Preliminary Phytochemical Screening and Antioxidant Activity of Commercial Moringa oleifera Food Supplements

Moringa oleifera has been reported to possess a high number of bioactive compounds; hence, several food supplements are commercially available based on it. This work aimed to analyze the phytochemical composition and antioxidant activity of commercial food supplements. The phenolic composition of methanolic extracts was determined by using high-performance liquid chromatography with diode-array and electrospray ionization mass spectrometric detection (HPLC-DAD-ESI-MSn), and the antioxidant activity was assessed by ABTS·+ and DPPH assays. Thirty-three compounds were identified, and all the main compounds were quantified, observing that the main contribution to the phenolic profile was due to kaempferol and quercetin glucosides. The antioxidant activity in both assays agreed with the phenolic content: the higher the phenolic levels, the higher the antioxidant activity. The obtained results were compared with those previously published regarding Moringa oleifera leaves to establish the potential benefits of food supplement consumption in the diet.


Introduction
Moringa is a plant cultivated in different countries such as India, Ethiopia, the Philippines, and Sudan, and is being grown in West, East, and South Africa, tropical Asia, Latin America, the Caribbean, and the Pacific Islands. It is also known in the world as "the tree of life" because it has various parts which are used as sources of food and medicines [1]. There are 13 species of this plant, which encompass a very diverse range of growth habits or forms, from herbs and shrubs to large trees. Although they vary greatly in their form, it is very easy to distinguish a member of Moringa from any other plant. Large pinnate leaves characterize these species, where each leaf is divided into many leaflets. The fruits form a long and woody capsule that, when it reaches maturity, slowly opens into three valves that separate one from the other along their length, remaining attached only to the base of the fruit [2].
Of the species discussed above, Moringa oleifera, is the best known and most used. It is not very long-lived, about 20 years, and reaches a height of between 5-10 m. This species is native to South Asia, where it grows in the Himalayan foothills, but is widely cultivated across the tropics. Numerous studies have highlighted the advantageous influences of this plant on human health [3], which is cultivated for its edible leaves, flowers, and nutritious pods, with M. oleifera leaf being the most utilized part [4]. In recent years, M. oleifera leaves have been extensively studied due to their enormous potential as sources of functional compounds with health-promoting properties [5], especially various biological activities such as antioxidant, anti-inflammatory, anti-diabetic, anti-cancer, cardioprotective, hypocholesterolemic, hepatoprotective, antifungal, antiviral, antidepressant, and anti-asthmatic activities [6][7][8]. In addition, M. oleifera leaves are useful in treating neurodysfunctional diseases such as Alzheimer's disease, epilepsy, and ischemic stroke [9,10]. Before performing the sample extraction, the content of 10 capsules was mixed, and 10 tablets were ground and mixed to ascertain representativity. Then, three sub-samples of each supplement were extracted and analyzed independently. Ultrasound-assisted extraction was done by placing 2.5 g of dry material in 50 mL MeOH for 10 min (Qsonica Sonicators; Newton, CT, USA) with a power of 55 W and a frequency of 20 kHz (50% power). Each sample was extracted in triplicate. Then, solutions were filtered through Whatman No.1 filters and the solvent was evaporated under reduced pressure in a rotary evaporator at 40 • C. Dried extracts (DE) were stored at −20 • C until analysis.

Chromatographic Analysis
The instrumentation and the chromatographic conditions are described in detail in the Supplementary Materials. Briefly, an HPLC system was connected to a DAD detector and an ion trap mass spectrometer equipped with an electrospray ionization interface, operating in negative ion mode.
MS data and analytical standards were used for compounds' identification, whereas the quantitation was performed using UV data to construct the calibration graphs. Calibration graphs for chlorogenic acid, neochlorogenic acid, coumaric acid, quercetin, kaempferol, rutin, and vicenin-2 were prepared at concentrations 0.5-100 mg L −1 in MeOH. Chromatograms were recorded at 320 nm for phenolic acids and 350 nm for flavonoids. The mentioned analytical standards were used to quantify the exact compound or compounds of the same chemical family. A chromatogram showing the analytical standards used is given in Figure S1 (Supplementary Materials).

Antioxidant Capacity Assays
The antioxidant capacity of the selected food supplements was studied by ABTS ·+ and DPPH assays. The results were expressed in mg Trolox equivalents per 100 g of dried extract (mg TE/g DE), mmol TE/g DE, and IC50 (50% inhibition). Details for each assay are given in Supplementary Materials.

Statistical Analysis
Statistical analysis was carried out using SPSS Statistics software v.22 (IBM SPSS Statistics for Windows, IBM Corp., Armonk, NY, USA). The analyses were performed in triplicate, and data are expressed as mean ± standard deviation. A one-way analysis of variance (ANOVA) with Tukey's HSD post-hoc test (p < 0.05) was used to look for statistical differences among results in the quantification of compounds and antioxidant activities. Different superscripts in the corresponding tables indicate significant differences in the extracts (p < 0.05).

Results and Discussion
In this work, we selected food supplements containing M. oleifera leaves and extracts of M. oleifera seeds. The phenolic profile was characterized by HPLC-DAD-ESI-MS n , and the main compounds were quantified. Then, the antioxidant capacity was evaluated by ABTS ·+ and DPPH assays.

HPLC-ESI-MS n Analysis of Food Supplements' Extracts
The characterization of the extracted compounds was performed by mass spectrometry, using negative ion mode (the most sensitive mode for phenolic compounds). The identification was carried out using analytical standards and data available in the scientific literature. Compounds were numbered regarding their order of elution, keeping the same numbering in all samples ( Table 2). The base peak chromatogram of a food supplement is shown in Figure 1. As can be seen in Table 2, most of the characterized compounds were flavonoid glycosides, 19 out of 33 identified compounds. The phenolic profile agrees with previous reports on the composition of M. oleifera leaves [15,25]. Following is a brief description of the identification.
Antioxidants 2023, 12, x FOR PEER REVIEW 4 of 1 triplicate, and data are expressed as mean ± standard deviation. A one-way analysis o variance (ANOVA) with Tukey's HSD post-hoc test (p < 0.05) was used to look for statis tical differences among results in the quantification of compounds and antioxidant activ ities. Different superscripts in the corresponding tables indicate significant differences in the extracts (p < 0.05).

Results and Discussion
In this work, we selected food supplements containing M. oleifera leaves and extract of M. oleifera seeds. The phenolic profile was characterized by HPLC-DAD-ESI-MS n , and the main compounds were quantified. Then, the antioxidant capacity was evaluated by ABTS ·+ and DPPH assays.

HPLC-ESI-MS n Analysis of Food Supplements' Extracts
The characterization of the extracted compounds was performed by mass spectrom etry, using negative ion mode (the most sensitive mode for phenolic compounds). The identification was carried out using analytical standards and data available in the scien tific literature. Compounds were numbered regarding their order of elution, keeping th same numbering in all samples ( Table 2). The base peak chromatogram of a food supple ment is shown in Figure 1. As can be seen in Table 2, most of the characterized compound were flavonoid glycosides, 19 out of 33 identified compounds. The phenolic profile agree with previous reports on the composition of M. oleifera leaves [15,25]. Following is a brie description of the identification.

Phenolic Acids
Compound 4 exhibited deprotonated molecular ion at m/z 315 and suffered the neutral loss of 162 Da to yield dihydroxybenzoic acid at m/z 153 (comparison with an analytical standard of protocatechuic acid), so it was characterized as its hexoside. Compounds 5 and 11 were identified as neochlorogenic acid and chlorogenic acid by comparison with analytical standards. Compound 6 exhibited the transition 179→135, typical of caffeic acid (checked with a caffeic acid analytical standard), so it was tentatively characterized as a derivative.
Compounds 9, 12, and 14 were identified as 3-p-coumaroylquinic acid, 3-feruloylquinic acid, and 4-p-coumaroylquinic acid, respectively, based on the hierarchical scheme proposed by Clifford et al. [42].Compounds 9, 12, and 14 were identified as 3-p-coumaroylquinic acid, 3-feruloylquinic acid, and 4-p-coumaroylquinic acid, respectively, based on the hierarchical scheme proposed by Clifford et al. [42].  It is worth mentioning that although some authors mentioned gallic acid as one of the main compounds in M. oleifera leaves [43], we did not find this compound in any of the analyzed supplements. This is in line with the findings of other authors, who did not find gallic acid either [25].

Other Compounds
Compound 1 was identified as citric acid by comparison with an analytical standard. Compound 2 was characterized as a disaccharide (probably diglucoside) due to the neutral loss of 162 Da (341→ 179) and the characteristic fragments of hexoside moieties (m/z 179, 161, 143, and 119) [47]. Compound 3 was characterized as the glucosinolate glucomoringin, previously reported in M. oleifera [48]. Compound 5 exhibited deprotonated molecular ion at m/z 315 and suffered the neutral loss of 162 Da to yield dihydroxybenzoic acid at m/z 153, so it was characterized as its hexoside. Compound 13 was tentatively characterized as roseoside (vomifoliolglucoside or drovomifoliol-O-β-D-glucopyranoside) based on bibliographic information [49]. Compound 35 was identified as N-feruloyltyramine [50]. This compound was only detected in food supplement S4, due to the presence of black pepper fruit, which contains this compound [51]. Hence, it was absent in all the supplements that contained only M. oleifera.

Quantification of Phytochemicals
The most abundant compounds were flavonoids, followed by phenolic acids. The following analytical standards were used: chlorogenic acid, coumaric acid, and neochlorogenic acid for phenolic acids; and quercetin, kaempferol, rutin, and vicenin-2 (an apigenin glucoside) for flavonoids. The results are shown in Table 3.
Food supplements S1, S2, and S5 presented more than 10 mg g −1 DE of total individual phenolic content (the sum of all the phenolics quantified by HPLC), with S5 presenting the highest amount of phenolics. However, the other supplements presented a lower concentration of phenolics, with S3 presenting the lowest concentration. Although all of them are made from M. oleifera leaves (except S4), these differences make it clear that the preparation of food supplements is different, as these contents of phenolics are not supposed to be based only on the origin of M. oleifera species. However, in all of them, the profile is similar: more than 85% of the phenolics are flavonoids (again, except in S3, with only 73% of phenolics). Among flavonoids, the main compounds are kaempferol and quercetin glycosides, in agreement with the results reported in M. oleifera leaves by other authors [15,25,53]. Values are reported as mean ± SD of three parallel experiments. Bold values represent the sum of each type of components. Means in the same line not sharing the same letter are significantly different at p < 0.05 probability level. Hex = hexoside (usually glucoside, but also galactoside); der. = derivative; FQA = feruloylquinic acid; AHex = acetylhexoside.
Sultana et al. [53] reported a total amount of flavonoids of 6.13 mg mg −1 , similar to our results (2.4-12.5 mg g −1 DE). These same authors reported concentrations of quercetin and kaempferol of 0.281 and 0.0402 mg g −1 , respectively, whereas we found levels of 1.4-4.7 mg g −1 DE for quercetin (sum of all glycosides) and 0.56-8.2 mg g −1 DE for kaempferol (sum of all glycosides). These differences are due to the high levels of myricetin reported by Sultana et al., whereas we did not find this flavonoid in any of the analyzed extracts.
Singh et al. [43] reported concentrations of 0.08-0.5 mg g −1 for chlorogenic acid, 0.05-0.5 mg g −1 for ferulic acid, 0.07-0.2 mg g −1 for kaempferol and 0.03-0.8 mg g −1 for quercetin. Whereas the levels of chlorogenic acid and ferulic acid are similar to the ones reported in this work (Table 3), the levels found for flavonoids by these authors were much lower, due to the different extractants used (water in their work, in contrast to methanol in ours). Other authors also reported the levels of specific phenolic compounds in M. oleifera [54]; however, the concentrations were given in terms of fresh weight, making the comparison not straightforward. Hence, it can be observed that a comparison in terms of the main compounds can be made (quercetin and kaempferol were the main contributors to the phenolic profile), whereas comparisons of concentration are difficult to perform.
After performing the quantitation of the most abundant compounds, we also calculated the relative contribution of all compounds using the method of area normalization. Peak areas of each compound were obtained using the precursor ion, [M-H] -(Extracted Ion Chromatograms). Then, the relative contribution (in percentage) of each compound was calculated and the heat map (the darker the color, the higher the abundance) was constructed (Table 4). It can be observed that these data agree with the quantification (Table 3), observing that kaempferol and quercetin glycosides represented the highest percentage of phenolic contribution to the extracts.

Antioxidant Activity
The antioxidant capacity was evaluated utilizing the ABTS ·+ and DPPH assays. We expressed the results in g TE (Trolox equivalents) per 100 g DE (Figure 2 and Supplementary Materials, Table S1), mmol TE/g DE (Table 5) and IC50 (amount needed to inhibit 50% of ABTS ·+ or DPPH; Table 6). The reason to express the results of the assays in different ways is to ease comparison with other authors, as there is not consensus to express these assays in the same units.  In general, the antioxidant activity observed was in-line with the phenolic content. In this sense, supplement S5 had the highest activity, S1 and S3 presented similar capacity, and S3 and S6 had the lowest antioxidant capacity. However, there are some discrepancies; S3 and S6 presented the same antioxidant activity (no significant differences), even though S3 had less content of phenolics. This difference may be explained by the diverse antioxidant activity displayed by individual phenolics. In this case, both supplements had the same amount of quercetin-O-hexoside, which probably explains the similar activity. However, in general terms, the highest the phenolic content, the highest the antioxidant effect.  Trihydroxy-octadecenoic acid 1.85 1.71 5.74 6.10 1.87 0.81 Hex = hexoside (usually glucoside, but also galactoside); Rut= rutinoside; dHex = deoxyhexoside (usually rhamnoside, but also furanoside); Glc = glucoside. In general, the antioxidant activity observed was in-line with the phenolic content. In this sense, supplement S5 had the highest activity, S1 and S3 presented similar capacity, and S3 and S6 had the lowest antioxidant capacity. However, there are some discrepancies; S3 and S6 presented the same antioxidant activity (no significant differences), even though S3 had less content of phenolics. This difference may be explained by the diverse antioxidant activity displayed by individual phenolics. In this case, both supplements had the same amount of quercetin-O-hexoside, which probably explains the similar activity. However, in general terms, the highest the phenolic content, the highest the antioxidant effect.  (6) Braham et al. [24] reported DPPH values of 0.53 and 0.56 mmol TE/g for M. oleifera dried leaves, by using 70% and 50% ethanol as extraction solvents, respectively. Oldoni et al. [16] found DPPH values of 0.34 mmol TE/g of extract, obtained with 80% ethanol. Lin et al. [55] and Wu et al. [27] reported values of 0.17-0.47 and 0.07-0.15 mmol TE/g for M. oleifera dried leaves in the DPPH assays, respectively, with different concentrations of ethanol as the extractant and different extraction methodologies.
On the other hand, Lin et al. [55] and Wu et al. [27] reported values of 0.23-0.49 and 0.05-0.07 mmol TE/g for M. oleifera dried leaves in the ABTS ·+ assays, respectively, and Oldoni et al. [16] found a value of 0.93 mmol TE/g of extract in the ABTS ·+ assay.
When comparing our results with those previously reported by other authors, it is necessary to consider that there are differences in the solvent and the methodology used for the extraction, and in the forms of expression of results (DE in our work, in contrast to dried sample weight or extract weight in the previous works). In addition, previous studies revealed the significant influence of seasons and agroclimatic locations on the content of bioactive compounds with antiradical activity in M. oleifera leaves [56]. Therefore, it can be said that the results obtained in the present work in food supplements for the DPPH assay (0.05-0.20 mmol TE/g DE) and ABTS ·+ assay (0.07-0.21 mmol TE/g DE), using methanol for extraction purposes, are of the same order as those previously reported by other authors.
In another work [57], values for IC50 of 1.02 and 1.60 mg mL −1 for ABTS ·+ and DPPH assays were reported in methanol extracts of M. oleifera leaves. In general, these values are better than the ones found in food supplements (Table 6). However, food supplement S5 presented a similar antioxidant activity in the ABTS assay (1.26 mg mL −1 ) and slightly lower in the DPPH assay. These results agreed with the fact that S5 presented the highest phenolic concentration (Table 3).

Conclusions
In this work, we have reported the phenolic composition and antioxidant activity of six food supplements (sold in different presentations) based on M. oleifera, and compared the results obtained with those from other authors who analyzed M. oleifera fresh leaves. We found similarities in terms of phenolic profile: the main compounds were derivatives (mainly glucosides) of quercetin and kaempferol. Interestingly, we found malonyl-hexoside and acetyl-hexoside, which are not common flavonoids (the most abundant ones are usually hexoside, pentoside, deoxyhexoside, and rutinoside). However, in terms of quantitative analysis, although quercetin and kaempferol compounds were the most abundant (in agreement with previous works), the concentrations varied significantly between samples. This was an expected result, as the exact origin of M. oleifera plants (as well as season and agroclimatic conditions) and the preparation procedure, not provided by the different manufacturers, are probably different. Regarding the antioxidant capacity, in general, a good potential was obtained for most of the supplements; also, the results were different among them. However, as expected, there was a correlation between phenolic content and antioxidant activity: the higher the phenolic content, the higher the antioxidant activity. In our opinion, the consumption of these food supplements seems to provide a valuable source of antioxidants to the diet, although it is clear that not all the supplements provide the same amount of phenolics (which is equivalent to the antioxidant benefits).