Natural Antioxidants in Foods and Medicinal Plants: Extraction, Assessment and Resources

Abstract

Natural antioxidants are widely distributed in food and medicinal plants. These natural antioxidants, especially polyphenols and carotenoids, exhibit a wide range of biological effects, including anti-inflammatory, anti-aging, anti-atherosclerosis and anticancer. The effective extraction and proper assessment of antioxidants from food and medicinal plants are crucial to explore the potential antioxidant sources and promote the application in functional foods, pharmaceuticals and food additives. The present paper provides comprehensive information on the green extraction technologies of natural antioxidants, assessment of antioxidant activity at chemical and cellular based levels and their main resources from food and medicinal plants.

1. Introduction

In biological system, reactive oxygen species (ROS) and reactive nitrogen species (RNS), such as superoxide, hydroxyl, and nitric oxide radicals, can damage the DNA and lead to the oxidation of lipid and proteins in cells [1,2,3]. Normally, antioxidant system occurring in human body can scavenge these radicals, which would keep the balance between oxidation and anti-oxidation. Nonetheless, the exposure of cigarette smoking, alcohol, radiation, or environmental toxins induces the production of excessive ROS and RNS, which disrupt the balance between oxidation and anti-oxidation and result in some chronic and degenerative diseases [3,4,5]. The increment of intake of exogenous antioxidants would ameliorate the damage caused by oxidative stress through inhibiting the initiation or propagation of oxidative chain reaction, acting as free radical scavengers, quenchers of singlet oxygen and reducing agents [6].

The exogenous antioxidants are mainly derived from food and medicinal plants, such as fruits, vegetables, cereals, mushrooms, beverages, flowers, spices and traditional medicinal herbs [7,8,9,10,11,12,13,14,15,16,17,18]. Besides, the industries processing agricultural by-products are also potentially important sources of natural antioxidants [19]. These natural antioxidants from plant materials are mainly polyphenols (phenolic acids, flavonoids, anthocyanins, lignans and stilbenes), carotenoids (xanthophylls and carotenes) and vitamins (vitamin E and C) [6,20]. Generally, these natural antioxidants, especially polyphenols and carotenoids, exhibit a wide range of biological effects, such as anti-inflammatory, antibacterial, antiviral, anti-aging, and anticancer [2,20,21,22,23,24,25,26,27,28,29,30,31].

Considering their important health effects, the efficient extraction methods of natural antioxidants, appropriate assessment of antioxidant activity as well as their main resources from food and medicinal plants are drawing great attention in food science and nutrition. To improve the extraction efficiency of antioxidant components from plant materials, several green non-conventional methods have been developed for reducing operational time and usage of organic solvents, such as ultrasound-assisted extraction, microwave-assisted extraction, enzyme-assisted extraction, pressurized liquid extraction, supercritical fluid extraction, high hydrostatic pressure extraction, pulsed electric field extraction and high voltage electrical discharges extraction. Moreover, to further assess the antioxidant capacities of extracts from natural products, especially those frequently consumed by people, different evaluation assays have been developed, e.g., Trolox equivalence antioxidant capacity (TEAC) assay, ferric ion reducing antioxidant power (FRAP) assay, oxygen radical absorbance capacity (ORAC) assay, inhibiting the oxidation of low-density lipoprotein (LDL) assay, cellular antioxidant activity assay, etc. These assays have been used for ranking the antioxidant plants and recommending best antioxidant foods for consumption. This review is aimed at summarizing the extraction methods of natural antioxidants, assessment methods of antioxidant activity and their main resources from food and medicinal plants.

2. Extraction Methods of Antioxidants from Foods and Medicinal Plants

Extraction is the first and crucial step for studying the natural antioxidants from plants (Figure 1). Many extraction factors play important roles in the extraction efficiency, such as type and concentration of extraction solvent, extraction temperature, extraction time, and extraction pH. Among them, the solvent is one of the most influential factors. Numerous solvents have been used for the extraction of antioxidants from food and medicinal plants. The selection of solvents is based on the chemical nature and polarity of antioxidant compounds to be extracted. Most of the phenolics, flavanoids and anthocyanins are hydrosoluble antioxidants. The polar and medium polar solvents, such as water, ethanol, methanol, propanol, acetone and their aqueous mixtures, are widely used for extraction [32,33,34,35]. Carotenoids are lipid-soluble antioxidants, and common organic solvents, such as the mixtures of hexane with acetone, ethanol, methanol, or mixtures of ethyl acetate with acetone, ethanol, methanol, have been used for extraction [36,37,38].

Various extraction procedures, including conventional extraction methods and non-conventional extraction methods, can be chosen to extract antioxidants from food and medicinal plants. The conventional extraction methods are mainly hot water bath, maceration and Soxhlet extraction, which are very time-consuming and require relatively large amounts of organic solvents with low extraction yields [39,40,41]. Furthermore, the long heating process such as hot water bath and Soxhlet extraction may lead to the degradation of the thermolabile compounds. To obtain antioxidants from plants in an energy-efficient and economically sustainable way, ultrasound, microwave, pressurized liquid, enzyme hydrolysis, supercritical fluids, high hydrostatic pressure, pulsed electric field, and high voltage electrical discharges have been studied as non-conventional methods (

Table 1. The application of non-conventional techniques in the extraction of antioxidants from some foods and medicinal plants.

Source	Compounds Extracted	Extraction Parameters		Extraction Improvement	Reference
		Non-Conventional Method	Conventional Methods
ultrasound-assisted extraction (UAE)
blueberry wine pomace	anthocyanins and phenolics	solvents: 70% ethanol and 0.01% hydrochloric acid; conditions: 400 w, 61.03 °C, 23.67 min	70% ethanol and 0.01% hydrochloric acid; 61 °C, 35 min without ultrasound treatment	increased total anthocyanins from 1.72 to 4.27 mg C3G/g (2.5-fold) and total phenolics from 5.08 to 16.41 mg gallic acid equivalent (GAE)/g (about 3.2-fold)	[42]
papaya	lycopene	solvents: 42.28% ethanol in ethyl acetate conditions: 40 kHz, 800 W, 26.09 min, 50.12 °C	40% ethanol in ethyl acetate (300 mL) 95 °C in a Soxhlet extractor	Recovery of lycopene increased from 68.3 ± 4.1 to 189.8 ± 4.5 μg/g	[43]
carrot	carotenoids	solvents: sunflower oil conditions: 22.5 W/cm2, 40 °C, 20 min	hexane at room temperature for one hour	obtained the β-carotene yield of 334.75 mg/L just in 20 min while the CSE method using hexane as solvent obtained the β-carotene yield of 321.36 mg/L after one-hour extraction	[44]
microwave-assisted extraction (MAE)
Achillea millefolium dust	antioxidants	solvents: 70% ethanol conditions: 170 W, 40 mL/g, 33 s	40% ethanol at room temperature (1:10, v/v) for 48 h	increased total polyphenol content from 135.26 ± 1.72 to 237.74 ± 2.08 mg GAE/g, total flavonoid content from 30.82 ± 2.35 to 42.95 ± 1.32 mg quercetin equivalents (QE)/g, 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging activity from 21.58% ± 0.88% to 71.72% ± 2.12%	[45]
Quercus bark	polyphenols	solvents: ethanol content 33%, methanol content 0.38% conditions: 50 Hz, 45 W, 60 min, pH 10.75, room temperature	the same extraction condition without microwave treatment	increased by 3 times and 2 times respectively for total phenolic content and antioxidant recoveries	[46]
enzyme-assisted extraction (EAE)
wine making by-products	phenolics	solvents: 70% acetone enzyme treatment with 2% viscozyme solution stirred for 12 h at 37 °C or 1 mg/mL pronase solution stirred for 1 h, then extraction with 70% (v/v) acetone in a gyratory water bath shaker at 30 °C for 20 min	the same extraction protocol without enzyme treatment	pronase and viscozyme increased the content of soluble phenolics while reducing the content of insoluble-bound phenolics	[32]
grape skins	flavonoids	solvents: buffer solution containing an appropriate amount of enzyme conditions: 10.52 mg/g Lallzyme EX-V, pH 2.0, extraction at 45 °C for 3 h	70% aqueous ethanol containing 1% formic acid for one day in the dark	improved recovery of anthocyanin contents (from 40,496.19 ± 58.18 to 41,752.95 ± 76.10 mg/kg) and flavan-3-ol contents (from 329.32 ± 2.46 to 345.94 ± 2.88 mg/kg)	[47]
tomato processing waste	lycopene	solvents: hexane/acetone/ethanol (50:25:25 v/v) conditions: 1.5% cellulase/2% pectinase at 4 h of incubation period	without enzyme treatment	increased the yield of lycopene from less than 200 to 847.33 μg/g (cellulase treatment) and to 1262.56 μg/g (pectinase treatment)	[48]
pressurized liquid extraction (PLE)
Aerial parts of Dracoceph-alum kotschyi	phenolics and flavonoids	solvents: methanol conditions: 74 °C, 34 bar pressure, 11.33 min static time, 17.45 min dynamic time, and 0.7 mL/min solvent flow rate	percolated with 1.0 L of methanol at room temperature (25 °C) according to the European Pharmacopeia	improved recovery of total phenolic (from 22.29 ± 0.05 to 30.92 ± 0.03 GAE mg/g), total flavonoids (from 5.042 ± 0.04 to 6.13 ± 0.07 QE mg/g) and luteolin content (from 9550 ± 0.3 to 13,247 ± 0.2 μg/g)	[49]
roots of Scutellaria pinnatifida	phenolics and flavonoids	solvents: methanol conditions: 65.8 °C, 39.2 bar pressure, 12.9 min static time, 18.9 min dynamic time, and 0.76 mL/min solvent flow rate	percolated with 1.0 L of methanol at room temperature	the total phenolic content increased from 196.66 to 396.94 mg/g, and the total flavonoid content increased from 91.3 to 127.78 mg/g	[50]
black bamboo leaves	antioxidants	solvents: 50% ethanol for the total phenolic (TP) and 75% ethanol for total flavonoid (TF) and 25% ethanol for DPPH radical scavenging ability conditions: 1500 psi, 200 °C, 25 min static time	reflux extraction method (~90 °C, 1 L solvent, 60 min)	improved extraction yields from 240 to 500 mg/1 g Dry black bamboo leaves (DL), TP contents from 1510 ± 3.2 to 2682 ± 0.9 mg/100 g, TF contents from 182 ± 2.7 to 657 ± 1.7 mg/100 g	[51]
palm pressed fiber	β-carotene	solvents: n-hexane conditions: 80 °C, 1500 psi, 2 × 10 min static extractions with flush volume 50%	extracted with n-hexane and chloroform in a Soxhlet apparatus for 8 h	obtained total β-carotene and vitamin contents comparable to Soxhlet extraction but with lower total organic solvent and rapid extraction process	[52]
supercritical fluid extraction (SFE)
myrtle leaves and berries	antioxidants	solvents: carbon dioxide conditions: 23 MPa, 45 °C and a CO2 flow of 0.3 kg/h using absolute ethanol as co-solvent with a flow rate of 0.09 kg/h	obtained by hydrodistillation using a Clevenger-type apparatus, for two hours	increased antioxidant capacity (about 20–40 times), polyphenolic contents (about 2 times) and myricetin-3-O-rhamnoside content (about 110–170 times in fruit and about 130–210 times in leaves)	[53]
Prunus persica leaves	phenolic compounds	solvents: carbon dioxide conditions: 60 °C, 150 bar and 6% ethanol co-solvent at a flow rate of 15 g/min and for a duration of 60 min	extracted 3 times with 30 mL of solvent system (acetone:methanol:water:formic acid, 40:40:20:0.1)	the radical scavenging activity value increased from 32.23% to 53.25%	[54]
high hydrostatic pressure extraction (HHPE)
prickly pear beverages prepared with 10% peel and 90% pulp	Phyto-chemical Compounds	400 or 550 MPa, room temperature, 0–16 min	thermally treated at 138 ± 1 °C for 2 s	increased TP content (16%–35%) and antioxidant activity (8%–17%) for Cristal (A) and Rojo San Martin varieties as well as increased the betaxanthin contents (6%–8%) and betacyanin content (4%–7%) for Rojo San Martin variety	[55]
Panax ginseng	phenolic compounds	600 MPa for 1 min at room temperature	conventional steaming	increased the total phenolic contents (from 1.13 to 1.37 mg maltol equivalent/g of red ginseng), especially maltol content (4.38 to 12.61 mg/100 g of red ginseng), also improved the ferrous ion chelating and superoxide dismutase activities	[56]
Pulsed electric field extraction (PEFE)
defatted canola seed cake	polyphenols	10% ethanol 30 V, 30 Hz and 10 s	microwave processing (5 min, liquid/solid ratio of 6.0 and 633.3 W)	less solvent usage, a shorter extraction time	[57]
Borago officinalis L. leaves	polyphenols	acidic water (pH 1.5) 1 to 7 kV/cm, 15–150 μs, 0.04 to 61.1 kJ/kg	the same extraction without PEF treatment	increased the TPC (1.3–6.6 times) and ORAC values (2.0–13.7 times)	[58]
orange peel	polyphenols	distilled water 5 kV/cm, 60 μs, 0.06 to 3.77 kJ/kg, pressurization at 5 bars for 30 min	the same extraction without PEF treatment	increased the naringin content from 1 to 3.1 mg/100 g and hesperidin content from 1.3 to 4.6 mg/100 g	[59]
high voltage electrical discharges extraction (HVEDE)
olive kernel	phenolic compounds	49% ethanol, 66 kJ/kg, pH 2.5	PEF with electric field strength E = 13.3 kV/cm and UAE at 400 W and 24 kHz	more effective polyphenol extraction (255 mg GAE/L for HVEDE versus 140 and 146 mg GAE/L for UAE and PEFE, respectively)	[60]

). In this section, the current application and developments of the non-conventional methods are summarized.

2.1. Ultrasound-Assisted Extraction (UAE)

Ultrasound assisted extraction (UAE) has been applied widely in the last three decades as an efficient extraction method in the food and pharmaceutical industries [61]. The mechanism is based on the cavitation phenomenon. The spread of ultrasound in liquid systems is via a series of compression and rarefaction waves, which can induce the production of cavitation bubbles within the fluid [62,63]. The size of these bubbles grow over the period of a few cycles until reach a critical point, then these bubbles collapse and release a great quantity of energy, which would generate extreme temperatures (5000 K) and pressures (1000 atmospheres) at room temperature. During the ultrasound assisted extraction of bioactive components from plant materials, the high temperature and pressure would destroy the cell walls, facilitate the release of bioactive compounds from plant cell walls and enhance the mass transport. The heat transfer of UAE is from outside of the plant cell to the inside, which is in the opposite direction of microwave assisted extraction.

Ultrasound frequency, intensity, temperature, and time can directly affect both extraction efficiency and yields. In addition, types of solvent, solvent volume, as well as sample characteristics such as moisture content of the sample and particle size are also the important factors for effective extraction [64]. Ultrasound-assisted extraction of anthocyanins and phenolic compounds in mulberry (Morus nigra) pulp was performed by Espada-Bellido et al. [65]. Several extraction variables, including methanol concentration (50%–100%), temperature (10–70 °C), ultrasound amplitude (30%–70%), cycle (0.2–0.7 s), solvent pH (3–7) and solvent–solid ratio (10:1.5–20:1.5) were optimized to obtain the high extraction yields. It was found that extraction temperature and solvent concentration were the most influential parameters for extraction of anthocyanins and phenolic compounds. Besides, different UAE conditions were suitable for different bioactive components. The optimum UAE conditions for anthocyanins are 76% methanol concentration, 12:1.5 solvent–solid ratio, 48 °C extraction temperature, cycle of 0.7 s and 70% ultrasound amplitude at pH 3, while the optimum UAE conditions for phenolic compounds were 61% methanol concentration, 11:1.5 solvent–solid ratio, 64 °C extraction temperature, cycle of 0.7 s and 70% ultrasound amplitude at pH 7. Under the UAE optimal conditions, the maximized extraction yields of total anthocyanins and total phenolic compounds were 149.95 μg/g and 1214.03 μg/g, respectively. Compared with conventional extraction, UAE showed several merits in terms of extraction yield and extraction time. Polyphenols from apple pomace were extracted using conventional extraction and UAE methods [66]. The results showed that ultrasound-assisted extraction increased the catechin equivalents by more than 20% and the release of flavan-3-ols and procyanidins were also enhanced by more than 25% in only 45 min. In another study, Liu et al. [67] used UAE and traditional boiling-water extraction methods to extract antioxidants from ** antioxidant parameter (TRAP), and inhibiting the oxidation of low-density lipoprotein (LDL).

3.1.1. Scavenging Free Radicals Assays

The Trolox equivalent antioxidant capacity (TEAC) assay is widely applied to evaluate the antioxidant ability to scavenge the ABTS radical [148,149,151,152]. According to the type of oxidation agent, there are two versions of this assay [153]. TEAC1: metmyoglobin-H₂O₂ oxidize ABTS to generate its colored ABTS•⁺ form; then, subsequent addition of antioxidants results in loss of the green color. TEAC2: potassium persulfate oxidize ABTS to generate its colored ABTS•⁺ form; then, subsequent addition of antioxidants results in loss of the green color. The version of TEAC1 is inaccurate because antioxidants also can react with the original HO• radical and the metmyoglobin except for the ABTS•⁺, which could cause an overestimation of antioxidant capacity. Therefore, the version of TEAC2 is more preferable. ABTS•⁺ has a UV-vis absorption maximum at 734 nm. The decrease of absorbance can be monitored spectrophotometrically. The difference of the absorbance tested is plotted versus the antioxidants concentrations. The antioxidant capacity was expressed as Trolox equivalents. Because ABTS•⁺ could react rapidly with antioxidants, the assay possesses the merits of rapidity and simplicity. Additionally, ABTS•⁺ is not influenced by ionic strength and is solvable in both organic and aqueous solvents, so it can be applied in multiple media to detect both hydrophilic and lipophilic antioxidant activities [154]. However, for slow reactions, the TEAC values tested is inaccurate when the duration of reaction is beyond 6 min [149].

As one of the few stable organic nitrogen radicals, the 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical is used to analyze the antioxidant activity [151,155]. The DPPH• posses a deep purple color and has a UV-vis absorption maximum at 515 nm [148,149]. The test compounds (antioxidants) reduce DPPH radical to DPPH-H and the solution color fades. The reducing ability can be assessed by measuring the decrease of its absorbance. In the end, the results are shown by EC₅₀ and TEC₅₀, that is, the necessary amount of antioxidant to decrease the initial DPPH concentration by 50% and the time taken to reach the steady state to EC₅₀ concentration [155]. DPPH assay is widely used in antioxidant capacity screening of fruit and vegetable juices or extracts, for it is easy, rapid and requires only a UV-vis spectrophotometer to test. Compared with ABTS assay, the DPPH radical is commercially available and does not have to be generated before assay such as ABTS•⁺. However, the application of DPPH assay is limited by its disadvantage. The linear reaction range of DPPH assay is narrow, only 2–3-fold. Moreover, for steric inaccessibility, antioxidants that possess strong antioxidant activities in lipid peroxidation system may react slowly or may even be inert to DPPH [148,149].

3.1.2. Reducing the Metal Ions Assays (FRAP and CUPRAC Assays)

The ferric-reducing antioxidant power (FRAP) assay measures directly the reducing capacity of antioxidants. In ferric-reducing antioxidant reactive system, the antioxidants can reduce a ferric tripyridyltriazine complex (Fe³⁺-TPTZ) to the ferrous complex (Fe²⁺-TPTZ) under pH 3.6 condition with a blank sample in parallel. The ferrous complex (Fe²⁺-TPTZ) is blue ferrous form and has a UV-vis absorption maximum at 593 nm. The ability of antioxidants in samples (FRAP value) is positive related to the increase in absorbance [151,156,157,158]. In general, this assay is suitable for some antioxidants that complete the reaction rapidly within 4 to 6 min, such as ascorbic acid and uric acid [159]. However, it has been demonstrated that the absorption of several dietary polyphenols in water and methanol slowly increased even after several hours, such as tannic acid, and caffeic acid [157]. In addition, FRAP cannot detect compounds that act by radical quenching (H transfer), particularly thiols and proteins. This results in a serious underestimation in serum sample [149].

The cupric reducing antioxidant capacity (CUPRAC) assay is similar to the FRAP method. CUPRAC method is conducted by mixing Cu(II)-neocuproine (Nc) chelate with antioxidant solution. The absorbance of the color Cu(I)-chelate as a result of redox reaction is measured at 450 nm after 30 min [160]. The application on this assay is less extensive than FRAP. However, this assay exhibits several merits in some ways. For example, the reagent in this assay is useful at pH 7, which is at physiological pH (as opposed to the Folin and FRAP assays, which work at pH 10 and pH 3.6, respectively). This method could be applied for the determination of both hydrophilic and lipophilic antioxidants because the Cu(II)-Nc is soluble in both aqueous and organic environments (unlike Folin and DPPH) [160,161]. In addition, CUPRAC assay can measure the reducing power of thiol-type antioxidants, such as glutathione and nonprotein thiols [150,162].

3.1.3. Folin–Ciocalteu Reagent (FCR) Assay

Folin–Ciocalteu reagent (FCR) assay is a widespread method for quantitative determination of phenolic compounds. The mechanism of Folin–Ciocalteu method is electron transfer (ET) [134,135]. It involves reducibility of phenols in alkaline solution (pH = 10), which is capable of turning yellow molybdotungsto-phosphoric heteropolyanion reagent into the blue resultant molybdotungsto-phosphate [148,163,164]. These blue pigments have a maximum absorption in the 700–760 nm range. The maximum absorption depends on the qualitative and/or quantitative composition of phenolic mixtures. The total phenols assay by FCR is simple, convenient, and has produced a large body of comparable data. Thus, it has become a routine assay in studying phenolic antioxidants from fruits, vegetables and medicine plants [163,165,166,167,168]. A large number of publications found excellent linear correlations between the “total phenolic profiles” by FCR and “the antioxidant activity” by other ET-based antioxidant capacity assay (e.g., FRAP, TEAC, etc.) [148,169,170].

3.1.4. Oxygen Radical Absorbance Capacity (ORAC) Assay

Generally, these assays estimate the capacity of antioxidants to protect a target molecule exposed to a free radical source. The oxygen radical absorbance capacity (ORAC) assay has been applied widely in the field of antioxidant and oxidative stress via H atom transfer [148,149,171,172]. In the basic assay, the peroxyl radical mixes with a fluorescent probe (FL; 3′,6′-dihydroxyspiro[isobenzofuran-1[3H],9′[9H]-xanthen]-3-one), then form a nonfluorescent product, which can be quantitated easily by fluorescence [149,171,173]. When an antioxidant is added into the mixture, peroxyl radical induced oxidation is inhibited and the decay of FL is prevented. Antioxidant capacity is reflected by determining the decreased rate and amount of product formed over time. Using AUC (area under curve) to reflect the antioxidant capacity is favorable because it applies to an antioxidant that has a lag phase or one that do not [171]. It is useful for a broad range of sample types, including raw fruit and vegetable extracts, plasma, and pure phytochemicals. Furthermore, the high-throughput assay is able to test several hundred samples daily by just using one plate-reader coupled with a multichannel automatic liquid handling system [148].

3.1.5. Total Radical Trap** Antioxidant Potential (TRAP) Assay

The total radical trap** antioxidant potential (TRAP) assay measures the ability of antioxidants to suppress the oxidation progress of 2,2′-azobis-2-methyl-propanimidamide, dihydrochloride (AAPH) or 2,2′-azobis(2-amidinopropane) dihydrochloride (ABAP) [148,174]. The variation in the reaction progress is monitored fluorometrically (λ_ex = 495 nm and λ_em = 575 nm). The fluorescence decay rate in the reaction slows after the addition of antioxidants compared with the rate before the antioxidants addition. The quantification is based on the lag-phase duration compared with the lag phase of Trolox [149,162]. The application of the lag phase is based on the assumption that the antioxidants show a lag phase and the length of the lag phase is positively correlated to antioxidant capacity. However, not every antioxidant component possesses an obvious lag phase and the potential of antioxidants that play a role after the lag phase is totally ignored.

3.1.6. Inhibiting the Oxidation of Low-Density Lipoprotein (LDL) Assay

Inhibition of induced lipid autoxidation has been developed as a measure of antioxidant capacity in a more physiologically relevant system [148,149,150]. Usually, the reaction solution contains free radical initiator (Cu(II) or 2,2′-azobis(2-amidinopropane) dihydrochloride (AAPH)), substrate (linoleic acid or LDL), and antioxidants. The autoxidation of linoleic acid or LDL is induced by Cu(II) or AAPH. The peroxidation of the lipid components is monitored at 234 nm by UV spectrometer for conjugated dienes. In the presence of a radical initiator, the reaction starts and the absorbance at 234 nm increases as the evidence of the accumulation of conjugated diene oxides. After the addition of antioxidants, the reaction rate slows down until the antioxidant is exhausted. In the period, the lag time is measured and used to evaluate antioxidant capacity.

Compared with other in vitro assays, the major advantage of this method is the use of a biological relevant substrate, which makes the results relevant to oxidative reactions in vivo [149]. Because LDL is isolated from blood samples, one of the major flaws of this method is the variability of the LDL samples, which can vary with different donors. Thus, this method is hard to be developed as a consistent, reproducible, high throughput antioxidant evaluation assay [148,149,150]. On the contrary, using linoleic acid or its methyl ester as an oxidation substrate would make the results more reproducible than using LDL. However, linoleic acid would form micelles in the presence of water, and the reaction progress in micelles cannot be monitored directly by UV absorbance, thus the accuracy of the method can be affected.

3.2. Cellular-Based Assays

The antioxidant capacity evaluated by chemical assays cannot completely reflect the behavior of the sample in vivo. It is necessary to estimate the effectiveness of antioxidants in more biologically relevant conditions. Animal models and human studies are more suitable for evaluation but more expensive and time-consuming [150]. As intermediate testing methods, cellular antioxidant activity (CAA) assay has been developed for evaluating the antioxidant capacities [175,176]. Dichlorofluorecin (DCFH) method is a commonly used CAA assay, which tests the capacity of antioxidants to prevent the oxidation of DCFH. In general, DCFH trapped within cells is easily oxidized to fluorescent dichlorofluorescein (DCF) by ABAP-generated peroxyl radicals in human hepatocarcinoma HepG2 cells. DCF could be monitored by fluorescence (λ_exc = 485 nm, λ_em = 538 nm). The decrease in cellular fluorescence is proportional to the antioxidant capacity of bioactive components. Except for HepG2 cells, several cell types have been applied for the CAA assay, such as human red blood cell, human endothelial EA.hy926, human colon cancer Caco-2 cell, human macrophage U937 cell and mouse macrophage RAW264.7 cell [150,176,177]. Besides, a CAA assay based on microfluidic cell chip with arrayed microchannels has been developed to assess plant antioxidants [178]. The microfluidic chip contains 288 round cell culture micro chambers and 48 independent parallel array channels. In this method, eight groups of different samples with six different concentrations could be tested simultaneously with multimode reader.

The assessment of antioxidant activity at cellular level is not limited to the test of ability of ROS/RNS scavenging but also includes tests of expression of antioxidant enzymes, inhibition of pro-oxidant enzymes, and activation vs. repression of redox transcription factors [150]. The antioxidant activities of the extracts prepared from five brown seaweeds was assessed in Caco-2 cells. Glutathione (GSH) content and antioxidant enzyme activity (catalase (CAT) and superoxide dismutase (SOD)) were evaluated. These cellular assays indicated that Pelvetia canaliculata could exert the antioxidant capacity mainly by preventing H₂O₂-mediated SOD depletion in Caco-2 cells [179]. Besides, antioxidant enzyme activities of glutathione peroxidase (GPx) and glutathione reductase (GR) were measured in three Argentinean red wines. Some protective effects of wine were observed in cells exposed to H₂O₂, which was attributed to the increased activity of antioxidant enzymes GPx and GR [180]. In addition, suppression of NF-κB activation as an anti-oxidant response has been observed in cultured cells with the treatment of phenols (e.g., curcumin) or food extracts (e.g., blueberries) [150,181,182]. In a study, it was observed that the activation of NF-κB and activator protein-1, as well as IL-8 release were suppressed in curcumin-treated alveolar epithelial cells. Additionally, in comparison with untreated cells, the levels of GSH and glutamylcysteine ligase catalytic subunit mRNA expression were increased [181].

4. Main Resources of Natural Antioxidants

Among various chemical based assays, the TEAC assay evaluates the ability to scavenge the free radical, the FRAP assay measures directly the reducing capacity of antioxidants, the total phenols assay by FCR assesses the phenolic contents from tested samples. To comprehensively study different aspects of antioxidants, the combination of TEAC, FRAP and FCR methods is frequently used to evaluate the antioxidant activity. As shown in

Table 3. The antioxidant capacities of some foods and medicinal plants.

Category	Varieties Showing Strong Antioxidant Activities	Assessment Method	Reference
antioxidant activities at chemical level
26 spices	oregano, cinnamon stick, clove, cinnamon, sage	Trolox equivalent antioxidant capacity (TEAC), Folin–Ciocalteu reagent (FCR)	[8]
62 fruits	Chinese date, pomegranate, guava, sweetsop, persimmon, Chinese wampee and plum, grape (red)	TEAC, ferric-reducing antioxidant power (FRAP), FCR	[11]
24 cereal grains	black rice, red rice, purple rice, buckwheat	TEAC, FRAP, FCR	[12]
49 Edible macro-fungi	Thelephora ganbajun, Boletus edulis, Volvariella volvacea, Boletus regius, and Suillus bovinus	TEAC, FRAP, FCR	[13]
56 vegetables	Chinese toon Bud, loosestrife, perilla leaf, cowpea, caraway, lotus root, sweet potato leaf, soy bean (green), pepper leaf, ginseng leaf, chives, and broccoli	TEAC, FRAP, FCR	[14]
223 medicinal plants	Acanthopanax gracilistylus, Agrimonia pilosa, Anemarrhena asphodeloides, Caesalpina sappan, Carthamus tinctorius, Dioscorea bulbifera, Fraxinus rhynchophylla, Lonicera japonica (flower), Magnolia officinalis, Mentha haplocalyx, Paeonia lactiflora (red), Polygonum multiflorum (Stem), Polygonum multiflorum (Root), Rhodiola sacra, Salvia miltiorrhiza, Tussilago farfara, Sargentodoxa cuneata	TEAC, FRAP, FCR	[15]
51 edible and wild flowers	Rosa rugosa, Limonium sinuatum, Pelargonium hortorum, Jatropha integerrima and Osmanthus fragrans, Orostachys fimbriatu, Chaenomeles sinensis, Calliandra haematocephala	TEAC, FRAP, FCR	[16]
50 fruit wastes	grape seed, hawthorn peel, longan peel, longan seed, mango peel, Chinese olive peel and sweetsop peel	TEAC, FRAP, FCR	[19]
antioxidant activities at cellular level
27 vegetables	beet, broccoli, and red pepper, eggplant, Brussels sprout, cabbage	cellular antioxidant activity (CAA) based on HepG2 cells	[183]
25 fruits	pomegranate and berries (wild blueberry, blackberry, raspberry, and blueberry)	CAA based on HepG2 cells	[184]
11 Thai pigmented (red and purple) and 2 nonpigmented rice varieties	hawm dowk mali deang (red), hawm deang sukhothai1 (red), hawm deang (red), man pu (red), red rose (red), klam moang (purple), klam chiang mai (purple)	CAA based on HL-60 cells	[185]
13 food legumes	black soybean, black bean, pinto bean, lentil, green pea, yellow soybean	CAA based on human gastric adenocarcinoma AGS cells	[192]
12 plantago species	P. lanceolata, P. himalaica, P. depressa, P. cornuti, P. jehohlensis	CAA based on HepG2 cells	[193]
seven cultivars of Aloe	Aloe. greenii, Aloe. arborescens	CAA based on HepG2 cells	[194]
five main Phyllanthus emblica L. cultivars	qingyougan, binggan and boligan	CAA based on HepG2 cells	[195]

, the antioxidant activities of many food and medicinal plants have been widely estimated, e.g., fruits, vegetables, cereal grains, edible and wild flowers, macro-fungi, medicinal plants, spices, etc., and the varieties showing strong antioxidant activities were displayed in Table 3 based on a combinative consideration of the results obtained by FRAP, TEAC and FCR assays. Overall, these results showed that different categories exhibited diverse antioxidant capacities and the variation was very large. The FRAP, TEAC and FCR values of 62 fruits varied from 0.11 ± 0.01 to 72.11 ± 2.19 μmol Fe(II)/g, 0.84 ± 0.03 to 80.68 ± 2.11 μmol Trolox/g, and 11.88 ± 0.11 to 585.52 ± 18.59 mg GAE/100 g, respectively [11]. The FRAP, TEAC and FCR values of 56 vegetables varied from 2.69 to 60.9 μmol Fe(II)/g, 6.93 to 33.63 μmol Trolox/g, and 4.99 to 23.27 mg GAE/g, respectively [14]. The FRAP, TEAC and FCR values of 24 cereal grains varied from 5.23 ± 0.23 to 126.19 ± 2.91 μmol Fe(II)/g, 0.62 ± 0.14 to 30.03 ± 1.10 μmol trolox/g, and 1.35 ± 0.15 to 9.47 ± 0.48 mg GAE/g, respectively [12]. The FRAP, TEAC and FCR values of 223 medicinal plants varied from 0.14 to 1844.85 μmol Fe(II)/g, 0.99 to 1544.38 μmol Trolox/g, and 0.19 to 101.33 mg GAE/g, respectively [15]. Obviously, among these varieties showing strong antioxidant activities, the antioxidant activities and total phenolic contents of medicinal plants were significantly higher than those of fruits, vegetables and cereals.

Additionally, the antioxidant activities of food and medicinal plants have also been evaluated by cellular antioxidant activity assays based on different cell types, and the varieties showing strong antioxidant activities were displayed in Table 3. The cellular antioxidant activities of 27 vegetables ranged from not detected (tomato) to 41.9 ± 6.2 μmol of QE/100 g (beet) [183]. The cellular antioxidant activities of 25 fruits ranged from 3.15 ± 0.21 (banana) to 292 ± 11 μmol of QE/100 g (wild blueberry) [184]. In both studies, these results showed that CAA values were significantly associated with total phenolic content. Surarit et al. [185] reported that the ethanolic bran extracts of 11 Thai pigmented (red and purple) and two non-pigmented rice varieties exerted the cellular antioxidant activities based on HL-60 cells in the following order: red > purple > non-pigmented rice, which showed the same order of phenolic and flavonoid contents in these rice extracts.

4.1. Natural Sources of Polyphenols

Polyphenols are abundant in food and medicinal plants, including phenolic acids, flavonoids, lignans, and stilbenes [20].

Phenolic acids comprise of derivatives of cinnamic acid (e.g., p-coumaric, caffeic, and ferulic) and derivatives of benzoic acid (e.g., gallic acid and hydroxybenzoic acids). Compared with the hydroxybenzoic acids, the hydroxycinnamic acids are more abundant in edible plants [20]. Fruits such as blueberries, kiwis, plums, cherries, apples are found to be rich in the hydroxycinnamic acids (0.5–2 g hydroxycinnamic acids/kg fresh wt). Caffeic acid is the most abundant phenolic acid and account for 75%–100% of the total hydroxycinnamic acid content in many fruits, whereas, ferulic acid is the most abundant phenolic acid in cereal grains and account for about 90% of total polyphenol content of wheat grain. The hydroxybenzoic acid content in edible plants is usually very low, except for certain red fruits, black radish, and onions. Due to the low content, they are not considered to be of great nutritional interest.

Flavonoids are abundant in most edible fruits and vegetables. Its subclasses include flavonols, flavanones, catechins, flavones, anthocyanidins and isoflavonoids. The concentration and type of flavonoids vary in different dietary sources [186]. Quercetin is usually the most abundant flavonol in edible plants. The richest dietary source of quercetin is onion. Tea and wine have relatively low amounts of quercetin. Other flavonols include myricetin (berries), isorhamnetin (onion), and kaempferol (broccoli). Flavanones almost only exist in citrus fruits. Hesperidin and narirutin are the main flavonoids of oranges and mandarins, whereas, naringin and narirutin are the main flavonoids of grapefruit. Catechins usually exist in the form of aglycones or esterified with gallic acid. The richest dietary sources of catechins are tea and red wine. In addition, the major flavones are apigenin and luteolin. The major dietary sources are red pepper and celery. Anthocyanins, such as pelargonidin, cyanidin, and delphinidin, contribute to the red, blue, or violet color of edible plants, such as plums, eggplant, and many berries. The isoflavonoids, such as isoflavones genistein and daidzein, mainly exist in legumes. The richest dietary source are soybean and soy products [186,187].

The richest dietary source of lignans is linseed containing secoisolariciresinol (up to 3.7 g/kg dry wt) and low quantities of matairesinol [20]. Other algae, leguminous plants (lentils), cereals (triticale and wheat), fruit (pears, prunes) and certain vegetables (garlic, asparagus, carrots) also have traces of these same lignans. Resveratrol is a stilbene, which has been extensively studied for its multiple bioactivities. Red wine is rich in resveratrol (0.3–7 mg aglycones/L and 15 mg glycosides/L).

4.2. Natural Sources of Carotenoids

Carotenoids are natural pigments, including β-carotene, lycopene, lutein, and zeaxanthin [188]. All colorful edible plants, especially dark green and yellow-orange leafy, are the good sources of carotenoids. Due to the lipid solubility of carotenoids, the absorption mainly depends on their preparation with oils or fats. Among the carotenoids, the β-carotene commonly occurs in edible plants that possesses the highest activity of provitamin A, such as acerola, mango, pumpkin, carrot, nuts, and oil palm. Lycopene is a kind of red pigment. It almost exists only in vegetables and algae tissues. Tomato products such as juices, soups, sauces, and ketchup, as well as their processing waste and peel are important sources of lycopene [189]. The lycopene (79%–91%) presented in tomatoes is mostly in the form of the trans isomer. Lutein and zeaxanthin are the most common xanthophylls in green and dark leafy vegetables such as broccoli, spinach, peas and lettuce [190]. Zeaxanthin also found in the red marine microalga P. cruentum (making up 97.4% of the total carotenoids) (Mezzomo et al., 2016; Raposo et al., 2015) [188,191].

5. Conclusions

Antioxidants derived from food and medicinal plants have been increasingly investigated for their various nutritional function and health benefits. In this review, the extraction methods of natural antioxidants, assessment methods of antioxidant activity as well as their main resources from food and medicinal plants are summarized. The non-conventional techniques described have potential to replace or enhance existing extraction techniques due to the less extraction time, energy consumption, and usage of harmful organic solvents, as well as higher extraction yields to recover antioxidant compounds from food and medicinal plants. Nevertheless, most of them are limited for industrial applications due to the high equipment costs and complicated installation procedures. Thus, establishing the balance between energetic and cost will be a key research in the future. To take advantage of the different extraction methods and to limit their drawbacks, the combinative application of multiple extraction technologies and the automated potential of these non-conventional extraction technologies would be the development tendency in future. For assessing the antioxidant activity of plant materials, several assays are suggested, such as the determination of total polyphenolic content by FCR, scavenging free radical ability by TEAC, metal-reducing activity by FRAP, and a kind of cellular-based assay. Furthermore, in order to make the comparison of different samples and studies possible, standardizing the operating conditions of the same analysis method and the expression of results are recommended.