family. It ranges from China to the Moluccas and has been used in food and traditional medicine in East and Southeast Asia [1]. The most important part of the mature fruit is the red flesh surrounding the seeds, called the aril, which is used as a colorant in rice or as a material for further processing into functional food ingredients. The seeds are not eaten and they are removed from the aril and are mostly considered waste [2]. However, in traditional medicine, Gac seeds are alleged to have a wide array of therapeutic effects for a wide variety of conditions, including fluxes, liver and spleen disorders, hemorrhoids, wounds, bruises, inflammation, swellling and infections [1,3]. Modern science has reported biological activities for Gac seed extracts, including being a gastroprotective agent [4,5] and accelerating the healing of gastric ulcers in rats [6], and possessing antitumour [7], anticancer [8] and anti-inflammatory [9,10] activities.

Gac seed saponins have been reported to be critical constituents in Gac seed extracts, which were responsible for their medicinal properties [9,11]. These constituents of Gac seeds have been investigated by several investigators: two saponins, referred to as momordica saponin I and II, have been isolated and characterised [12], in which momordica saponin I is a major gastroprotective ingredient [5]. Another saponin, karounidiol, a compound possessing cytotoxic activity against human cancer cell lines [13], has been reported to be present in Gac seeds [14]. The potential valuable pharmaceutical properties of the Gac seed saponins warrants investigating how they are best extracted from the seeds i.e., which extraction technique(s) will maximise the yield of saponins.

The conventional extraction technique, in which the solid material is suspended in extraction solvent with no assistance for breaking the cell structure of the solid material, is often associated with a long heating time, which risks the degradation of bioactive compounds. This has led to the proposed use of advanced techniques such as microwave-assisted extraction (MAE) and ultrasonic-assisted extraction (UAE) that are efficient in terms of extraction time and solvent consumption. Microwave heating or ultrasonic cavitation is able to disrupt the plant cell structure via an increase in the internal pressure of the cell and thereby, release the bioactive compounds [15,16]. However, in a comparative study being carried out by the same authors [17], it was found that while the MAE significantly improved Gac seed saponin extraction in comparison to the conventional method, UAE did not. MAE, therefore, is the technique which needs to be further optimised. The MAE method is likely to be effective for the extraction of saponins from the Gac seeds, as it has been reported that microwave assistance significantly improved the recovery of saponins from a wide range of plant sources such as *Phyllanthus amarus* [18], yellow horn [19], *Ganoderma atrum* [20], chick pea [21] and ginseng [22], among others.

The choice of the extraction solvent is also important. Low alcohols such as methanol and ethanol have usually been used as effective solvents for the extraction of saponins from plant materials. However, according to the US Food and Drug Administration [23], methanol belongs to the Class 2 solvents, which should be limited in pharmaceutical products because of their inherent toxicity. Ethanol, on the other hand, belongs to the Class 3 solvents [23], which are less toxic and of lower risk to human health and therefore, should be used instead of methanol for the extraction of plant bioactive compounds. Moreover, ethanol in form of wines has been traditionally used for maceration of Gac seeds, therefore, it is reasonable to investigate the efficiency of this solvent for modern extraction methods. In addition, ethanol is also an excellent microwave absorbing solvent and has been used to advantage in MAE [16].

When it comes to extraction of saponins from seeds, defatting is often carried out before the saponins are extracted [24]. Although the defatting might make it simpler for the saponin extraction in terms of technique, and does not greatly affect the saponin yield for some type of seeds, it can cause a great loss of saponin for others.

Therefore, in this study, the extraction of saponins from powdered Gac seed kernels was optimised using MAE with ethanol as the extraction solvent. The aim was to determine an appropriate ethanol concentration, ratio of solvent to seed powder and microwave power and time for saponin extraction. Whether or not defatting the Gac seed powder had an impact on the extraction of saponins, was also determined.

#### **2. Materials and Methods**

#### *2.1. Materials*

#### 2.1.1. Solvents, Reagents and Chemicals

Absolute ethanol (≥99.8%), methanol and chemicals including vanillin, sulphuric acid, and potassium persulfate were products of Merck (Bayswater, VIC, Australia) and 2,4,6-tris(2-pyridyl)-*s*-triazine; (±)-6-hydroxy-2,5,7,8-tetramethylchromane-2carboxylic acid (trolox), aescin, 2,2-diphenyl-1-picrylhydrazyl (DPPH), 2,2 -Azino-bis(3ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS) were products of Sigma-Aldrich Co. (Castle Hill, NSW, Australia).

#### 2.1.2. Gac Seed Kernel Powder

Gac seeds, were collected from 450 kg of fresh Gac fruit, from accession VS7 as classified by Wimalasiri, Piva, Urban and Huynh [25]. These fruits were bought at Gac fruit farms in Dong Nai province, Ho Chi Minh city, Vietnam (Latitude: 10.757410; Longitude: 106.673439). After their separation from the fresh fruit, the seeds were vacuum dried at 40 ◦C for 24 h to reduce moisture and increase the crispness of the shell to facilitate shell removal. The dried seeds were de-coated to obtain the kernels, which were then packaged in vacuum-sealed aluminum bags and stored at −18 ◦C prior to use.

#### 2.1.3. Preparation of Gac Seed Kernel Powder

The Gac seed kernels were ground in an electric grinder (100 g ST-02A Mulry Disintegrator), to produce powder, which could pass through a sieve of 1.4 mm. The powder was then freeze-dried using a Dynavac FD3 freeze dryer (Sydney, NSW, Australia) for 48 h at −45 ◦C under vacuum at a pressure loading of 10−<sup>2</sup> mbar (1 Pa) to reduce the moisture content to 1.21 ± 0.02%, as determined using a MOC63u moisture analyser (Shimazdu, Kyoto, Japan). This Gac seed kernel powder was referred to as 'full-fat powder' and was stored in vacuum-sealed polyethylene bags under vacuum at −20 ◦C until used.

#### 2.1.4. Preparation of Defatted Gac Seed Kernel Powder

To prepare defatted Gac seed kernel powder, the freeze-dried kernel powder was extracted three times for thirty minutes with hexane (1:5 *w*/*v*) on a magnetic stirrer at room temperature. Each time, the resulting slurry was suction-filtered and the final residue was air-dried for 12 h and stored in a desiccator at ambient temperature until used. This Gac seed kernel powder was referred to as 'defatted powder'.

#### *2.2. Methods*

The experimental design for the study is shown in Figure 1.

**Figure 1.** Experimental design for optimisation of saponin yield from Gac seeds.

#### 2.2.1. Microwave Assisted Extraction (MAE)

The MAE was performed using a R395YS Sharp Carousel microwave oven (Sharp Corporation, Bangkok, Thailand) bought from a local Target store (Tuggerah, NSW, Australia). Gac seed kernel powder was mixed with ethanol of various concentrations with water in a 100 mL conical flask. The suspension was left pre-leaching for 30 min at ambient temperature before microwave treatment was applied for varying number of cycles, which consisted of 10 s power ON and 15 s power OFF per cycle. The temperature of the suspension was recorded at the end of the MAE process.

#### 2.2.2. Extraction of Saponins from Full-Fat and Defatted Gac Seed Kernel Powders

Prior to weighing for extraction, the moisture content of the powder samples was measured using a MOC63u moisture analyser (Shimazdu, Kyoto, Japan), which was used in the determination of saponin yield.

The effect of the ethanol concentration was investigated for the MAE of saponins from both the full-fat and the defatted powders. The concentration of ethanol was varied (60%, 70%, 80%, 90% and 100%) but the solvent to powder ratio and the microwaving conditions were kept constant at 30 mL g−<sup>1</sup> and 600 W for four cycles, respectively (1st experiment in Figure 1). After finishing the extractions, the suspensions were rapidly cooled to ≤20 ◦C in an ice water bath and filtered through a 0.45 μm membrane filter. The clear extracts were collected and kept at −20 ◦C for analysis within a week.

#### 2.2.3. Extraction of Saponins from the Full-Fat Seed Kernel Powder

The full-fat powder was selected for the following two experiments since it resulted in a higher extraction of saponins for all the concentrations of ethanol; and 100% absolute ethanol was chosen because it resulted in the highest extraction of saponins from the full-fat powder.

Two experiments (Figure 1) were done using the full-fat powder and 100% ethanol as the extraction solvent to determine the effect of three individual parameters, (i) the ratio of solvent to powder (10, 20, 30, 40, 60, 80 and 100 mL g−1) (Figure 1, 2nd Experiment), (ii) microwave radiation power (360, 480, 600, 720 and 840 W) and (iii) microwave irradiation time (1, 2, 3 and four cycles) (Figure 1, 3rd Experiment), on the recovery of saponins from the Gac seed kernel powder was investigated. When one parameter was examined, the other was maintained constant; for the 2nd experiment (Figure 1), the microwave conditions were 600 W with four cycles and for the 3rd experiment, the ratio of ethanol to powder was 30 mL g−1. After finishing the extractions, the suspensions were rapidly cooled to

≤20 ◦C in an ice water bath and filtered through a 0.45 μm membrane filter. The clear extracts were collected and kept at −20 ◦C for less than a week before analysis.

#### 2.2.4. Verifying Optimal Conditions for Gac Seed Saponin Extraction

From the findings in the 3rd experiment, two possible optimal sets of microwave parameters were chosen for the extraction of saponins from the full-fat powder. Therefore, these two sets of microwave parameters were repeated to validate the findings. A control (no microwave) extract was also run with 100% ethanol and the optimal solvent to powder ratio but where the heat was provided using a water bath instead of the microwave oven. The water bath temperature was chosen to be 76 ◦C and the incubation was done for 100 s because it was the maximum temperature and incubation time achieved during the MAE using the two sets of microwave parameters. These three extracts were analysed for TSC and antioxidant capacity–measured with two assays, ABTS and DPPH. The energy consumption for these extracts was also estimated according to the Equation (1) as follows:

$$\mathcal{W}\_{\bar{l}} = P\_{\bar{l}} \times t\_{\bar{l}} \tag{1}$$

where *Wi* is the consumed electrical energy for the extraction method (kWh), *Pi* is the electrical power supplied for the extraction method (kW) and *ti* is the electricity consumption time for the extraction method (h).

#### *2.3. Analytical Methods*

#### 2.3.1. Determination of Total Saponin Content (TSC)

Determination of the total saponin content was conducted using the colorimetric method of Hiai, Oura and Nakajima [26] with slight modifications. The principle of this method is the reaction of sulphuric acid-oxidised saponins with vanillin to produce a distinctive red-purple colour, which is measured at 560 nm using a spectrophotometer.

To 0.25 mL of the appropriately diluted Gac seed ethanol extract samples, 0.25 mL 8% vanillin in ethanol (*w*/*v*) was added followed by 2.5 mL of 72% H2SO4 (*v*/*v*). The test tube was vortexed, covered, incubated at 60 ◦C for 15 min and cooled to ambient temperature in an iced-water bucket for 2 min. With a reagent blank as reference, the absorbance was measured at 560 nm using a Carry 50 Bio spectrophotometer (Varian Pty. Ltd., Mulgrave, VIC, Australia).

A standard curve of aescin (100–1000 μg/mL) was constructed to determine the saponin concentrations. The results were expressed as mg aescin equivalents (AE) per gram dry weight of Gac seed kernel powder (mg AE g<sup>−</sup>1).

#### 2.3.2. Determination of Antioxidant Capacity

The antioxidant capacity was tested for the optimal and control extracts using two assays: ABTS and DPPH.

#### ABTS Assay

The ABTS assay [27] was used as described by Tan et al. [28] with slight modifications. Stock solutions of 7.4 mM ABTS and 2.6 mM potassium persulfate were prepared and kept at 4 ◦C until use. Fresh working solution was prepared for each assay by mixing the 2 stock solutions in equal quantities and incubating them for 15 h in the dark at ambient temperature. Then, 1 mL of the working solution was diluted with ~30 mL of methanol to obtain an absorbance of 1.1 ± 0.02 units at 734 nm. To 0.15 mL of each standard, blank and appropriately diluted extract sample, 2.85 mL of the working solution was added. The tubes were incubated for 2 h in the dark at ambient temperature and the absorption was measured at 734 nm using a Carry 50 Bio spectrophotometer (Varian Pty. Ltd., Mulgrave, VIC,

Australia). Trolox was used as the standard and the results were expressed as mg Trolox equivalents per gram dry weight of Gac seed kernel powder (mg TE g<sup>−</sup>1).

#### DPPH Assay

The DPPH assay [29] was used as described by Tan et al. [28]. A stock solution of 0.6 M DPPH in methanol was prepared and kept at −20 ◦C until use. The working solution was prepared by mixing 10 mL of stock solution with ~45 mL of methanol to obtain an absorbance of 1.1 ± 0.02 units at 515 nm. To 0.15 mL of each standard, blank and appropriately diluted extract sample, 2.85 mL of the working solution was added. The tubes were allowed to stand for 3 h in the dark at ambient temperature and the absorption was measured at 515 nm using a Carry 50 Bio spectrophotometer (Varian Pty. Ltd., Mulgrave, VIC, Australia). Trolox was used as the standard and results were expressed as mg Trolox equivalents per gram of dry weight Gac seed kernel powder (mg TE g<sup>−</sup>1).

#### *2.4. Statistical Analyses*

Experiments were performed in triplicate and values were expressed as means ± SD and were assessed for statistical significance using the one-way ANOVA and Tukey's *Post Hoc* Multiple Comparison test using the IBM SPSS Statistics 24 program (IBM Corp., Armonk, NY, USA). Correlation and regression analyses were done using Microsoft Excel 2016. Differences between means, correlations and regressions were considered statistically significant at *p* < 0.05.

#### **3. Results**

#### *3.1. Effect of the Ethanol Concentration on the MAE of Saponins from Full-Fat and Defatted Gac Seed Kernel Powders*

The full-fat and defatted Gac seed kernel powders were extracted using MAE with the ethanol concentration ranging from 60% to 100% (in water) in the extraction solvent. Figure 2 shows that at the lower ethanol concentrations, from 60% to 80%, there was no significant difference in the measured TSC for the full-fat powder. The measured TSC was higher with 90% ethanol and the highest (100.3 mg AE g−1) with 100% ethanol as the extraction solvent. In contrast, changing the ethanol concentration from 60% to 100% did not increase the measured TSC of the defatted Gac seed kernel powder, which was lower than for the full-fat powder for all the ethanol concentrations. Therefore, the full-fat Gac seed kernel powder and 100% absolute ethanol, as the extraction solvent, gave the best MAE extraction of saponins and they were used in the subsequent experiments.

**Figure 2.** Effect of the ethanol concentration, in the extraction solvent used for microwave-assisted extraction (MAE), on the measured total saponin content (TSC) of the full-fat and defatted Gac seed kernel powders. The values are the means of three replicates for each extraction and columns not sharing the same superscript letter are significantly different at *p* < 0.05.

#### *3.2. Effect of the Ethanol to Sample Ratio on the MAE of Saponins from the Full-Fat Gac Seed Kernel Powder*

Seven ratios of 100% absolute ethanol to full-fat powder, from 10 to 100 mL g<sup>−</sup>1, were investigated. Figure 3 shows that increasing the ratio from 10 to 30 mL g−<sup>1</sup> had a significant effect on the measure TSC value after MAE, which increased by 30% from 70.4 to 100.8 mg AE g<sup>−</sup>1. However, increasing the ratio from 30 to 100 mL g−<sup>1</sup> resulted in less pronounced increases in the measured TSC. Therefore, although the measured TSC was slightly and significantly higher with the ratio of 100 mL g−<sup>1</sup> compared to 30 mL g−<sup>1</sup> (Figure 3), the ratio of ethanol to powder of 30 mL g−<sup>1</sup> was deemed to be the better ratio, from the conservation of solvent perspective, and it was chosen for investigating the microwave parameters.

**Figure 3.** Effect of the ethanol to powder ratio on the TSC of the full-fat Gac seed kernel powder measured using MAE. The values are the means of three replicates for each extraction and columns not sharing the same superscript letter are significantly different at *p* < 0.05.

#### *3.3. Effect of the Microwave Parameters on the MAE of Saponins from the Full-Fat Gac Seed Kernel Powder*

Four levels of microwave power (360, 480, 600 and 720 W) were investigated and at every power level, the number of irradiation cycles was also varied (1, 2, 3 and four cycles). Each cycle consisted of 10 s power ON (irradiation) followed by 15 s power OFF (no irradiation). The full-fat powder was used and the ratio of 100% ethanol to powder was 30 mL g−1. In general, Figure 4 shows that the measured TSC gradually increased as the power and irradiation time were increased for the MAE but that many of the values were not significantly different from each other. Notably, from 600 W to two cycles upwards (to the right in Figure 4), there was no significant increase in the measured TSC values. However, the two sets of parameters, which only shared the a superscript in Figure 4, 360 W and three cycles and 480 W and four cycles, were selected as possibly optimal for the MAE extraction of saponins from the full-fat Gac seed kernel powder.

**Figure 4.** Effect of microwave power and irradiation time (cycles) on the TSC of the full-fat Gac seed kernel powder measured using MAE. The values are the means of three replicates for each extraction and columns not sharing the same superscript letter are significantly different at *p* < 0.05.

#### *3.4. Correlations between the TSC and the MAE Temperature*

The temperature of the extracts at the end of each MAE in Figure 4 was recorded using a digital thermometer. Their temperature ranged from 43.4 to 75.6 ◦C. Correlation analysis revealed that the measured TSC of the extracts was positively correlated with the temperature of the extraction mixture at the end of the MAE (Figure 5).

**Figure 5.** Correlation between the TSC and the temperature of the extract at the end of various MAE treatments. The black dots: TSC at different temperature of the extracts.

Table 1 shows that the temperature of the extracts at the end of the MAE was almost all due (92.5%) to the number of irradiation cycles (length of the microwave irradiation time) during the MAE; in contrast, there was no correlation between the temperature and the microwave power. Consisted with this, the measured TSC of the extracts was positively correlated with the number of microwave irradiation cycles but wasn't correlated with the power used during the MAE. Moreover, there was no interaction between the microwave power and the number of irradiation cycles.


**Table 1.** Correlations between the TSC and the MAE parameters.

*3.5. Verification of the Optimal MAE Conditions for the Extraction of Saponins from Full-Fat Gac Seed Kernel Powder*

Two possible optimal sets of microwave parameters (360 W and three cycles, 480 W and four cycles) were chosen for the extraction of saponins from the full-fat powder (Figure 4). Notably, the temperature measured at the end of the MAE using the two sets of microwave parameters (360 W and three cycles, 480 W and four cycles) was 72.2 ± 1.2 and 75.6 ± 1.9 ◦C, respectively, and they were not significantly different from each other.

These two sets of microwave parameters were repeated to validate the findings. A control (no microwave) set of extracts was also run with 100% ethanol and the optimal solvent to powder ratio where the temperature measured at the 480 W and four cycles MAE (76 ◦C) was provided using a water bath instead of the microwave oven. Also, the time used for the control extraction was chosen to be 100 s in order to match the length of time used for the 480 W and four cycles MAE.

These three sets of extracts were analysed for their saponin content and their ABTS and DPPH antioxidant activities. The results revealed that there was no difference among the three extracts in saponin content and antioxidant capacity (Table 2). However, the ABTS values were low and the DPPH assay did not detect any antioxidant activity for any of these extracts (Table 2).

**Table 2.** Saponin content, antioxidant activities and energy consumption of the optimal MAE and control extracts.


The results are mean values ± standard deviations (*n* = 3) and the values not sharing the same superscript letter in the same column, are significantly different at *p* < 0.05. † Ethanol + Full-fat powder (30 mL g−1); MAE at 360 W, three cycles for 75 s. ‡ Ethanol + Full-fat powder (30 mL g−1); MAE at 480 W, four cycles for 100 s. § Ethanol + Full-fat powder (30 mL g−1); Shaking water bath at 76 ◦C for 100 s.

From the point of view of saving energy, the optimal treatment 1 MAE parameters (Table 2) of 360 W with three irradiation cycles of 10 s power ON and 15 s power OFF per cycle (total of 75 s), were the best microwave conditions for the extraction of saponins from full-fat Gac seed kernel powder (0.003 kWh). The conventional extraction for 100 s in a shaking water bath at the same temperature (76 ◦C) as at the end of the optimal treatment 2 MAE settings also gave the same results but this also required more energy than the optimal treatment 1 MAE parameters because of the energy needed (0.325 kWh) to bring the temperature of the 5 L water bath from 20 ◦C up to 76 ◦C.

#### **4. Discussion and Conclusions**

The full-fat Gac seed kernel powder was the more suitable material to use as defatting caused a considerable loss of saponins (~75%). The highest TSC of extracts were obtained with 100% absolute ethanol, a 30 mL g−<sup>1</sup> ratio of ethanol to full-fat Gac seed kernel powder and several sets of MAE conditions. Furthermore, when two sets of the MAE parameters, which gave the highest measured TSC in the extracts, were re-tested, the two extracts had the same TSC. However, from the point of view of saving energy, the optimal 1 MAE parameters of 360 W with three irradiation cycles of 10 s power ON and 15 s power OFF per cycle (total of 75 s), were the best conditions for the extraction of saponins from the full-fat powder.

It was concluded that, to extract Gac seed saponins, it was better to use the full-fat seed kernel powder rather than kernel powder from which the fat had been extracted. The Gac seed saponins appeared to be mainly associated with the fat component of the seeds because they were largely lost during the defatting process with hexane. Undoubtedly, most of the Gac seed saponins (75%) were highly non-polar, which is consistent with a previous finding that saponins were found in the unsaponifiable matter from Gac seed oil [14]. The saponin content of the full-fat Gac seed kernels was also similar to that reported for other oily plant extracts, such as eucalyptus [30] and *Phyllanthus amarus* [31], and significantly higher than the non-oily extract from the flesh of bitter melon [28].

Absolute ethanol was found to be the best concentration of ethanol for extracting the saponins from the full-fat Gac seed kernel powder. This is also consistent with the Gac seed saponins being hydrophobic in nature and consistent with Gac seeds having a high fat content [32]. This result is also consistent with the earlier findings that Gac seed oil has a high content of unsaponifiable matter [33] and that this unsaponifiable material contains triterpenoid saponins [14]. It may be that more non-polar class 3 solvents, such as 1-propanol, isobutyl alcohol and n-butanol, could further improve the extraction of saponins from full-fat Gac seed kernels. However, because of the higher costs of these solvents, ethanol would be the solvent of choice for recovery of the Gac seed saponins on economic grounds.

The ratio of 30 mL ethanol per 1 g of Gac seed powder was the ratio of choice because, at this ratio, the saponin yield was improved significantly compared to the two lower ratios and there was not much improvement at the higher ratios. Although it varies for different plant materials, in the conventional extraction method, the higher the ratio of solvent volume to solid sample the better the extraction of compounds is. However, in the case of MAE, a higher solvent: sample ratio may not necessarily give a better yield due to non-uniform distribution and exposure to microwaves [16].

Varying the MAE parameters did not greatly affect the saponin extraction, mostly likely due to ethanol being a very good solvent [16] for the extraction of the Gac seed saponins and for increasing in temperature even under mild microwave conditions. When the optimal MAE conditions were compared with a control extraction (no microwave), it was found that the same level of saponin extraction was achieved irrespective of the heating source. However, from the point of view of saving energy, the optimal 1 MAE parameters were the best conditions for the extraction of saponins from full-fat Gac seed kernel powder. The energy saving characteristic of MAE has been confirmed in numerous reports [15]. This is due to the heat is generated inside the materials and then comes outwards, whereas in conventional heating the surface is heated first. Thus, the microwave heating is rapid and effective as heat is transferred directly to the material.

The antioxidant capacity of the Gac seed saponin extracts was low. This is possibly due to the more lipophilic nature of the Gac seed saponins, which is consistent with the findings that lipophilic compounds such as carotenoids, which do not show DPPH-radical scavenging activity [34], and tocopherols, which do not show much activity in the ABTS assay [35].

In conclusion, this study demonstrated that the extraction parameters play an important role in the extraction of saponins from Gac seeds. Accordingly, the MAE optimal parameters for the extraction of saponins were a ratio of 30 mL of 100% absolute ethanol per g of full-fat Gac seed kernel powder with the microwave set at 360 W for three irradiation cycles of 10 s power ON and 15 s power OFF per cycle. These parameters are recommended for the extraction of enriched saponins from Gac seeds for potential application in the nutraceutical and pharmaceutical industries.

**Author Contributions:** Conceptualization, A.V.L., M.H.N. and P.D.R.; Methodology, A.V.L.; Validation, A.V.L.; Formal Analysis, A.V.L.; Investigation, A.V.L.; Data Curation, A.V.L. and P.D.R.; Writing-Original Draft Preparation, A.V.L.; Writing-Review & Editing, P.D.R., M.H.N. and S.E.P.; Supervision, P.D.R., M.H.N. and S.E.P.

**Funding:** This research received no external funding.

**Acknowledgments:** A.V.L. acknowledges the University of Newcastle and VIED for their financial support.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

*Medicines* Editorial Office E-mail: medicines@mdpi.com www.mdpi.com/journal/medicines

MDPI St. Alban-Anlage 66 4052 Basel Switzerland

Tel: +41 61 683 77 34 Fax: +41 61 302 89 18