Phenolic profiling and in vitro bioactivity of Moringa oleifera leaves as affected by different extraction solvents

Contents lists available atScienceDirect

Food Research International

Phenolic pro

filing and in vitro bioactivity of Moringa oleifera leaves as

a

ffected by different extraction solvents

Gabriele Rocchetti

, Jorge Pamplona Pagnossa

, Francesca Blasi

, Lina Cossignani

,

Roberta Hilsdorf Piccoli

, Gokhan Zengin

, Domenico Montesano

, Pier Sandro Cocconcelli

,

Luigi Lucini

a_{Department for Sustainable Food Process, Università Cattolica del Sacro Cuore, Via Emilia Parmense 84, 29122 Piacenza, Italy} b_{Food Science Department, University of Lavras (UFLA), Campus Universitário, CEP 37.200-000 Lavras, MG, Brazil}

c_{Department of Pharmaceutical Sciences, Food Science and Nutrition Section, University of Perugia, Via S. Costanzo 1, 06126 Perugia, Italy} d_{Department of Biology, Faculty of Science, Selcuk University, Campus, Konya, Turkey}

A R T I C L E I N F O

Keywords: Moringa Food metabolomics Enzymatic activity

Antioxidant activity, antimicrobial activity

A B S T R A C T

In this work the (poly)-phenolic profile of Moringa oleifera leaves was comprehensively investigated through untargeted metabolomics, following a homogenizer-assisted extraction (HAE) using three solvent systems, i.e. methanol (HAE-1), methanol-water 50:50 v/v (HAE-2) and ethyl acetate (HAE-3). This approach allowed to putatively annotate 291 compounds, recording mainlyflavonoids and phenolic acids. Thereafter, antioxidant capacity, antimicrobial activity and enzyme inhibition were assayed in the different extracts. HAE-1 extract showed the highest total phenolic content (31.84 mg/g), followed by HAE-2 (26.95 mg/g) and HAE-3 (14.71 mg/g). In addition, HAE-1 and HAE-2 extracts exhibited an expressive activity against Bacillus cereus and Listeria innocua. The HAE-2 leaf extract was characterized by the highest DPPH and ABTS values (being 49.55 and 45.26 mgTE/g), while ferric reducing antioxidant power was found to be higher in HAE-1 (58.26 mgTE/g). Finally, the enzyme inhibitory effects of M. oleifera leaf extracts were investigated against five enzymes, namely acetylcholinesterase (AChE), butyrylcholinesterase (BChE), tyrosinase,α-amylase and α-glucosidase. All of the tested extracts exhibited inhibitory effects on AChE and BChE with a higher activity for HAE-3 and HAE-1, whilst HAE-1 showed the higher impact on tyrosinase, glucosidase and amylase activities. Taken together, these findings suggest that M. oleifera leaf extracts are a good source of bioactive polyphenols with a potential use in food and pharma industries.

1. Introduction

Moringa oleifera Lam. (Moringa) represents an important multi-tasking crop (Falowo et al., 2018). This plant is considered a precious species because itsflowers, leaves, seeds and pods are edible and ex-tremely rich in macro/micronutrients and bioactives (Montesano, Cossignani, & Blasi, 2019). In addition, various applications in humans (food fortificant, cooking oil, cosmetic), agriculture (water purification, biofuel, fertilizer), and livestock (feed ingredient, fodder, antibiotic) have been reported (Falowo et al., 2018; Valenga, Boschen, Rodrigues, & Maia, 2019).

The different M. oleifera parts are mainly used as functional in-gredients for the production of dietary supplements (Abraham, Abu, & Gernah, 2013; Hassan, Bayoumi, Abd El-Gawad, Enab, & Youssef, 2016), mainly due to their wide health-promoting properties. In this

regard, Moringa leaves are characterized by high amounts of different (poly)-phenolic compounds (mainly flavonoids) (Djande, Piater, Steenkamp, Madala, & Dubery, 2018; Rocchetti, Blasi et al., 2019; Rocchetti, Lucini, Rodriguez, Barba, & Giuberti, 2019), together with other antioxidant compounds such as ascorbic acid and carotenoids. However, the bioactive composition of plants depends on several fac-tors such as plant physiological stage, pedoclimatic conditions, and geographical origin (Blasi, Urbani, Simonetti, Chiesi, & Cossignani, 2016; Lombardi et al., 2017), despite the extraction methods and conditions play a crucial role.

Recently, extraction techniques being safer and eco-friendlier have been developed (Putnik et al., 2018). For this reason, non-conventional extraction methods have been used for the recovery of phenolic from Moringa leaves, as a valid alternative to conventional procedures (Rocchetti, Blasi et al., 2019; Rocchetti, Lucini, et al., 2019; Zhu, Li, He,

https://doi.org/10.1016/j.foodres.2019.108712

Received 25 June 2019; Received in revised form 1 September 2019; Accepted 25 September 2019 ⁎_{Corresponding authors.}

E-mail addresses:gabriele.rocchetti@unicatt.it(G. Rocchetti),domenico.montesano@unipg.it(D. Montesano),luigi.lucini@unicatt.it(L. Lucini).

Available online 31 October 2019

0963-9969/ © 2019 Elsevier Ltd. All rights reserved.

Montesano, & Barba, 2018). In addition, as reviewed byRani, Husain, and Kumolosasi (2018), various research has been conducted to eval-uate the potential of M. oleifera extracts as both antioxidant and anti-microbial agents. For example,Wang, Chen, and Wu (2016)evaluated the antibacterial activities of M. oleifera extracts (e.g. methanol, chloroform, ethyl acetate and aqueous) against both Gram-negative and Gram-positive bacteria.

Interestingly, it has been reported also that polyphenols from M. oleifera might be able to inhibit intestinal activities of two key-enzymes involved in the digestion of carbohydrates such asglucosidase and α-amylase (Adisakwattana & Chanathong, 2011). Beyond nutritional benefits, flavonoids display also a variety of biological activities both in vitro and in vivo. For example,Uriarte-Pueyo and Calvo (2011)reported that these bioactives are potential acetylcholinesterase inhibitors, while Chiocchio et al. (2018)evaluated also their inhibitory activity vs. tyr-osinase, a biological-target of cosmetic interest for the reduction of hyperpigmentation and melanin biosynthesis.

In a recent work (Rocchetti, Blasi et al., 2019; Rocchetti, Lucini, et al., 2019), we evaluated the impact of different extraction methods, namely maceration, homogenizer-assisted extraction, rapid solid-liquid dynamic extraction, microwave-assisted extraction and ultrasound-as-sisted extraction, on polyphenols of M. oleifera leaves, revealing that homogenizer-assisted extraction (HAE; with methanol 100% and me-thanol:water 50:50, v/v) produced the highest phenol contents, when compared to other extraction methods. In this work, we selected again the most performing extraction method (HAE), using methanol 100% and methanol: water 50:50, v/v as extraction solvents. In addition, ethyl-acetate was also used in order to use a solvent with a different polarity, able to provide a different extraction efficiency compared with methanol and the hydro-alcoholic mixture. Therefore, the aim of this work was to evaluate the impact of the extraction solvent (i.e., me-thanol 100%, meme-thanol/water 50:50 v/v, and ethyl acetate 100%) on the comprehensive recovery of phenolics from M. oleifera leaves. To this purpose, untargeted metabolomics was used to depict in an unbiased manner the phenolic composition. Afterwards, enzymatic, antioxidant and antimicrobial activities have been evaluated on the same extracts. Thus, the novelty of the work lies with the application of untargeted metabolomics to correlate the phenolic composition of different M. oleifera leaf extracts with the corresponding biological activity. The data reported here could be of great importance for the utilization of M. oleifera leaf extracts as food preservatives and nutraceuticals. 2. Material and methods

2.1. Samples

Fresh leaves of M. oleifera Lam. were provided by the Sud Rienergy S.r.l. - Favella Group, a farm located in Southern Italy (Corigliano Calabro, Cosenza, Italy– 39°129 13′ 27,69″ N and 9° 01′ 29,69″ E). Three samples of Moringa leaves were randomly collected in the orchard from different plants from June to September 2017. The sam-ples were taxonomically identified by Nicola Rizzo (Sud Rienergy farm) and representative specimens were deposited at the Orto Botanico, Centro di Ateneo per i Musei Scientifici, University of Perugia (Italy). Intact leaves were dried in a ventilated oven at 60 °C for 24 h (Olabode, Akanbi, Olunlade, & Adeola, 2015), up to constant weight. Afterwards, dried leaves were milled in a blender and passed through 250μm sieve to obtain a thin powder having a moisture 10 ± 1%. Samples were preserved in amber glass containers, in a dry place at room tempera-ture, until further analyses.

2.2. Extraction of polyphenols

A Homogenizer-Assisted-Extraction (HAE) was used to extract polyphenols from M. oleifera leaves. In particular, three replicates (1.0 g) of dried Moringa leaves were extracted in 10 mL of methanol

100% (HAE-1), methanol/water 50:50 v/v (HAE-2), and ethyl acetate 100% (HAE-3) by using an Ultra-turrax (Ika T25, Staufen, Germany) at 5000g for 3 min. The extracts were then centrifuged (Eppendorf 5810R, Hamburg, Germany) at 10,000g for 10 min at 4 °C. Subsequently, the extracts were concentrated using a rotary evaporator equipment until total dryness, followed by a 2 mL resuspension in each solvent. The resulting solutions were collected in amber vials until further analyses. 2.3. Untargeted profiling of M. oleifera leaves by UHPLC-QTOF mass spectrometry

An aliquot of each Moringa leaf extract wasfiltered using 0.22 μm cellulose syringefilters and then injected in a 1290 liquid chromato-graph coupled with a G6550 mass spectrometer detector via a Dual Electrospray Jet Stream ionization system (all from Agilent Technologies, Santa Clara, CA, USA). The instrumental conditions for the analysis of polyphenols in plant extracts by untargeted metabo-lomics were optimized in a previous work (Rocchetti et al., 2017; Rocchetti, Blasi et al., 2019; Rocchetti, Lucini, et al., 2019). The mass spectrometer worked in FULL SCAN mode with a nominal resolution at 30,000 FWHM and in positive polarity. For the mass-acquisition, a range of 100–1100 m/z was set up. The injection volume was 6 μL considering three replicates for each sample.

Raw metabolomic data were aligned and deconvoluted using the Agilent Profinder B.06 software and then annotated according to the ‘find-by-formula’ algorithm, using the combination of monoisotopic accurate mass and the entire isotopic pattern against Phenol-Explorer 3.6, the comprehensive database on phenolic compounds available online (http://phenol-explorer.eu/) (Rocchetti, Blasi et al., 2019; Rocchetti, Lucini, et al., 2019). The approach used allowed great con-fidence in the annotation. In particular, the compounds were identified according to a Level 2 of identification (i.e., putatively annotated compounds; Salek, Steinbeck, Viant, Goodacre, & Dunn, 2013). The Agilent Profinder B.06 software was used also for the post-acquisition datafiltering: only those compounds putatively annotated within 100% of replications in at least one condition were considered.

Afterwards, the phenolics ascribed into classes were then quantified using methanolic standard solutions (methanol/water 80:20, v/v) of single pure standard compounds analyzed through the same method (Rocchetti, Blasi et al., 2019; Rocchetti, Lucini, et al., 2019). The standards prepared were representative of the following phenolic sub-classes: anthocyanins (cyanidin), flavanols and flavonols (catechin), flavones (luteolin), phenolic acids (ferulic acid), lignans (sesamin), al-kylphenols (cardol), stilbenes (resveratrol) and tyrosols. Results were finally expressed as mg phenolic equivalents/g dry matter (DM). 2.4. Bacterial strains and antimicrobial activity

Antimicrobial activity was tested using well diffusion assay method as described byValgas, Souza, Smânia, and Smânia (2007), with slight modifications, employing 24 h cultures of four indicator bacterial strains: Listeria innocua UC-8409; Bacillus cereus UC-4044; Salmonella Enteritidis S64; and Salmonella Typhimurium S190. The strains were obtained from Culture collection of Università Cattolica del Sacro Cuore (Piacenza, Italy) and LABENT (Laboratory of Enterobacteria of Oswaldo Cruz Foundation, Rio de Janeiro, Brazil). The chosen test microorgan-isms were inoculated into sterile BHI agar medium (Oxoid, Canada) by uniformly mixing 100μL of the inoculums containing 1 × 107_CFU/mL (600 nm) in Petri dishes. M. oleifera concentrated extracts of methanol 100% (HAE-1), methanol/water 50:50 v/v (HAE-2), and ethyl acetate 100% (HAE-3) were used as testing antimicrobial agents and negative controls were the respective solvents of extraction. The positive control of ampicillin used was 1 mg/mL in water. Aliquots of 15μL of the Moringa concentrated extracts were added into 5 mm diameter wells made on agar medium and plates were maintained at room temperature for 2 h to allow the diffusion of the solutions into the medium. The

plates were then incubated at 37 °C for 48 h and the zones of inhibition (ZOI) were measured. The test was carried out considering triplicate values, and mean diameter of the inhibition zones was expressed in millimeters.

2.5. Determination of in vitro antioxidant and enzyme inhibitory effects The metal chelating, phosphomolybdenum, ferric reducing anti-oxidant power (FRAP), cupric ion-reducing antianti-oxidant capacity (CUPRAC), ABTS, and DPPH activities of the extracts were assessed following the methods described byUysal et al. (2017). The antioxidant activities were reported as trolox equivalents, whereas EDTA was used for metal chelating assay. The possible inhibitory effects of the extracts against cholinesterase (by Ellman’s method), tyrosinase, α-amylase and α-glucosidase were evaluated using standard in vitro bioassays (Uysal et al., 2017).

2.6. Statistical analysis

A one-way analysis of the variance (ANOVA) was done using the software PASW Statistics 25.0 (SPSS Inc.) to investigate significant differences (p < 0.05, Duncan's post hoc test) when considering the results of each in vitro assay and total phenolic contents. Pearson's correlation coefficients (p = 0.01, two-tailed) were also calculated by using the same statistical software.

The Mass Profiler Professional B.12.06 (Agilent technologies) was then used for the elaboration of the untargeted UHPLC/QTOF data by unsupervised hierarchical cluster analysis (HCA), based on the fold-change heat map, as previously reported (Rocchetti, Blasi et al., 2019; Rocchetti, Lucini, et al., 2019). The raw metabolomic dataset was ex-ported and elaborated into SIMCA 13 software (Umetrics, Malmo, Sweden) by supervised orthogonal projections to latent structures dis-criminant analysis (OPLS-DA) multivariate statistics. Data normal-ization and cross-validation of the OPLS-DA model rigorously followed the experimental plan reported in our previous work (Rocchetti, Blasi et al., 2019; Rocchetti, Lucini, et al., 2019). In addition, model para-meters regarding both goodness of fit and prediction (R2Y and Q2Y,

respectively) were also recorded. Afterwards, the variables selection methods namely VIP (i.e., variable importance in projection) was used to select those polyphenols having the highest discrimination potential (VIP score > 1), i.e., the most affected by the extractions used. Finally, a Fold-Change (FC) analysis was carried out considering the UHPLC/ QTOF data on polyphenols, in order to assess the impact of each ex-traction solvent on the different phenolic sub-classes.

3. Results and discussion

3.1. Screening of M. oleifera leaf polyphenols by different extraction solvents

The untargeted phenolic profile of the different Moringa extracts provided 291 annotated compounds, being 39 anthocyanins, 66 fla-vones, 47 flavonols, 15 lignans, 7 alkylphenols, 49 lower-molecular-weight compounds, 63 phenolic acids (mainly hydroxycinnamics) and 5 stilbenes. The polyphenols annotated are provided insupplementary material, together with their composite mass spectra. The dried leaves of M. oleifera are known to be a great source of phenolic compounds (Rocchetti, Blasi et al., 2019; Rocchetti, Lucini, et al., 2019; Vergara-Jimenez, Almatrafi, & Fernandez, 2017); the mainflavonoids reported in Moringa leaves are myricetin, quercetin and kaempferol derivatives, while gallic and chlorogenic acid are reported to be the most abundant phenolic acids. In this regard, looking at the phenolic profile outlined by UHPLC-QTOF mass spectrometry, we found great abundances of dihydromyricetin 3-O-rhamnoside, quercetin 3-O-rhamnoside, glyco-sidic forms of kaempferol, gallic acid 4-O-glucoside and isomeric forms of caffeoylquinic acid (supplementary material), thus corroborating what reported in literature.

The metabolomic dataset was then elaborated by means of un-supervised hierarchical cluster analysis (HCA; produced from the fold-change heat map) in order to group samples according to intrinsic si-milarities in their measurements. The HCA heat map is reported as Fig. 1; as can be observed, a clear separation of the different Moringa extracts was achieved, revealing that each solvent tested (i.e. MeOH 100%, methanol: water 50:50, v/v, and ethyl acetate) was able to Fig. 1. Non-averaged unsupervised hierarchical cluster analysis (HCA) on the phenolic profile of different Moringa leaf extracts (similarity: ‘Euclidean’; linkage rule: ‘Ward’). Compounds intensity was used to build up heat map, on the basis of which the clusters were generated.

promote the selective extraction of specific cluster of compounds. In particular, according to the different polarity index of the solvents used (methanol: water 50:50, v/v > MeOH > ethyl acetate), we found that HAE-3 samples showed the most differential profile (outlined in the first cluster of HCA;Fig. 1), while HAE-1 and HAE-2 provided similar fold-change values and then included in the same cluster (i.e. the second one) of the HCA, according to their phytochemical profiles. The semi-quantitative values for each phenolic class were then obtained from the UHPLC-QTOF data and are reported inTable 1, expressed as mg phenolic equivalents/g dry matter (DM). Overall, HAE-1 extract was characterized by the highest total phenolic content (31.84 mg/g), fol-lowed by HAE-2 (26.95 mg/g) and HAE-3 (14.71 mg/g) extracts. Looking in detail to each phenolic subclass, flavonoids (such as an-thocyanins and flavonols/flavanols) were better extracted by polar solvents, whilst HAE-1 samples showed the highest values (p < 0.05) of phenolic acids and tyrosol equivalents (i.e. 10.60 and 9.90 mg/g DM, respectively). Interestingly, HAE-3 and HAE-2 samples presented the highest values of stilbene equivalents. Thesefindings fitted with pre-vious works (Rocchetti, Blasi et al., 2019; Rocchetti, Lucini, et al., 2019; Castro-López et al., 2017), highlighting the potential of different ex-traction methods (conventional vs non-conventional) to recover poly-phenols from Moringa leaves. It is also important to emphasize that our data are difficult to compare with the existing literature, considering that only few works adopted a similar analytical workflow (Lin et al.,

2019; Makita, Chimuka, Steenkamp, Cukrowska, & Madala, 2016); in this regard, the most of research carried out on Moringa leaves still used in vitro spectrophotometric assays (such as Folin-Ciocalteu) to assess the total phenolic composition of this matrix. However, as suggested by Granato et al. (2018), in vitro spectrophotometric methods should be used only as preliminary investigation, and necessarily followed by high-resolution chromatographic methods to depict in detail the phe-nolic profile. In previous studies,Zullaikah et al. (2019)demonstrated an enhancement of the extraction of total phenolic compounds (mainly flavonoids) and in vitro antioxidant activity from Moringa oleifera leaves by using a subcritical water-ethanol mixture, whilst other authors (Leone et al., 2015; Lin et al., 2019) observed differences in terms of nutrients and phenolic compounds in M. oleifera leaves grown in dif-ferent countries, thus confirming that besides the extraction conditions, the actual phenolic composition of M. oleifera leaves is strongly affected by several other factors, such as genetic background and/or pedo-cli-matic conditions.

Afterwards, considering the clear differences outlined by both un-supervised HCA and semi-quantitative values for each phenolic class, a supervised multivariate statistical approach, named OPLS-DA, was used to group and discriminate each sample according to the corresponding phenolic profile. The OPLS-DA score plot obtained by using the dif-ferent extraction solvent as class membership criterium is provided in Fig. 2. This plot confirmed what previously reported, i.e. that each extraction solvent was able to promote different phytochemical pro-files. In particular, the first latent vector (t0) discriminated HAE-2 and HAE-3 from HAE-1, while the second latent vector (t1) outlined a dif-ferential profile of HAE-2 sample. Overall, the OPLS model built was characterized by excellent accuracy parameters, being R2_Y (goodness-of-fit) = 0.99 and Q2_{Y (goodness-of-prediction) = 0.98. Besides, the} model was cross validated and inspected for outliers (supplementary material). The following variable selection method VIP (i.e. variables importance in projection) was used to highlight the (poly)-phenolic compounds most affected by the different extraction conditions. In this regard, the VIP markers, grouped in phenolic subclasses, are reported in Table 2, together with their score (> 1.1), standard error and log Fold-Change (FC). As can be observed from the table, 68 compounds were found to be the most affected by the different extraction solvents used. In particular, the VIP markers included flavonoids (anthocyanin and flavone equivalents) and phenolic acids (mainly hydroxycinnamics). The VIP approach was particularly useful in identifying those com-pounds selectively extracted by the different solvents. For example, HAE-3 promoted the highest recovery of two stilbenes when compared Table 1

Total phenolic contents considering each phenolic subclass in the different Moringa leaf extracts tested (i.e. HAE-1, HAE-2 and HAE-3). Data (expressed on a dry matter basis, DM) are presented as mean values (mg/g equivalents) ± standard deviation (n = 3). Superscript letters within each row indicate homogeneous sub-class as resulted from ANOVA (p < 0.05), Duncan's post-hoc test.

Phenolic class HAE-1 (MeOH 100%) HAE-2 (MeOH: H2O 50:50, v/v) HAE-3 (EtOAc 100%) Anthocyanins 2.37 ± 0.35b _{2.57 ± 0.15}b _{0.87 ± 0.05}a Flavonols 2.40 ± 0.10b _{4.93 ± 0.35}c _{0.53 ± 0.06}a Flavones 3.40 ± 0.69a _{2.90 ± 0.17}a _{2.87 ± 0.40}a Lignans 1.17 ± 0.15ab _{1.97 ± 1.10}b _{0.07 ± 0.05}a Tyrosols 9.90 ± 2.06b _{6.01 ± 1.65}a _{3.07 ± 0.23}a Alkylphenols 1.47 ± 0.06c _{1.01 ± 0.10}b _{0.10 ± 0.00}a Phenolic acids 10.60 ± 1.30b _{6.23 ± 2.55}a _{5.77 ± 1.01}a Stilbenes 0.53 ± 0.40a _{1.33 ± 0.06}b _{1.43 ± 0.45}b Total 31.84 26.95 14.71

Fig. 2. Orthogonal projection to latent structures discriminant analysis (OPLS-DA) on different Moringa leaf extracts phenolic profiles. Individual replications are given in the class prediction model score plot.

Table 2

Discriminant polyphenols identified by VIP (variable importance in projection) selection method following supervised OPLS-DA modelling. Compounds are provided together with VIP scores (measure of variables importance in the OPLS model) and LogFC values.

Subclass VIP marker VIP score LogFC (HAE-3 vs HAE-1) LogFC (HAE-3 vs HAE-2) LogFC (HAE-1 vs HAE-2) Anthocyanins Pelargonidin 1.22 ± 0.26 −14.66 −12.50 2.16 Delphinidin 3-O-sambubioside 1.19 ± 0.19 −16.07 −14.86 1.23 Peonidin 3-O-(6″-acetyl-glucoside) 1.17 ± 0.23 −21.84 −22.47 0.62 Pelargonidin 3,5-O-diglucoside 1.16 ± 0.27 22.26 22.89 0.62 Pelargonidin 3-O-glucosyl-rutinoside 1.15 ± 0.42 20.93 21.56 0.63 Petunidin 3-O-(6″-p-coumaroyl-glucoside) 1.12 ± 0.65 −17.74 1.87 19.62 Cyanidin 3-O-xyloside 1.11 ± 0.27 −18.87 −18.92 −0.04 Delphinidin 3-O-glucoside 1.11 ± 0.16 −5.25 −22.04 −16.79 Cyanidin 3-O-(6″-malonyl-glucoside) 1.10 ± 0.17 −16.57 −19.38 −2.81 Delphinidin 3-O-(6″-acetyl-galactoside) 1.10 ± 0.59 22.15 22.78 0.62 Malvidin 3-O-glucoside 1.10 ± 0.19 −19.13 −21.29 −2.16 Dihydrochalcones Phloretin 2′-O-xylosyl-glucoside 1.21 ± 0.28 −23.89 −20.64 3.25 Dihydroflavonols Dihydromyricetin 3-O-rhamnoside 1.11 ± 0.15 1.43 −2.21 −3.64

Dihydroquercetin 1.10 ± 0.27 1.25 −19.66 −20.91 Flavanols Procyanidin dimer B7 1.18 ± 0.43 −15.99 1.87 17.87

(+)-Gallocatechin 1.10 ± 0.71 −22.61 1.88 24.49 Flavanones Pinocembrin 1.14 ± 0.33 12.01 22.06 10.04 6-Geranylnaringenin 1.13 ± 0.44 24.01 17.83 −6.18 Eriodictyol 1.12 ± 0.62 −23.27 −20.09 3.17 Naringin 1.10 ± 0.38 −19.18 −18.58 0.59 Eriocitrin 1.10 ± 0.26 1.25 −19.67 −20.92 Flavones Luteolin 1.19 ± 0.37 −20.51 1.87 22.39 Hispidulin 1.14 ± 0.45 −1.74 0.08 1.83 Sinensetin 1.12 ± 0.66 −2.20 19.87 22.07 Luteolin 7-O-glucoside 1.11 ± 0.15 1.54 −1.95 −3.51 Geraldone 1.11 ± 0.19 19.59 −2.09 −21.68 Luteolin 7-O-malonyl-glucoside 1.11 ± 0.16 −16.52 −19.81 −3.28 Tangeretin 1.10 ± 0.59 −2.20 −0.89 1.30 Flavonols Kaempferol 7-O-glucoside 1.19 ± 0.22 −17.54 −16.19 1.34 Quercetin 3-O-xylosyl-rutinoside 1.15 ± 0.36 19.07 19.70 0.62 Quercetin 3-O-rhamnoside 1.11 ± 0.15 1.54 −1.95 −3.50 Quercetin 3-O-glucosyl-xyloside 1.11 ± 0.16 1.25 −18.32 −19.57 Methylgalangin 1.11 ± 0.19 19.59 −2.09 −21.68 Isoflavonoids 6″-O-Malonylgenistin 1.11 ± 0.16 −11.83 −21.42 −9.59 Lignans Secoisolariciresinol-sesquilignan 1.17 ± 0.23 20.67 21.30 0.62 7-Hydroxysecoisolariciresinol 1.13 ± 0.61 −22.25 1.87 24.13 7-Oxomatairesinol 1.10 ± 0.59 −2.20 −0.89 1.30 Alkylphenols 4-Vinylphenol 1.13 ± 0.24 −7.40 2.92 10.33 Curcuminoids Curcumin 1.17 ± 0.41 −23.10 −13.16 9.93 Demethoxycurcumin 1.11 ± 0.15 22.73 −2.31 −25.05 Furanocoumarins Bergapten 1.10 ± 0.46 3.76 5.44 1.68 Hydroxycinnamaldehydes Ferulaldehyde 1.10 ± 0.56 3.36 1.10 −2.34 Hydroxycoumarins Coumarin 1.16 ± 0.49 −19.26 1.87 21.14 Mellein 1.10 ± 0.56 3.36 1.02 −2.34 Hydroxyphenylpropenes Estragole 1.17 ± 0.48 −22.84 −19.73 3.11 Acetyl eugenol 1.10 ± 0.26 1.25 −19.37 −20.62 Naphtoquinones 1,4-Naphtoquinone 1.17 ± 0.21 5.80 7.09 1.29 Tyrosols Hydroxytyrosol 1.18 ± 0.41 −6.83 14.28 21.12 3,4-DHPEA-AC 1.15 ± 0.47 −13.51 −11.02 2.49 Hydroxybenzoic acids Gallic acid 4-O-glucoside 1.21 ± 0.30 −23.39 1.88 25.27

Gallic acid 1.20 ± 0.31 −9.19 −0.08 9.10 3-Hydroxybenzoic acid 1.17 ± 0.23 22.64 23.27 0.62 Syringic acid 1.16 ± 0.28 25.64 26.26 0.62 Hydroxycinnamic acids 1,2-Diferuloylgentiobiose 1.21 ± 0.30 −18.79 1.87 20.67

p-Coumaric acid ethyl ester 1.17 ± 0.23 19.70 13.54 −6.17 24-Methyllathosterol ferulate 1.17 ± 0.34 −20.86 −5.14 15.72 Verbascoside 1.17 ± 0.24 23.26 23.89 0.63 p-Coumaroyl glycolic acid 1.17 ± 0.25 21.92 22.55 0.63 Sitosterol ferulate 1.15 ± 0.38 −9.43 0.04 9.48 3/4/5-Feruloylquinic acid 1.14 ± 0.63 −6.66 −3.15 3.51 1-Sinapoyl-2-feruloylgentiobiose 1.11 ± 0.15 1.25 −20.36 −21.61 2-S-Glutathionyl caftaric acid 1.11 ± 0.17 1.25 −19.34 −20.59 3/4/5-Sinapoylquinic acid 1.11 ± 0.19 1.25 −19.95 −21.20 Chicoric acid 1.11 ± 0.29 −18.35 −18.62 −0.27 1,2-Disinapoylgentiobiose 1.10 ± 0.20 1.25 −16.55 −17.80 Hydroxyphenylacetic acids Homoveratric acid 1.16 ± 0.47 −13.51 −11.02 2.49 Stilbenes Pterostilbene 1.21 ± 0.18 24.78 3.09 −21.68

with HAE-1, namely pterostilbene and resveratrol 3-O-glucoside (being LogFC = 24.8 and 19.7, respectively), whilst HAE-1 was the best source of gallic acid 4-O-glucoside (LogFC 25.3) when compared with HAE-2 leaf extract. Therefore, although HAE-1 was the best condition allowing the highest recovery of total phenolics (i.e. 31.84 mg/g DM), in this work we confirmed that each extraction solvent promoted the recovery of exclusive phenolic compounds with a different efficiency (Table 2). 3.2. Antimicrobial activity of M. oleifera extracts

Antimicrobial activity tests showed different ZOI of bacterial growth varying between extracts of M. oleifera leaves when considering Gram-positive and Gram-negative bacteria. The results obtained are reported inTable 3. HAE-2 leaf extracts exhibited an inhibitory activity against B. cereus and L. innocua when compared with both ampicillin 1 mg/mL solution ZOI (9 mm) and the other treatments (i.e. HAE-1 and HAE-3). Overall, ZOI of HAE-1 and HAE-2 leaf extracts were wider than the positive control. In fact, the two previous treatments exhibited a great potential in the control of contaminant/pathogenic Gram-positive bacteria. Conversely, the concentrated M. oleifera leaf extracts did not considerably inhibited the growth of any Gram-negative strains tested (Table 3); in fact, ampicillin solution ZOI displayed a larger area (16 mm) than all extracts. All negative controls did not exhibit any ZOI (< 5 mm).

As its typical mode of action, most of the bioactive compounds act altering the permeability and the integrity of the bacterial cell. It is widely established in literature that Gram-negative are more tolerant to plant bioactives than Gram-positive due to the presence of the peri-plasmic membrane (Fernàndez-Pèrez, Tenorio, & Ruiz-Larrea, 2018). In this work, the different antimicrobial activities observed could also be related to the differences in the envelope structures of the tested Fir-micutes and Proteobacteria. In a previous work,Viera, Mourão, Angelo, Costa, and Vieira (2010)stated that M. oleifera extracts were found to be more bactericidal than soursop extracts, considering 13.0 mm halo as efficient against Staphylococcus aureus, Vibrio cholerae and Escherichia coli. As previously reported, many phenolic compounds have been widely characterized and studied for their antimicrobial and anti-oxidant efficiencies, such as flavonoids (e.g. myricetin, quercetin and catechin) (Hendra, Ahmad, Sukari, & Oskoueian, 2011),

hydroxybenzoic and hydroxycinnamic acids (Vongsak et al., 2013). In this regard, a phytochemical screening done byOnyekaba, Chinedu, and Fred (2013) revealed thatflavonoids, terpenoids, phenolics and alkaloids characterizing M. oleifera leaves extracts possessed a marked antibacterial potential against E. coli and Pseudomonas aeruginosa.

3.3. In vitro enzymatic and antioxidant assays

The in vitro antioxidant capacities of Moringa extracts were in-vestigated by different spectrophotometric assays. The results obtained are given inTable 4. In the present work, HAE-2 extract exhibited the strongest ability in both DPPH and ABTS assays, while HAE-3 had the lowest anti-radical abilities. Regarding the reducing power ability, we performed three reducing power assays and the activity can be ranked as HAE-1 > HAE-2 > HAE-3. According to the results (Table 4), the alcoholic extracts of M. oleifera leaves had the best radical scavenging and reducing power abilities, as reported also by Prabakaran, Kim, Sasireka, Chandrasekaran, and Chung (2018) and Nobossé, Fombang, and Mbofung (2018). In terms of metal chelating abilities, the best activity was exhibited by HAE-2 leaf extracts, and the least active was HAE-3. As a result, the observed significant activity for 1 and HAE-2 extracts could be linked to their phytochemical profiles. For example, the level of anthocyanins, lignans, flavonols, phenolic acids in these extracts were higher compared with HAE-3, and these compounds have been widely reported as radical scavenger or reductive agents (Heleno, Martins, Queiroz, & Ferreira, 2015; Khoo, Azlan, Tang, & Lim, 2017; Panche, Diwan, & Chandra, 2016; Yashin et al., 2018).

In recent years, numerous enzymatic inhibitors are used to scruti-nize plants with the aim to successfully identify a plant of significant therapeutically value. For instance, AChE and BChE are known for the treatment of Alzheimer’s disease (Pereira Rocha et al., 2018), tyrosinase against skin disorders (Zolghadri et al., 2019), whilst amylase and glucosidase are normally responsible in the hydrolysis of starch into sugars. Consequently, these key enzymes are vital for the human bio-logical system in preventing the onset of major health issues (Copeland, Harpel, & Tummino, 2007). The enzyme inhibitory effects of M. oleifera extracts were investigated against five enzymes, namely acet-ylcholinesterase (AChE), butyracet-ylcholinesterase (BChE), tyrosinase, α-amylase andα-glucosidase. The results are summarized inTable 5. In the present work, all of the tested extracts exhibited inhibitory effects on AChE and BChE. The higher inhibitory effects on these enzymes were found in HAE-3 and HAE-1. In terms of tyrosinase inhibitory ef-fect, the extracts decreased in the order: HAE-1 > HAE-2 > HAE-3. In this work, the best inhibitory abilities of amylase and glucosidase were observed for HAE-1 and HAE-3 leaf extracts, respectively. In the light of literature, we observed few data on enzyme inhibition potential of M. oleifera leaves. Accordingly, some researchers have reported that M. oleifera leaves had significant enzyme inhibition properties (Khan, Parveen, Chester, Parveen, & Ahmad, 2017; Natsir, Wahab, Laga, & Arif, 2018). Therefore, thesefindings could provide a new horizon on the biological effects of M. oleifera leaves to be potentially exploited by food and pharmaceutical industries.

Table 3

ZOI of bacterial growth as resulted by antimicrobial activity test of M. oleifera leaf extracts (i.e. HAE-1, HAE-2 and HAE-3) when considering Gram-positive and Gram-negative bacteria. Data are presented as mean values ± standard deviation (n = 3). Superscript letters within each column indicate homo-geneous sub-class as resulted from ANOVA (p < 0.05), Duncan's post-hoc test.

ZOI (mm) Bacillus cereus Listeria innocua Salmonella Enteritidis Salmonella Typhimurium HAE-1 11.67 ± 0.47b _{10.21 ± 0.08}c _{5.67 ± 0.47}a _{5.33 ± 0.47}a HAE-2 15 ± 0.00c _{9.10 ± 0.62}b _{5.00 ± 0.00}a _{5.00 ± 0.00}a HAE-3 7.67 ± 0.47a _{5.00 ± 0.00}a _{5.00 ± 0.00}a _{5.00 ± 0.00}a Ampicillin (0.1%) 9.00 ± 0.00b _{9.00 ± 0.00}b _{16.00 ± 0.00}b _{16.00 ± 0.00}b

ZOI: Zone of inhibition. The mean ZOI of 5.00 mm is related to absence of inhibition (standard diameter of wells in agar medium over Petri dishes).

Table 4

In vitro antioxidant activity of the different Moringa leaf extracts. Data (expressed on a dry matter basis, DM) are presented as mean values ± standard deviation (n = 3). Superscript letters within each column indicate homogeneous sub-class as resulted from ANOVA (p < 0.05), Duncan's post-hoc test. TE: Trolox equivalent; EDTAE: EDTA equivalent.

Sample DPPH (mgTE/g) ABTS (mgTE/g) CUPRAC (mgTE/g) FRAP (mgTE/g) Phosphomolybdenum (mmol TE/g) Metal chelating (mg EDTAE/g) HAE-1 45.38 ± 0.66b _{37.17 ± 0.84}b _{136.17 ± 1.33}c _{58.26 ± 1.11}c _{1.41 ± 0.10}c _{14.30 ± 2.23}b

HAE-2 49.55 ± 0.55c _{45.26 ± 0.68}c _{99.29 ± 2.13}b _{54.24 ± 0.54}b _{0.81 ± 0.12}b _{54.78 ± 1.57}c

3.4. Correlations

Pearson’s correlation coefficients were inspected in order to provide potential correlations among the different parameters under in-vestigation (supplementary material). Flavonols alkylphenols, lignans, tyrosols and phenolic acids were the classes of compounds mostly re-lated to the in vitro antioxidant activities. In addition, alkylphenols presented strong correlations (p < 0.01) withfive in vitro antioxidant assays, being DPPH (0.88), ABTS (0.85), phosphomolybdenum (0.94), CUPRAC (0.99) and FRAP (0.96). Besides,flavonol equivalents were the phenolic subclasses most correlated (p < 0.01) with DPPH (0.89) and ABTS (0.92). Therefore, it was possible to observe a distinct cor-relation between the in vitro antioxidant activities and the phenolic content, which indicates that these compounds are directly involved in such activities. However, no correlation was observed between the antioxidant activities and theflavonoid contents, which demonstrated that the presence of a considerable amount of these compounds in Moringa leaf extracts does not always imply a corresponding high an-tioxidant potential.

Generally, the observed enzyme inhibitor effects were dissimilar to antioxidant assays. For example, HAE-3 possessed the lowest anti-oxidant ability however the best inhibitory effect on AChE was pro-vided by that same extract. Besides, cholinesterase inhibitory effects were also not correlated with the phytochemicals, presenting weak correlation for flavones when considering AChE and BChE activities. Similar to cholinesterase, amylase and glucosidase inhibitory effects were found to be not related to polyphenols. However, tyrosinase in-hibitory effects were strongly correlated with alkylphenols (0.84) and tyrosol (0.66). Thesefindings were supported byWen et al. (2013), who reported some tyrosols as anti-melanogenic agents. Taken together, the observed enzyme inhibitory effects could be explained with the com-plex nature of phytochemicals or their possible interactions (i.e. sy-nergistic and antagonistic).

Afterwards, we found thatflavonols, lignans and alkylphenols were significantly (p < 0.01) related to the ZOI of L. innocua, thus leading to attribute the bactericidal potential to these bioactive compounds. Similar results were observed in ZOI of B. cereus, showing significative correlations (p < 0.05) between the amounts of tyrosols (0.71), phe-nolic acids (0.78) and anthocyanins (0.72). Besides, looking to the re-sults on Gram-negative, it is possible to hypothesize that the amount of lignans (p < 0.05) may impact the absence of inhibition on S. Typhimurium antimicrobial tests, as previously observed byKawaguchi et al. (2009), showing that 89% of lignans stereoisomers tested against Salmonella choleraesuis were characterized by no bactericidal activities. 4. Conclusions

In this work, an untargeted metabolomic approach based on UHPLC-QTOF-mass spectrometry was used to depict the phenolic pro-file of three different Moringa leaf extracts obtained by using methanol, methanol:water 50:50 v/v and ethyl acetate, revealing a great abun-dance offlavonoids and phenolic acids. Afterwards, the following use of both unsupervised and supervised multivariate statistics applied to the metabolomic dataset depicted the phenolic classes being mostly af-fected by the different extractions tested. Accordingly, each extraction

solvent was found to promote a different extraction efficiency, re-cording the highest total phenolic content in methanol 100% leaf ex-tracts (i.e., (31.84 mg/g). In addition, enzymatic, antioxidant and an-timicrobial activities were assessed. Methanol 100% and methanol:water 50:50 v/v extracts exhibited an expressive activity against all Gram-positive bacteria tested (i.e., Bacillus cereus and Listeria innocua). Furthermore, the in vitro antioxidant assays revealed that methanol:water 50:50 v/v extracts showed the highest DPPH and ABTS values, being 49.55 and 45.26 mgTE/g, respectively, whilst methanol 100% were characterized by the highest values for CUPRAC and FRAP activities (i.e., 136.17 and 58.26 mgTE/g) and when considering phosphomolybdenum assay (1.41 mmolTE/g). Interestingly, ethyl-acetate leaf extracts showed the higher inhibitory effects on AChE and glucosidase (being 4.68 mg GALAE/g and 24.51 mmol ACAE/g, re-spectively) when compared with the other extracts. In addition, strong correlations (p < 0.05; p < 0.01) among each phenolic class and the various biological activities were also recorded. Therefore, it is possible to conclude that Moringa leaf extracts could represent an interesting food matrix able to be used as food supplement and preservative. However, the selection of the extraction solvent to recover specific phenolic classes should be carefully evaluated, considering that the ethyl-acetate extracts, i.e., those characterized by the lower total phe-nolic contents, were also very effective as enzymatic inhibitors. Declaration of Competing Interest

The authors declare no conflict of interest. Acknowledgments

The authors wish to thank the “Romeo and Enrica Invernizzi” foundation for supporting the metabolomic platform and the Sud Rienergy Soc. Agr. S.r.l. - Favella Group (Corigliano Calabro, CS, Italy) for providing the botanical material.

Appendix A. Supplementary material

Supplementary data to this article can be found online athttps:// doi.org/10.1016/j.foodres.2019.108712.

References

Abraham, S. I., Abu, J. O., & Gernah, D. I. (2013). Effect of Moringa oleifera leaf powder supplementation on some quality characteristics of wheat bread. Food and Nutrition Sciences, 270–275 443036.

Adisakwattana, S., & Chanathong, B. (2011). Alpha-glucosidase inhibitory activity and lipid-lowering mechanisms of Moringa oleifera leaf extract. European Review for Medical and Pharmacological Sciences, 15(7), 803–808.

Blasi, F., Urbani, E., Simonetti, M. S., Chiesi, C., & Cossignani, L. (2016). Seasonal var-iations in antioxidant compounds of Olea europaea leaves collected from different Italian cultivars. Journal of Applied Botany and Food Quality, 89, 202–207.

Castro-López, C., Ventura-Sobrevilla, J. M., González-Hernández, M. D., Rojas, R., Ascacio-Valdés, J. A., Aguilar, C. N., & Martínez-Ávila, G. C. C. (2017). Impact of extraction techniques on antioxidant capacities and phytochemical composition of polyphenol-rich extracts. Food Chemistry, 237, 1139–1148.

Chiocchio, I., Mandrone, M., Sanna, C., Maxia, A., Tacchini, M., & Poli, F. (2018). Screening of a hundred plant extracts as tyrosinase and elastase inhibitors, two en-zymatic targets of cosmetic interest. Industrial Crops and Products, 122, 498–505.

Table 5

Enzyme inhibitory activity of the different Moringa leaf extracts. Data (expressed on a dry matter basis, DM) are presented as mean values ± standard deviation (n = 3). Superscript letters within each column indicate homogeneous sub-class as resulted from ANOVA (p < 0.05), Duncan's post-hoc test. GALAE: Galatamine equivalent; KAE: Kojic acid equivalent; ACAE: Acarbose equivalent.

Sample AChE inhibition (mg GALAE/ g)

BChE inhibition (mg GALAE/ g)

Tyrosinase inhibition (mg KAE/ g)

Amylase inhibition (mmol ACAE/g)

Glucosidase inhibition (mmol ACAE/g)

HAE-1 4.19 ± 0.31b _{5.26 ± 0.21}b _{112.39 ± 1.37}b _{0.62 ± 0.00}b _{18.92 ± 1.30}b

HAE-2 3.26 ± 0.15a _{2.23 ± 0.54}a _{111.80 ± 0.50}b _{0.50 ± 0.02}a _{12.97 ± 2.43}a

Copeland, R. A., Harpel, M. R., & Tummino, P. J. (2007). Targeting enzyme inhibitors in drug discovery. Expert Opinion on Therapeutic Targets, 11(7), 967–978.

Djande, C. Y. H., Piater, L. A., Steenkamp, P. A., Madala, N. E., & Dubery, I. A. (2018). Differential extraction of phytochemicals from the multipurpose tree, Moringa olei-fera, using green extraction solvents. South African Journal of Botany, 115, 81–89.

Falowo, A. B., Mukumbo, F. E., Idamokoro, E. M., Lorenzo, J. M., Afolayan, A., & Muchenje, V. (2018). Multi-functional application of Moringa oleifera Lam. in nutri-tion and animal food products: A review. Food Research Internanutri-tional, 106, 317–334. Fernàndez-Pèrez, R., Tenorio, C., & Ruiz-Larrea, F. (2018). Effect of wine polyphenol

extracts on the growth of Escherichia coli. In: Exploring microorganisms: Recent advances in applied microbiology (pp. 226–230). Irvine, California & Boca Raton, Florida, USA: Brown-Walker Press, Universal Publisher, Inc.

Granato, D., Shahidi, F., Wrolstad, R., Kilmartin, P., Melton, L. D., & Finglas, P. (2018). Antioxidant activity, total phenolics andflavonoids contents: Should we ban in vitro screening methods? Food Chemistry, 264, 471–475.

Hassan, F. A. M., Bayoumi, H. M., Abd El-Gawad, M. A. M., Enab, A. K., & Youssef, Y. B. (2016). Utilization of Moringa oleifera leaves powder in production of yoghurt. Journal of Dairy Science, 11, 69–74.

Heleno, S. A., Martins, A., Queiroz, M. J. R., & Ferreira, I. C. (2015). Bioactivity of phenolic acids: Metabolites versus parent compounds: A review. Food Chemistry, 173, 501–513.

Hendra, R., Ahmad, S. A., Sukari, M. Y., & Oskoueian, S. E. (2011). Flavonoid analyses and antimicrobial activity of various parts of Phaleria macrocarpa (Scheff.) Boerl fruit. International Journal of Molecular Sciences, 12, 3422–3431.

Kawaguchi, Y., Yamauchi, S., Masuda, K., Nishiwaki, H., Akiyama, K., Maruyama, M., ... Koba, Y. (2009). Antimicrobial activity of stereoisomers of butane-type lignans. Bioscience, Biotechnology, and Biochemistry, 73(8), 1806–1810.

Khan, W., Parveen, R., Chester, K., Parveen, S., & Ahmad, S. (2017). Hypoglycemic po-tential of aqueous extract of Moringa oleifera leaf and in vivo GC-MS metabolomics. Frontiers in Pharmacology, 8, 577.

Khoo, H. E., Azlan, A., Tang, S. T., & Lim, S. M. (2017). Anthocyanidins and anthocyanins: Colored pigments as food, pharmaceutical ingredients, and the potential health benefits. Food & Nutrition Research, 61(1), 1361779.

Leone, A., Fiorillo, G., Criscuoli, F., Ravasenghi, S., Santagostini, L., Fico, G., ... Bertoli, S. (2015). Nutritional characterization and phenolic profiling of Moringa oleifera leaves grown in Chad, Sahrawi Refugee Camps, and Haiti. International Journal of Molecular Sciences, 16(8), 18923–18937.

Lin, H., Zhu, H., Tan, J., Wang, H., Wang, Z., Li, P., & Zhao, C. (2019). Comparative analysis of chemical constituents of Moringa oleifera leaves from China and India by ultra-performance liquid chromatography coupled with quadrupole-time-of-flight mass spectrometry. Molecules, 24, 942.

Lombardi, G., Cossignani, L., Giua, L., Simonetti, M. S., Maurizi, A., Burini, G., ... Blasi, F. (2017). Phenol composition and antioxidant capacity of red wines produced in Central Italy changes after one-year storage. Journal of Applied Botany and Food Quality, 90, 197–204.

Makita, C., Chimuka, L., Steenkamp, P., Cukrowska, E., & Madala, E. (2016). Comparative analyses offlavonoid content in Moringa oleifera and Moringa ovalifolia with the aid of UHPLC-qTOF-MSfingerprinting. South African Journal of Botany, 105, 116–122.

Montesano, D., Cossignani, L., & Blasi, F. (2019). Sustainable Crops for Food Security: Moringa (Moringa oleifera Lam.). Encyclopedia of Food Security and Sustainability, 1, 409–415.

Natsir, H., Wahab, A. W., Laga, A., & Arif, A. R. (2018). Inhibitory activities of Moringa oleifera leaf extract againstα-glucosidase enzyme in vitro. Journal of Physics: Conference Series (Vol. 979, No. 1, p. 012019). IOP Publishing.

Nobossé, P., Fombang, E. N., & Mbofung, C. M. (2018). Effects of age and extraction solvent on phytochemical content and antioxidant activity of fresh Moringa oleifera L. leaves. Food Science & Nutrition, 6(8), 2188–2198.

Olabode, Z., Akanbi, C. T., Olunlade, B., & Adeola, A. A. (2015). Effects of drying tem-perature on the nutrients of Moringa (Moringa oleifera) leaves and sensory attributes of dried leaves infusion. Direct Research Journal of Agriculture and Food Science, 3, 117–122.

Onyekaba, T. U., Chinedu, O. G., & Fred, A. C. (2013). Phytochemical screening and investigations of antibacterial activities of various fractions of the ethanol leaves extract of Moringa oleifera Lam (Moringaceae). International Journal of Pharmaceutical, Chemical & Biological Sciences, 3.

Panche, A. N., Diwan, A. D., & Chandra, S. R. (2016). Flavonoids: An overview. Journal of Nutritional Science, 5.

Pereira Rocha, M., Rodrigues Valadares Campana, P., de Oliveira Scoaris, D., de Almeida,

V. L., Dias Lopes, J. C., Fonseca Silva, A., ... Gontijo Silva, C. (2018). Biological ac-tivities of extracts from Aspidosperma subincanum Mart. and in silico prediction for inhibition of acetylcholinesterase. Phytotherapy Research, 32(10), 2021–2033.

Prabakaran, M., Kim, S. H., Sasireka, A., Chandrasekaran, M., & Chung, I. M. (2018). Polyphenol composition and antimicrobial activity of various solvent extracts from different plant parts of Moringa oleifera. Food Bioscience, 26, 23–29.

Putnik, P., Lorenzo, J. M., Barba, F. J., Roohinejad, S., Jambrak, A. R., Granato, D., ... Bursać Kovačević, D. (2018). Novel food processing and extraction technologies of high-added value compounds from plant materials. Foods, 7(7), 106–122. Rani, N. Z. A., Husain, K., & Kumolosasi, E. (2018). Moringa genus: A recview of

phy-tochemistry and pharmacology. Frontiers in Pharmacology, 9, 108.https://doi.org/10. 3389/fphar.2018.00108.

Rocchetti, G., Lucini, L., Chiodelli, G., Giuberti, G., Montesano, D., Masoero, F., & Trevisan, M. (2017). Impact of boiling on free and bound phenolic profile and anti-oxidante activity of comercial gluten-free pasta. Food Research International, 100, 69–77.

Rocchetti, G., Blasi, F., Montesano, D., Ghisoni, S., Marcotullio, M. C., Sabatini, S., ... Lucini, L. (2019). Impact of conventional/non-conventional extraction methods on the untargeted phenolic profile of Moringa oleifera leaves. Food Research International, 115, 319–327.

Rocchetti, G., Lucini, L., Rodriguez, J. M. L., Barba, F. J., & Giuberti, G. (2019). Gluten-freeflours from cereals, pseudocereals and legumes: Phenolic fingerprints and in vitro antioxidant properties. Food Chemistry, 271, 157–164.

Salek, R. M., Steinbeck, C., Viant, M. R., Goodacre, R., & Dunn, W. B. (2013). The role of reporting standards for metabolite annotation and identification in metabolomic studies. Gigascience, 2, 13.

Uriarte-Pueyo, I., & Calvo, M. I. (2011). Flavonoids as acetylcholinesterase inhibitors. Current Medicinal Chemistry, 18, 5289–5302.

Uysal, S., Zengin, G., Locatelli, M., Bahadori, M. B., Mocan, A., Bellagamba, G., ... Aktumsek, A. (2017). Cytotoxic and enzyme inhibitory potential of two Potentilla species (P. speciosa L. and P. reptans Willd.) and their chemical composition. Frontiers in Pharmacology, 8, 290.

Valenga, M. G. P., Boschen, N. L., Rodrigues, P. R. P., & Maia, G. A. R. (2019). Agro-industrial waste and Moringa oleifera leaves as antioxidants for biodiesel. Industrial Crops & Products, 128, 331–337.

Valgas, C., Souza, S. M. D., Smânia, E. F., & Smânia, A., Jr (2007). Screening methods to determine antibacterial activity of natural products. Brazilian Journal of Microbiology, 38(2), 369–380.

Vergara-Jimenez, M., Almatrafi, M. M., & Fernandez, M. L. (2017). Bioactive components in Moringa oleifera leaves protect against chronic disease. Antioxidants, 6(4), 91.

Viera, G. H., Mourão, J. A., Angelo, A. M., Costa, R. A., & Vieira, R. H. (2010). Antibacterial effect (in vitro) of Moringa oleifera and Annona muricata against Gram positive and Gram negative bacteria. Journal of the São Paulo Institute of Tropical Medicine, 52(3), 129–132.

Vongsak, B., Sithisarn, P., Mangmool, S., Thongpraditchote, S., Wongkrajang, Y., & Gritsanapan, W. (2013). Maximizing total phenolics, totalflavonoids contents and antioxidant activity of Moringa oleifera leaf extract by the appropriate extraction method. Industrial Crops and Products, 44, 566–571.

Wang, L., Chen, X., & Wu, A. (2016). Mini review on antimicrobial activity and bioactive compounds of Moringa oleifera. Medicinal Chemistry, 6, 578–582.

Wen, K. C., Chang, C. S., Chien, Y. C., Wang, H. W., Wu, W. C., Wu, C. S., & Chiang, H. M. (2013). Tyrosol and its analogues inhibit alpha-melanocyte-stimulating hormone induced melanogenesis. International Journal of Molecular Sciences, 14(12), 23420–23440.

Yashin, A. Y., Yashunskii, D. B., Vedenin, A. N., Nifant’ev, N. E., Nemzer, B. V., & Yashin, Y. I. (2018). Chromatographic determination of lignans (antioxidants) in food pro-ducts. Journal of Analytical Chemistry, 73(5), 399–406.

Zhu, Z., Li, S., He, J., Montesano, D., & Barba, F. J. (2018). Enzyme-assisted extraction of polyphenol from edible lotus (Nelumbo nucifera) rhizome knot: Ultra-filtration per-formance and HPLC-MS2 profile. Food Research International, 111, 291–298.

Zolghadri, S., Bahrami, A., Hassan Khan, M. T., Munoz-Munoz, J., Garcia-Molina, F., Garcia-Canovas, F., & Saboury, A. A. (2019). A comprehensive review on tyrosinase inhibitors. Journal of Enzyme Inhibition and Medicinal Chemistry, 34(1), 279–309. Zullaikah, S., Naulina, R. Y., Meinawati, P., Fauziyah, K., Rachimoellah, M., Rachmaniah,

O., ... Prasetyo, E. N. (2019). Enhanced extraction of phenolic compounds from Moringa oleifera leaves using subcritical water ethanol mixture. IOP Conference Series Materials Science and Engineering, 543, 012021.https://doi.org/10.1088/1757-899X/ 543/1/012021.