Antioxidant and Anti-Inflammatory Activities of the Crude Extracts of Moringa oleifera from Kenya and Their Correlations with Flavonoids

Moringa oleifera Lam. (M. oleifera) is commonly distributed and utilized in tropical and sub-tropical areas. There has been a large number of reports on the antioxidant and anti-inflammatory activity of its leaves, but only a few about its seeds and roots. Hence, in this work we aimed to systematically compare the antioxidant and anti-inflammatory activities of the ethanol crude extracts of leaves, seeds, and roots of M. oleifera from Kenya, and further correlate the differential activities with the chemical constituents from these three parts. The antioxidant activities were measured by using three different assays (DPPH (2,2-diphenyl-1-picrylhydrazyl), ABTS (2,2′-azinobis-(3-ethylbenzthiazoline-6-sulfonic acid) and FRAP (Ferric-Reducing Antioxidant Power), respectively). Results showed that the leaf extracts displayed the highest DPPH radical scavenging and FRAP total reducing power activities with IC50 values of 1.02 ± 0.13 mg/mL and 0.99 ± 0.06 mM Fe2+/g, respectively; the leaf and root extracts exhibited potential ABTS radical scavenging activities with the IC50 values of 1.36 ± 0.02 and 1.24 ± 0.03 mg/mL. Meanwhile, the leaf and seed extracts (11.1–100 µg/mL) also exerted obvious anti-inflammatory activities, as indicated by the inhibition of NO production. To further reveal correlations between these differential activities with the chemical constituents in the three organs, the total flavonoids content (TFC) of the three different extracts were evaluated, and the TFC of leaves, seeds and roots were found to be 192.36 ± 2.96, 5.89 ± 0.65 and 106.79 ± 2.12 mg rutin equivalent (RE)/g, respectively. These findings indicated the important impacts of the total flavonoid contents on antioxidant and anti-inflammatory activities. Additionally, we further determined the phytochemical profiles of M. oleifera by HPLC-UV/ESI-MS/MS, and identified most of the chemical constituents of leaves as flavonoids. In summary, the leaves of M. oleifera are a better potential natural source of antioxidants and anti-inflammatory agents, and very promising for development into the health promoting dietary supplements.


Introduction
Moringa oleifera Lam. (M. oleifera) is widely distributed and utilized in tropical and sub-tropical regions of the world, and is mainly native to India and Africa. It is often referred to as drumstick tree or horseradish tree, and also called "miracle tree", or "natural gift", or "mother's best friend", due to the high nutrients in leaves with protein, minerals, and β-carotene [1]. The M. oleifera leaves can be eaten in multiple ways, and can maintain nutrient levels for a long time when stored as dried powder. Therefore, some reginal and international relief organizations are putting a great deal of emphasis on developing the leaves of M. oleifera as a nutritional supplement in some Africa countries [2].
There has been a large amount of reports about the antioxidant and anti-inflammatory activities of M. oleifera leaves, but only a few referred to other parts of M. oleifera. For example, Ndhlala et al. [28] had determined the antioxidant activities of leaf extracts of thirteen M. oleifera cultivars from Thailand, South Africa and United States of America, etc. Siddhuraju et al. [29] had examined the antioxidant activities of M. oleifera leaves grown in three different agroclimatic regions (India, Nicaragua and Niger, respectively). Coppin et al. [5] had compared the anti-inflammatory activities of three samples of M. oleifera leaves from sub-Sahara Africa. In this respect, those previous reports showed the M. oleifera leaves are rich in flavonoids and phenolic compounds with high antioxidant and anti-inflammatory activities, whereas there is no such characterization of M. oleifera collected in Kenya, let alone a systematic comparison of its phytochemicals of three different organs (leaves, seeds and roots) as well as the correlation to their different biological activities.
Hence, the objective of this work was to systematically compare the antioxidant and anti-inflammatory activities of three different organs of M. oleifera from Kenya, and to correlate them with the chemical constituents of the organ itself. To this end, this study firstly measured the antioxidant activities of the three organs (leaves, seeds and roots) of M. oleifera and their total flavonoids content, and further determined their anti-inflammatory activities, and then finally evaluated their phytochemical profiles by LC-MS. We expect that the systematic comparison of three different organs of M. oleifera from Kenya could provide some new findings to explain or explore its traditional use or new development, and meanwhile provide valuable information on M. oleifera as a natural source of antioxidants and anti-inflammatory agents to promote their applications for functional food or medicinal values in the near future.

Plant Materials
The three different fresh parts (young leaves, seeds, and roots) of M. oleifera (6 years old) were collected from a farm in Machakos County (Kenya) with rich black soil and plenty of water in June of 2018. These fresh plant materials were dried in the sun, and pulverized by a disintegrator, then packed in sealed polyethylene bags and stored in a fridge at 4 • C until use. The specimens were authenticated by a botanist (Prof. Guangwan Hu) of Wuhan Botanical Garden of Chinese Academy of Sciences, and were stored in the herbarium of the Key Laboratory of Plant Germplasm Enhancement and Specialty Agriculture with the voucher specimen numbers (No. 2018001-2018003).

Chemicals and Reagents
Rutin was purchased from J&K Scientific Ltd.

Preparation of Sample Extraction
Firstly, 5 g dried powders of three different parts of M. oleifera (leaves, seeds, and roots) were immersed in 40 mL 90% ethanol in a 50 mL centrifuge tube for 3 h, and then extracted in an ultrasonic bath (200 W, 40 KHz) for 30 min. Next, the extracts were centrifuged for 5 min at 5000 rpm, and then the extraction process was repeated twice more as described above. After the extraction, all the supernatants were collected and concentrated to dry residues at 45 • C using a rotary evaporator under vacuum. The dried residues were kept at 4 • C until used for further tests. For further LC-MS analysis, each dried sample was dissolved in the methanol solution, and then filtered with a 0.22 µm membrane before injection into the LC-MS system.

DPPH Assay
The DPPH free radical scavenging activities of the crude extracts from three different organs of M. oleifera were evaluated according to a previous method of Zhu et al. [30] with some modifications. Firstly, 10 µL of the adequate diluted samples (dissolved in methanol) or the positive control solutions (Trolox and Ascorbic acid, 31.25-1000 µM) were mixed with 190 µL of methanol solution of DPPH (0.1 mM) in a 96-well plate. After 30 min of reaction at room temperature in darkness, the absorbance of the reaction mixture was recorded at 517 nm with a multifunctional microplate reader (Tecan Infinite M200 PRO, TECAN, Männedorf, Switzerland). Meanwhile, methanol was used as a blank control in this assay, and all the extracts and controls were tested in triplicate. The percentage of inhibition of DPPH free radical was estimated and computed using the following equation: where OD C is the absorbance value of the blank control, OD S is the absorbance value of the tested sample or positive control, and the IC 50 value was acquired when the DPPH free radicals were inhibited by 50%.

ABTS Assay
The ABTS radical scavenging assay of the three different parts extracts of M. oleifera was conducted using a slightly modified method reported by Zhu et al. [30]. The working ABTS radical cation (ABTS + ) solution was prepared by mixing ABTS solution (7 mM in H 2 O) and potassium persulfate (4.9 mM in H 2 O) in equal quantities, and allowed them to react at room temperature in darkness for twelve to sixteen hours. For this assay, the ABTS + solution was appropriately diluted with methanol to acquire an absorbance value of 0.700 ± 0.100 at 734 nm. Then, the mixture solution consisting of 190 µL of ABTS + solution and 10 µL of appropriately diluted samples, or Vitamin C solution (31.25 to 1000 µM), was incubated in a 96-well plate for 30 min in darkness before reading its absorbance at 734 nm. Methanol was used as blank and Vitamin C as positive control. All of the extracts and controls were tested in triplicate. The IC 50 value of ABTS assay was acquired at different concentrations; the calculation formula was identical to the DPPH radical scavenging method described in Section 2.4.1.

FRAP Assay
The FRAP assay was done in accordance to the previous method of Benzie and Strain [31]. Firstly, the working stock solution (Fe 3+ -TPTZ solution) was composed of 20 mM FeCl 3 ·6H 2 O solution, 10 mM TPTZ solution in 40 mM HCl and 300 mM acetate buffer (3.1 g C 2 H 3 NaO 2 ·3H 2 O and 16 mL C 2 H 4 O 2 , pH 3.6) at a ratio of 1:1:10 (v/v/v), and then stored at 37 • C before use. Thereafter, 10 µL of appropriately diluted samples were added to 30 µL of ultrapure water followed by 260 µL of the fresh Fe 3+ -TPTZ stock solutions in a 96-well plate. After 10 min for incubation at 37 • C, the absorbance of the mixture was acquired at 593 nm by triplicate tests with a multifunctional microplate reader. The calibration curve was established by using FeSO 4 ·7H 2 O (62.5, 125, 250, 500, 1000, and 2000 µM) as the standard, and the FRAP activity was expressed as mM Fe 2+ /g of sample.

Determinations of Total Flavonoids Content (TFC)
The TFC in the dried extracts of three organs of M. oleifera were measured by using the colorimetric method by Zou et al. [32] with minor modifications; rutin was used as the standard. Firstly, 60 µL of adequate diluted sample solution, 360 µL of methanol and 20 µL of a 5% NaNO 2 solution were mixed in a 1.5 mL EP tube and shaken gently for 6 min. Next, 40 µL of 10% AlCl 3 ·6H 2 O was added to stand for another 6 min, and followed by the 120 µL of 4% NaOH solution. After 15 min, the absorbance of reaction mixtures was read at 510 nm with an UV/VIS spectrophotometer (UV-1100, MAPADA, Shanghai, China). Methanol was used as the blank. The TFC of each sample was expressed as mg of rutin equivalents per gram of sample (mg RE/g sample). For each sample, this assay was repeated three times.

Anti-Inflammatory Activities of M. oleifera
The murine macrophage RAW264.7 cells were obtained from the American Type Culture Collection (ATCC), and cultured at 37 • C in the DMEM medium supplemented with the 10% FBS and 1% penicillin-streptomycin in a humidified incubator containing 5% CO 2 . The nitrite production was measured according to the Griess reagent protocol [33]. Briefly, the cells were transferred into the 96-well plate (4-6 × 10 4 cells in each well) and incubated for 24 h. Then, those cells were cultured with serial dilution (11.11, 33.33, 100 µg/mL) of leaves and seeds extracts of M. oleifera for 2 h. Next, those cells were added with the 10 ng/mL of LPS for another 24 h. Then, 100 µL of the culture medium were mixed with the equal volume of the Griess reagent, and after 15 min of reaction, the mixture absorbance was read at 540 nm with a microplate reader. Aspirin (10.0 µg/mL) was used as the positive control.

HPLC-UV/ESI-MS/MS Analysis
A TSQ Quantum Access MAX mass spectrometer coupled with a Thermo Accela 600 series HPLC system (Thermo Fisher Scientific, San Jose, CA, USA) was employed for the LC-MS analysis, which was equipped with both UV detector and ESI source. The separation of M. oleifera extracts was acquired with a Waters Sunfire RP-C18 column (4.6 × 250 mm, 5 µm) at 25 • C. 0.1% (v/v) formic acid in ultrapure water (mobile phase A) and acetonitrile (mobile phase B) were used as mobile phases. The HPLC gradient condition was optimized as follows: 8%-30% B in 0-30 min and 30%-95% B in 30-40 min for extracts of leaves; 5%-95% B in 0-40 min for extracts of seeds; and 5%-30% B in 0-20 min and 30%-95% B in 20-30 min for that of roots; respectively. All the injection volume of samples was 10 µL, the flow rate was selected at 0.8 mL/min. The ESI-MS conditions were set as following: mass range (m/z) was from 150 to 1500, spray voltage was 3 kV, capillary temperature was 350 • C, vaporizer temperature was 300 • C, sheath gas (N 2 ) was 40 psi, auxiliary gas (N 2 ) was 10 psi, and the negative scan mode was applied. Full scan and the data-dependent mode were selected to obtain the mass spectra data. All compounds in the extracts of M. oleifera were tentatively identified by comparing the retention time, parent ion, and mass fragments in references or available standards.

Statistical Analysis
In this work, all tests were done in triplicate and the data were represented as the mean ± standard deviation (SD). The statistical software SPSS 17.0 (IBM Corp., New York, NY, USA) was applied for statistical analysis.

Antioxidant Activity of M. oleifera Leaves, Seeds and Roots
Because of the complexity of natural phytochemicals and their different modes of action, it is inaccurate to assess the overall antioxidant potential only by a single method. Therefore, in this work we used three different methods (DPPH, ABTS, and FRAP assay, respectively) to assess and compare the antioxidant potential of three organs of M. oleifera [34]. The results are shown in Table 1. It was interesting to find that the leaves, seeds and roots of M. oleifera had the IC 50 values of 1.02 ± 0.13, >64, 3.33 ± 0.11 mg/mL with DPPH radical scavenging activities, whereas the ABTS values were 1.36 ± 0.02, 40.35 ± 1.47 and 1.24 ± 0.03 mg/mL, and the FRAP values were 0.99 ± 0.06, 0.02 ± 0.00 and 0.20 ± 0.01 mM Fe 2+ /g, respectively. The results of DPPH and FRAP methods were indicated that the antioxidant activities of M. oleifera leaves were the highest, the antioxidant activities of seeds were the lowest, and the roots were in-between. Meanwhile, the roots of M. oleifera had the highest potential antioxidant activities in ABTS assay. Part of these results was similar with those reported by Atawodi et al. [35], but with different activities due to different species, different organs, or different geographic regions. In their study, the descending order of antioxidant activities was root barks > leaves > stem of M. oleifera in Nigeria. However, they did not test the seeds.

Total Flavonoids Content (TFC)
The TFC of the dried extracts of M. oleifera leaves, seeds and roots was listed in Table 1. As shown in Table 1, the TFC of leaves, seeds and roots were 192.36 ± 2.96, 5.89 ± 0.65 and 106.79 ± 2.12 mg rutin equivalent (RE)/g, respectively. It was obvious to see that the descending order of the TFC of plant parts was: leaf > root > seed (32.66:18.13:1). As known from Table 1, the antioxidant activity of leaves is the highest, followed by the roots and the seeds. These indicated the positive impacts of the total flavonoid contents on antioxidant activities. In a previous study, the descending order of the TFC of the ethanol extracts of M. oleifera in Nigeria was: root > bark > leaf > flower > seed [35]. Compared with this study, the seeds from both Kenya and Nigeria had the lowest TFC, while the TFC of leaves was higher than roots in Kenya and the opposite occurred in Nigeria. The differences may be caused by the differences in climate and soil compositions of the different growing areas [36,37].

Anti-Inflammatory Activity
The anti-inflammatory activity was evaluated by determining the production of the nitric oxide (NO) in RAW 264.7 macrophage cells. As shown in Figure 1, the NO in the leaf and seed extracts treated groups was significantly lower than that of the NC group in a dose-dependent manner, within the concentration range investigated (11.11, 33.33, and 100 µg/mL). The leaf extracts at the concentration 100 µg/mL, gave lower NO production than the positive control aspirin (p = 0.024). In contrast, the seed extracts at the same concentration resulted in higher NO production than aspirin, but without significant difference (p = 0.086). Thus, it appeared that the leaf extracts of M. oleifera had a better anti-inflammatory potential than the seeds.

Anti-Inflammatory Activity
The anti-inflammatory activity was evaluated by determining the production of the nitric oxide (NO) in RAW 264.7 macrophage cells. As shown in Figure 1, the NO in the leaf and seed extracts treated groups was significantly lower than that of the NC group in a dose-dependent manner, within the concentration range investigated (11.11, 33.33, and 100 µg/mL). The leaf extracts at the concentration 100 µg/mL, gave lower NO production than the positive control aspirin (p = 0.024). In contrast, the seed extracts at the same concentration resulted in higher NO production than aspirin, but without significant difference (p = 0.086). Thus, it appeared that the leaf extracts of M. oleifera had a better anti-inflammatory potential than the seeds.

Analysis of M. oleifera Leaves, Seeds and Roots by HPLC-UV/ESI-MS/MS
Although the TFCs of three different parts of M. oleifera were determined and compared in our study, the phytochemical profiles or flavonoids profiles have not yet been revealed or compared so far. To further explore and explain the chemical differences in these three organs, LC-MS/MS analysis was conducted. As a result, the base peak chromatograms (BPC) of the extracts of M. oleifera leaves, seeds and roots with the LC-MS analysis were shown in Figure 2, where the peaks were numbered orderly according to their retention time. As per the analysis, a total of 17 compounds in three organs of the M. oleifera were identified or tentatively characterized in the ESI negative ion mode after comparing with the data from the references or purified standards. Table 2 summarized mass spectra data of all the compounds in extracts, which included saccharides, nitrile glycosides, organic acids, phenolic acids, phenylpropanoids, glucosinolates and flavonoids, and their chemical structures were presented in Figure 3. As shown in Figure 2 and Table 2, there were two shared constituents of peak 3 and peak 8 identified in all the three organs of M. oleifera. However, there were obvious differences in the metabolite profiles of the extracts from leaves, seeds and roots of M. oleifera, which could provide a good clue to explain their different antioxidant power, and the main phytochemicals responsible for it. It is also found the M. oleifera leaves are rich in flavonoids which may contribute largely to their better antioxidant and anti-inflammatory activities, while the seeds are not. In this sense, the present study for the first time conducted the phytochemical comparisons of three different organs of M. oleifera (leaves, seeds and roots) which could have a good correlation to their distinct antioxidant or anti-inflammatory activities, as shown in Table 1 and Figure 1.

Analysis of M. oleifera Leaves, Seeds and Roots by HPLC-UV/ESI-MS/MS
Although the TFCs of three different parts of M. oleifera were determined and compared in our study, the phytochemical profiles or flavonoids profiles have not yet been revealed or compared so far. To further explore and explain the chemical differences in these three organs, LC-MS/MS analysis was conducted. As a result, the base peak chromatograms (BPC) of the extracts of M. oleifera leaves, seeds and roots with the LC-MS analysis were shown in Figure 2, where the peaks were numbered orderly according to their retention time. As per the analysis, a total of 17 compounds in three organs of the M. oleifera were identified or tentatively characterized in the ESI negative ion mode after comparing with the data from the references or purified standards. Table 2 summarized mass spectra data of all the compounds in extracts, which included saccharides, nitrile glycosides, organic acids, phenolic acids, phenylpropanoids, glucosinolates and flavonoids, and their chemical structures were presented in Figure 3. As shown in Figure 2 and Table 2, there were two shared constituents of peak 3 and peak 8 identified in all the three organs of M. oleifera. However, there were obvious differences in the metabolite profiles of the extracts from leaves, seeds and roots of M. oleifera, which could provide a good clue to explain their different antioxidant power, and the main phytochemicals responsible for it. It is also found the M. oleifera leaves are rich in flavonoids which may contribute largely to their better antioxidant and anti-inflammatory activities, while the seeds are not. In this sense, the present study for the first time conducted the phytochemical comparisons of three different organs of M. oleifera (leaves, seeds and roots) which could have a good correlation to their distinct antioxidant or anti-inflammatory activities, as shown in Table 1 and Figure 1.

Saccharides and Nitrile Glycosides
As shown in Table 2

Saccharides and Nitrile Glycosides
As shown in Table 2

Saccharides and Nitrile Glycosides
As shown in Table 2 [40]. Besides, the fragment ion at m/z 116 was generated by the loss of the hexose moiety ([M−H−162] − ). Hence, peak 5 was temporarily proposed as niazirin [41].  indicating the presence of a glycoside for peak 2. Thus, peak 2 was tentatively identified as methyl 4-caffeoylquinate [41].

Phenolic Acids and Flavonoids
Peak 9 was tentatively identified as 3-caffeoylquinic acid according to the MS/MS data reported by Clifford et al. [43]. In this study, 5 peaks (peak 13, 14, 15, 16 and 17) were identified as flavonoids. For peak 13, compared with the mass spectra data of rutin standard under the same condition of LC-MS, it was definitely identified as rutin. Peak 14 and 15 showed common product ions at m/z 301 and 300 due to the quercetin core, and m/z 463 corresponded to quercetin glycoside. Similarly, peak 16 and 17 shared common product ions at m/z 285 and 284 due to the kaempferol core, and m/z 447 corresponded to kaempferol glycoside. By comparing with the mass spectra data in previously reported literature, peak 14-17 were tentatively identified as quercetin 3-O-glucoside, quercetin-acetyl-glycoside, kaempferol 3-O-glucoside and kaempferol-acetyl-glycoside [2], respectively. In this study, there were no flavonoids identified in root of M. oleifera, but in the TFC assay, the TFC of root of M. oleifera was 106.79 ± 2.12 mg rutin equivalent (RE)/g, and the reason for this may be due to the presence of ortho-diphenolic hydroxyl analogues, such as quinic acid in the roots, since the ortho-diphenolic hydroxyl compounds can coordinate with aluminum ions, and thus influence the determination of TFC.

Conclusions
In this study, the antioxidant assays with DPPH, ABTS and FRAP, anti-inflammatory activity tests by means of the effects on the NO production, and phytochemical analysis with HPLC-UV/ESI-MS/MS were conducted to compare and correlate the antioxidant and anti-inflammatory activities with chemical compositions among the leaves, seeds and roots of M. oleifera in Kenya. To the best of our knowledge, this is the first study aiming to reveal the correlations between the phytochemical profiles and antioxidant and anti-inflammatory activities of three different organs of M. oleifera from Kenya. As a result, we revealed that leaves of M. oleifera are rich in flavonoids and phenolic acids, which explained the best potential antioxidant and anti-inflammatory activities among three organs. This current study revealed the positive correlation between the flavonoids content in three different organs of M. oleifera with their antioxidant and anti-inflammatory activities, which suggested that the leaves from M. oleifera could be a more potential natural source of antioxidants and anti-inflammatory source than that of other organs, and is very promising for development into health promoting dietary supplements now and in the near future.