Evaluation of Garlic Cultivars for Polyphenolic Content and Antioxidant Properties

Two phenolic compound parameters (total phenolic and flavonoid contents) and 5 antioxidant parameters (DPPH [2, 2-diphenyl-1-picrylhydrazyl] radical scavenging activity, HRSC (hydroxyl radical scavenging capacity), FRAP (ferric ion reducing antioxidant power), CUPRAC (cupric ion reducing antioxidant capacity), and MCA (metal chelating activity) were measured in bulbs and bolts of 43 garlic cultivars. The bulbs of cultivar ‘74-x’ had the highest phenolic content (total phenolic, flavonoids) and the strongest antioxidant capacity (DPPH, FRAP, and CUPRAC), followed by bulbs of cultivar ‘Hanzhong purple’; the bulbs of cultivar ‘Gailiang’ had the lowest phenolic content and antioxidant capacity (FRAP, CUPRAC, MCA). The bolts of ‘Hanzhong purple’ also had higher phenolic content. Principal components analysis (PCA) separated the cultivars into 3 groups according to phenolic and flavonoid contents and strength of antioxidant activity. The first group had higher HRSC, FRAP, and flavonoid content; the second group had higher total phenolic content and MCA; some cultivars in the third group had higher HRSC and FRAP. All 8 test garlic bulb extracts successfully prevented Human Vascular Endothelial Cell death and significantly prevented reactive-oxygen species (ROS) formation in oxidative stress model, in which cultivar ‘74-x’ had highest protection capability, following by cultivar ‘Hanzhong purple’, and the bulbs of cultivar ‘No. 105 from Korea’ had the lower protection capability against cell death and ROS formation. The protection capability in vivo of these garlic cultivars was consistent with their phenolic content and antioxidant capacity.


Introduction
Garlic (Allium sativum L.) is one of the most commonly produced vegetables worldwide. According to the United Nations Food and Agriculture Organization (FAO), approximately 10 million metric tons of garlic is produced annually on approximately 1 million hectares (2.5 million acres) of land. China is by far the largest producer of garlic, producing over 75% of world tonnage. Garlic is a source of various biologically active phytomolecules, including organosulfur compounds, phenolic acids, allyl thiosulfinates, flavonoids, and vitamins. The health properties of garlic depend on its bioactive compounds and especially on phenolic compounds [1,2], which have interesting pharmacological properties, are present in relatively high amounts [3]. Many researches have been conducted to assess the dietary role of polyphenolic substances, and their characteristics, metabolic pathways, and biological effects [4,5], so garlic has been widely used to scavenge Reactive oxygen species (ROS) [6] and treat a variety of diseases including heart disease and cancer [7]. The extract of garlic has significant antioxidant activity and protective effects against oxidative DNA damage [8], decreasing fibrinogen and increasing antioxidant activity in the plasma of rats [9] and reducing the radiation sensitivity of normal tissues that are adjacent to tumors [10], so the extract might be useful in preventing endothelial dysfunction [11]. Garlic is thus widely used to protect humans against oxidative stress [12], reduce the risk of chronic diseases [13], prevent disease progression, and treat or prevent atherosclerosis [7,14,15] and cancer [14].
Interest in natural antioxidants and particularly in dietary antioxidants, which are present in vegetables and contribute to protection against oxidative stress in humans, is increasing. Garlic possesses potential health-promoting effects due to its high phenolic phytochemical content and is a source of natural antioxidants [16]. The total phenolic acid content of a local garlic cultivar grown near Namhae-gun, Korea was 17.86 mg/kg of dry matter (dm) and the total flavonoid content was 29.95 mg/kg dm [4]. The total phenolic content varied from 3.4 mg gallic acid equivalents (GAE)/g dm to 10.8 mg GAE/g dm in different garlic cultivars grown at four locations in Andalusia, Spain [3]. The content of phenolic compounds in garlic thus varies greatly with genetic, agronomic, and environmental factors [17], and it is well known that cultivar is the primary factor that determines this variation. Bulb firmness, pH, soluble solids, moisture content, and sugar content differed across 14 garlic cultivars [18]; other traits that vary across cultivars when grown under the same environmental conditions include the leaf number before bolting, flowering date, final stem length, flower/topset ratio, and pollen viability [19]. Variation in allicin, allyl methyl thiosulfinate, and allyl trans-1-propenyl thiosulfinate has also been observed in 93 garlic ecotypes [20]. Although some studies have investigated phenolic compounds and antioxidant activity in garlic, very few reports have compared the distribution of phenolic compounds among garlic cultivars in China, or between garlic bulbs and bolts among these cultivars. It is necessary to assess this trait in garlic germplasm resources to assure their effective utilization [21][22][23][24].
This study was designed to determine the variability in total phenolics, flavonoids, and antioxidant capacity among bulbs and bolts of 43 garlic cultivars for further development and utilization of these cultivars.

Materials
The 43 cultivars employed in this study were selected from the garlic germplasm nursery of Northwest A&F University College of Horticulture, Yangling, China; 40 of the cultivars were introduced from China, while 1 was from Ethiopia, 1 was from Thailand, and 1 was from Korea. The numeric identifiers of these 43 major garlic cultivars are shown in Table 1. The cultivars were planted in the garlic germplasm nursery, Northwest A&F University. The row spacing was 20 cm, plant spacing was 15 cm, and depth of the seed furrow was 5 cm. Human umbilical vein endothelial cells (ECV-304) were supplied by assistant professor Dong who worked in the College of Veterinary Medicine, Northwest A&F University.

Samples
Fresh bolts of the 32 bolting cultivars (the others were nonbolting cultivars) were sampled and collected in April 2009 and 2010 and the bulbs of all mature garlic cultivars were sampled in July 2009 and 2010 after air drying in the field. For each cultivar, samples were prepared by mixing equal amounts of garlic powder from 5 bulbs and 10 bolts selected randomly. The garlic cloves or bolts were manually peeled and immediately ground into powder in liquid nitrogen. The powder was stored at 270uC until analysis.

Preparation of phenolic extracts
The garlic bulbs and bolts were first treated using the method described by Park et al. [8] with minor modifications. Raw garlic was freeze-dried after steam blanching and was soaked in distilled water in a sealed flask for 14 d in a 70uC water bath. The extracts were centrifuged at 40006 g for 10 min at 4uC using a Sorvall RC-5C Plus centrifuge (Kendro Laboratory Products, Newton, CT, USA). The supernatants were stored at 220uC in the dark until use.
For testing cytotoxicity and ROS formation in vitro, the phenolic extracts were prepared from garlic bulbs and the extracts were freeze-dried in the freeze dryer (FD8505, Sigma-Aldrich Chemical Co., Shanghai, China) and then the phenolic extracts concentrations were adjusted to 1 mg/mL. Photometric determination of total phenolic content (TPC) and total flavonoid content (TFAC) Total phenolic content was determined in the bulbs and bolts of garlic cultivars using the Folin-Ciocalteu method [25] and a Shimadzu UV-1700 analyzer (Shimadzu Co.). Garlic extracts (100 mL) were diluted using 5.9 mL water and then mixed. Next, 200 mL Folin-Ciocalteu reagent was added into the mixture, and 2 mL sodium carbonate solution was added 1 min later. The mixture was allowed to react at room temperature in the dark for 120 min, and then the absorbance was measured at 735 nm. Gallic acid was used as a standard, and the results were expressed as milligrams of gallic acid equivalents (GAE) per gram of garlic fresh weight (FW). The total flavonoid content (TFAC) was determined according to the method of Sellappan et al. [26] with some modifications. One milliliter of garlic extract was added to 200 mL of 0.5 mol/L sodium carbonate solution. The mixture was allowed to stand for 5 min and was then added to 200 mL of 0.3 mol/L aluminum chloride crystalline and incubated for an additional 5 min. Then, 1.0 mL NaOH (1 M) was added to the reaction system and the absorbance was measured against a blank at 510 nm. The results were calculated and expressed as micrograms of rutin equivalents (RAE) per gram dry weight (DW).

Measurement of antiradical properties
We followed a previously described method for measuring DPPH free radical-scavenging capacity [27,28], with minor revisions. First, 100 mL garlic extracts were diluted with 900 mL water. DPPH methanolic solution was then added and the mixture was kept in the dark for 30 min. The absorbance at 515 nm was recorded to determine the concentration of remaining DPPH, and the results were expressed as micromoles of Trolox equivalents per gram of garlic mass.
The percentage inhibition of DPPH of the test sample and known solutions of Trolox were calculated by the following formula: Where Ai is the absorbance of the sample at 515 nm, and Ac is the absorbance of the blank at 515 nm.
HRSC was estimated by the methods of Prasad et al. [13]. Briefly, 3 mL distilled water and 100 mL FeSO 4 (0.02 M) were added to a microfuge tube. Next, 45 mL H 2 O 2 (0.15%) solution and 1 mL SA (8 M) were added. The final volume of the reaction mixture was then added to 100 mL sample solution and kept in the dark for 50 min at 20uC. The absorbance at 510 nm was recorded, and HRSC was calculated as follows: FRAP was measured as previously described [29] with minor modifications. First, 100 mL sample, 2.5 mL phosphate buffer (0.2 M, pH 6.6), and 2.5 mL potassium ferricyanide solution (1%) were sequentially mixed, and the mixture was then incubated in a 50uC water bath for 20 min before cooling. Next, 2.5 mL potassium ferricyanide (1%) was added, and the mixture was incubated in a 50uC water bath for 20 min. After cooling, 2.5 mL 3-chloroacetic acid (10%) was added and the solution was mixed. Then, 2.5 mL the mixture was extracted into a new tube, and 2.5 mL distilled water was added. Finally, 300 mL FeCl 3 (0.1%) was added to this mixture and the reaction was allowed to proceed for 5 min at room temperature in the dark. The absorbance of the product was measured at 700 nm; FRAP was expressed as this absorbance.
CUPRAC was determined using previously described methods [30], with minor modifications. Briefly, 0.1 mL garlic extract was mixed with 1 mL CuSO 4 (5 mM), 1 mL neocuproine (3.75 mM), 1 mL ammonium acetate (1 mM), and 1 mL distilled water and kept in the dark for 30 min in a 37uC water bath. Absorbance was then measured at 450 nm. Results are expressed in milligrams of Trolox per liter of extract.
Metal chelating capacity was measured using the methods of Prakash et al. [29]. One-hundred microliters of extract was mixed with 3.9 mL distilled water, 100 mL FeCl 2 ?4H 2 O (2 mM), and 50 mL ferrozine (5 mM). The reaction mixture was kept in the dark for 10 min and the Fe 2+ concentration was then monitored at 562 nm. The percentage chelating capacity was expressed as follows: Where A is absorbance at 562 nm.
ECV-304 viability was tested microscopically by plasma membrane disruption, as determined by the trypan blue (0.1% w/v) exclusion test [31]. Human Vascular Endothelial Cell viability was assessed at 3 hours, and the cells were at least 90% viable before use.
ECV-304 reactive-oxygen species (ROS) generation induced by tertiary butyl-hydroperoxide was determined by adding dichlorofluorescin diacetate (DCFH-DA) to the Human Vascular Endothelial Cell. DCFH-DA penetrated cell and was hydrolyzed to form non-fluorescent dichlorofluorescin (DCFH). DCFH was then oxidised by ROS to form the highly fluorescent dichlorofluorescein (DCF). After incubation with tertiary butyl hydroperoxide, 1 mL samples of cell were withdrawn at 3 hours and centrifuged at 10006 g for 5 minute. The cells were resuspended in DMEM and 10 mM DCFH-DA was added. Cells were allowed to incubate at 37uC for 20 minutes. The fluorescent intensity of DCF was determined by VictorTM X5 multilabel reader (PerkinElmer Life and Analytical Sciences) measuring the excitation and emission wavelengths, which were 490 and 520 nm, respectively.

Statistical analysis
All data are expressed as the mean 6SD of 3 replicates, and statistical analyses were subjected to ANOVA procedures (DPS v7.55 for Windows). Significant differences among cultivars were detected by Duncan's multiple range tests. P-values ,0.05 were considered significant.
Frequency distribution histograms of polyphenolic content were constructed according to the distribution interval among cultivars, where the interval was equal to the Value maximum minus the Value minimum divided by N [32].
Principal component analysis (PCA) was used to detect clustering and to establish relationships between cultivars and polyphenolic content and antioxidant properties. To eliminate the influence of dimension, data were classified into 10 grades for analysis: grade 1,X22d and grade 10.X+2d, where the interval of each grade was 0.5d, and d was the standard deviation.

Results and Discussion
The distribution of garlic polyphenolic content and antioxidant properties among different cultivars The frequency distributions of TPC and TFAC among bulbs and bolts of different garlic cultivars are shown in Fig. 1. TPC in bulbs exhibited a partially normal distribution and ranged from 21.27 to 33.96 mg GAE/g FW in most cultivars (Fig.1A). Previous studies have shown TPC values of 5.63 mg GAE/g in aged garlic extract [3,8] and 15.23 mg GAE/g FW [32], which are much lower than the TPC observed in our 43 garlic cultivars. However, TPC in bolts of 78.1% of hardneck garlic cultivars was much lower than that in bulbs, ranging from 9.24 to 15.48 mg GAE/g, and these results are in accordance with those of Nuutila et al. [16], who detected a TPC of 0.075-0.080 mg GAE/g in different Allium species.
The TFAC in both bulbs and bolts exhibited an almost normal distribution, ranging from 0.11 to 0.59 mg rutin DW/g FW and 0.075 to 0.675 mg rutin DW/g FW, respectively (Fig.1B). There was no significant difference in TFAC between the bulbs and bolts of different cultivars.

Polyphenolic content and antioxidant properties of different cultivars
The extracts of bulbs from over 40 cultivars were evaluated for total phenolic content and antioxidant activities to determine the extent of genotypic variation. The coefficient of variation in TPC in bulbs and bolts was 3.06%, indicating that the TPC varied significantly among different cultivars (Fig. 2). TPC in bulbs of the 43 garlic cultivars varied from 17.16 to 42.53 mg GAE/g. Bulbs of cultivar '74-x' had the highest TPC, followed by the 'Hanzhong purple' cultivar; cultivar 'Gailiang' had the lowest TPC. These results suggest that cultivar '74-x' may be a better source of phenolic compounds than other garlic cultivars ( Fig. 2A).
The TFAC in bulbs also varied significantly (more than 4-fold) among cultivars (Fig. 2B), with a coefficient of variation of 2.70%. TFAC ranged from 0.15 (cultivar 'No. 105 from korea') to 0.60 mg rutin DW/g (cultivar '74-x' respectively). Cultivar '74-x' also had the highest TFAC in bulbs of the 43 cultivars, followed by the cultivar 'Hanzhong purple', while cultivars 'No. 105 from korea' and 'Gailiang' had the lowest TFAC. This result was consistent with that of TPC.
Variability of TPC and TFAC in bulbs of different cultivars could be attributed to various cultivar characteristics. Clove size must be considered, as it indirectly affects the final concentration of phenolic compounds. In agreement with our results, previous reports have shown that different garlic cultivars had different yields [33], allicin content [21], and polyphenolic content [32], although Sterling and Eagling [34] reported that the variety and origin of garlic were not important factors affecting the agriculture traits of this crop.
TPC in garlic bolts of 32 cultivars varied from 10.17 mg GAE/ g (cultivar 'Xiangfan ershui early') to 22.66 mg GAE/g (cultivar 'Hanzhong purple'), which was much higher than that in leaves of Allium roseum L., where the polyphenol content was 1.23 mg GAE/ g aged extract [35]; the TPC in bulbs was higher than that in bolts. In these 32 cultivars, TFAC in garlic bolts varied from 0.06 to 0.67 mg rutin DW/g (cultivars 'Baihe early' and 'Sicuan red' respectively) 'Hanzhong purple', which had the highest TPC also had a higher TFAC. The coefficients of variation for TPC and TFAC were 3.47% and 5.85%, respectively. Normally, organisms have several different antioxidant systems that may be involved in various interactions; thus, the methods used to analyze antioxidant activity should accurately reflect all of the antioxidants in a complex system. In this study, 5 in vitro assays (DPPH, HRSC, FRAP, CUPRAC, and metal chelating capacity) were used as complementary methods to evaluate the potential antioxidant activity of garlic cultivars. Significant differences were observed among different cultivars in these assays ( Table 2). Among the methods available, those based on the elimination of free radicals, particularly on scavenging of DPPH radicals have been used most frequently. The DPPH-scavenging activity of garlic bulbs ranged from 3.60% to 45.63%, which was exceeded the range (5.07% to 11.36%) reported for the Allium genus [35]. In contrast, DPPH values were much higher (66.48%) in the leaves of A. sativum L. and in fried garlic (60.85%) [6].
Cultivars 'No. 97 from Guizhou', 'Baihe early', and 'Xiangfan ershui early' had relatively low antioxidant activities, while cultivar '74-x', which had the highest TPC among all the cultivars, showed the highest antioxidant activity. The DPPH-scavenging activity of cultivar '74-x' was 12.68-fold higher than that of cultivar 'No. 97 from Guizhou'. Significant differences in HRSC were found among cultivars; and the strongest scavenging activity (81.31%) was observed in bulbs of cultivar '87-x' and the lowest NOH scavenging capacity (37.87%) was shown in bulbs of cultivar 'Pengxian late'.
Currently, the most frequently used method for measuring antioxidant activity is FRAP, which is often expressed as absorbance value, i.e., the higher the absorbance, the stronger the reducing capability. Significant differences based on FRAP assays were observed among cultivars. The lowest absorbance value in bulbs was 0.273 (cultivar 'Gailiang'), and the TPC of this cultivar was also the lowest among all cultivars. The highest absorbance value was 0.613 (cultivar 'Hanzhong purple'), followed  In contrast to FRAP, which is based on the ferric-ferrous system, the CUPRAC method is based on the potential to reduce copper ions. We found that cultivar '74-x' possessed the highest reducing activity as measured by CUPRAC (8305.43 mg Trolox DW/g), and cultivar 'Bijie' had the lowest reducing activity (3200.40 mg Trolox DW/g; 38.53% that of cultivar '74-x'). The metal chelating capacity in bulbs ranged from 1.71% (cultivar 'Gailiang') to 22.98% (cultivar 'Sichuan red').
Antioxidant activities in bolts differed greatly according to the assay used ( Table 2)

PCA of polyphenolic content and antioxidant properties
PCA allows a large number of variables to be reduced to just a few that accounts for most of the variance in the observed results. Thus, we performed PCA to determine which garlic cultivars were associated with TPC and antioxidant activity in their bulbs (Fig.3).  (Fig 3B).
Almost all of cultivars in the first group had the higher positive value of PC1 so it indicated that the first group had higher HRSC, FRAP and flavonoid content. The second group had lower negative value of PC1 so it indicated that the second group had higher total phenolic content and MCA. Some cultivars such as

Correlation between phenolic compounds and antioxidant capacity
Correlation analyses were performed among the phenolic compounds and antioxidant capacity parameters, and between each phenolic compound and antioxidant capacity measurement for the 43 cultivars. These analyses showed significant positive correlations between TPC and TFAC, and among DPPH, FRAP, and CUPRAC in both bulbs (Table 3) and bolts (Table 4). There were also significant correlations among DPPH, FRAP, and CUPRAC antioxidant assay measurements both in bulbs and bolts. These findings indicated that the DPPH, FRAP, and CUPRAC methods were stable and reliable in measuring antioxidant capacities of garlic. The correlation between HRSC scavenging activities and polyphenolic compounds was weak, which could be explained by the fact that the presence of other non-flavanol compounds had a strong influence on overall scavenging activity.
In vitro Protection against tertiary butyl hydroperoxide induced cytotoxicity and ROS formation in Human Vascular Endothelial Cells (Oxidative Stress Model) The bulb extracts of eight garlic cultivars (extract concentrations 1 mg/mL) were tested and compared for cytoprotection against oxidative stress cell death and ROS formation in Human Vascular Endothelial Cells. It showed that all the bulb extracts tested successfully prevented cell death and the order of protection against cell death was, from most protective to least protective after 3 hours incubation: '74-X' extract . 'Hanzhong Purple' extract = 'Suzhou white' extract . 'Ningqiang mountain garlic' extract. 'Cangshan' extract = 'Russian garlic' extract . 'No. 105 from Korea' extract = 'No. 97 from Guizhou' extract (     Actually, the order of protection of garlic bulb extracts against oxidative stress cell death was almost the same tendency of TPC and TFAC in these test cultivars even the antioxidant capacity. The TPC order was '74-X'. 'Hanzhong Purple' .'Ningqiang mountain garlic' . 'Cangshan' . 'No. 97 from Guizhou' . 'Russian garlic' . 'No. 105 from Korea' . 'Suzhou white'; and the TFAC of these test cultivars was '74-X'. 'Hanzhong Purple' . 'Ningqiang mountain garlic' . 'No. 97 from Guizhou' . 'Cangshan' . 'Suzhou white' . 'Russian garlic' . 'No. 105 from Korea'. It was showed that protection capability of these garlic bulb extracts against oxidative stress cell death was consistent with their phenolic content and antioxidant capacity. Due to the variety of polyphenolic compounds found in garlic, it is likely that multiple protective mechanisms may act against oxidative and carbonyl induced cytotoxicity in in vitro models. One possible mechanism whereby garlic extracts may decrease oxidative stress in cells is in the prevention of lipid peroxidation [36].

Conclusions
On the basis of PCA, the garlic cultivars examined in this study could be divided into 3 groups. Group 1 contained 23 cultivars with stronger HRSC and FRAP, as well as higher flavonoid contents. The second group consisted of 16 cultivars with a higher TPC and MCA, and the third group consisted of 4 cultivars with stronger HRSC and FRAP. In addition, significant correlations among different antioxidant assays were observed in both bolts and bulbs. These antioxidant properties were highly correlated with the presence of the primary phenolic compounds. It was showed that the bulb extracts of eight test garlic cultivars successfully prevented Human Vascular Endothelial cell death and ROS formation, in which cultivar '74-x' and 'Hanzhong purple' had superior protective effects and cultivar 'No. 105 from Korea' had the lower protection capability against cell death and ROS formation, which was consistent with their phenolic content and antioxidant capacity.