Profiling of organosulphur compounds using HPLC-PDA and GC/MS system and antioxidant activities in hooker chive (Allium hookeri)

Abstract This study investigates the comprehensive organosulphur compounds and evaluation of antioxidant activities in Allium hookeri, called hooker chive. The non-volatile and volatile organosulphur compounds were determined by high-performance liquid chromatography-photodiode array detector (HPLC-PDA) and gas chromatography-mass spectrometry (GC-MS), respectively. Among 11 non-volatileorganosulphur compounds, methiin and cycloalliin were major compounds. A total of 42 volatile compounds were identified and allyl methyl sulphides and dimethyl sulphides were primarily volatiles. The total phenols ranged from 19.32 to 71.19 mg gallic acid equivalents (GAE)/g. The antioxidant capacities expressed IC50 values ranged from 9.59 to 38.54 mg/mL for DPPH radical scavenging activity and from 0.39 to 2.36 for ferric reducing antioxidant power assay. The results were proposed as useful tools for evaluating the nutraceutical of hooker chive.

Generally, only a few Allium plants are commonly known as edible vegetables, notably, onion (Allium cepa), garlic (Allium sativum), chive (Allium schoenoprasum), leek (Allium ampeloprasum var. porrum) and rakkyo (Allium chinense). recently, Allium hookeri, commonly called hooker chive, was introduced to cultivate in the southern region of South Korea (rhyu & Park 2013). Hooker chive is a wild herb originating from India and Myanmar. The root of the plant is used in food and medicines in Asia; the leaves are used for teas and seasonings (Sangtam et al. 2012). The health benefits of hooker chive include anti-inflammatory and anti-cancer effects that are related with bioactive compounds, such as organosulphur and phenolic compounds (Bae & Bae 2012). However, hooker chive, despite the known health benefits of its various bioactive compounds, remains under-utilised in the average diet and rarely researched worldwide. Although rhyu and Park (2013) reported the characterisation of alkyl thiosulphinates in hooker chive using HPLC-ESI-MS, only the amount of diallyl thiosulphinate (allicin) was quantified. Therefore, it is essential that the bioactive compounds in hooker chive should be quantified for the assessment as a functional food. Accordingly, this study reported the identification and quantification of organosulphur compounds in hooker chive by liquid chromatography-photodiode array detector (HPLC-PDA) system. In addition, volatile compounds in hooker chive were identified using gas chromatography-mass spectrometry (GC-MS). Moreover, antioxidant capacities were estimated by ferric reducing antioxidant power and DPPH radical scavenging assays. Thus, the overall nutraceutical values of hooker chive were investigated.

Determination of non-volatile organosulphur compounds
For simultaneous analysis of all 11 non-volatile organosulphur compounds with an HPLC-PDA system, the analytical method was performed following Yoo et al. (2010). However, identification and quantification of non-volatile organosulfur compounds were difficult due to complex matrix effect of hookeri chive. For this reason, the analytical methods for determination of non-volatile organosulphur compounds were conducted following Yoo et al. (2010) and Lee et al. (2013).
The non-volatile organosulphur compounds were classified into three groups: γ-glutamyl peptides, S-alk(en)yl-l-cysteines and S-alk(en)yl-l-cysteine sulphoxides (ACSo). The contents of non-volatile organosulphur compounds are shown in Table S1 (supplementary material  Table S1). The levels of organosulphur compounds varied according to the domestic or imported sample. The amounts of γ-glutamyl peptides, GSMC, GSAC, GSPC and γ-GPA, were 0.13 to 2.05, 0.34 to 2.63, 0.24 to 1.67 and 0.01 to 0.15 mg/g of dry weight (DW), respectively. The leaves had significantly greater γ-glutamyl peptide levels (except GSPC) than the roots (p < 0.05). The imported roots had higher levels of γ-glutamyl peptides than domestic sample roots. Ichikawa et al. (2006) determined the levels of organosulphur compounds in garlic cloves: GSAC, GSPC and GSMC levels were 5.43 to 14.87, 7.89 to 15.12 and 0.78 to 1.27 mg/g of DW, respectively. The contents of GSAC and GSPC were considerably lower compared with their levels in garlic (A. sativum L.). However, the GSMC level of the leaves was greater than that of garlic, ranging from approximately 1.6 to 2.6-fold greater. The level of S-alk(en) yl-l-cysteines, S-methyl cysteine (SMC) and trans-1-propenyl cysteine (TPC) was also quantified. The SMC and TPC contents in the roots and leaves were 0.01 to 0.71 and 0.03 to 0.57 mg/g of DW, respectively. The concentrations of S-alk(en)yl-l-cysteines in the leaves of domestic sample were significantly greater than those in the root (p < 0.05). The contents of the S-alk(en)yl-l-cysteine sulphoxides (ACSo), alliin, allicin, cycloalliin and methiin were also quantified. The amounts of alliin and allicin in the root sample were 3.66 to 4.02 and 0.30 to 0.35 mg/g of DW, respectively. The alliin and allicin levels ranged from 12.44 to 25.05 and 3.25 to 4.60 mg/g of DW, respectively, in garlic (Ichikawa et al. 2006;Yoo et al. 2010). The level of allicin in the root samples was significantly lower by an approximate factor of 14 when compared with the level in garlic. The level of methiin was 3.99 to 6.39 mg/g of DW.
The roots of the imported samples had the highest methiin levels compared with the levels in the roots and leaves. The cycloalliin content ranged from 1.31 to 11.62 mg/g of DW in hooker chive. In the domestic and imported samples, the levels of cycloalliin in the roots were not statistically different, whereas they were considerably higher in the leaves than in the roots (p < 0.05). The methiin content ranged from 1.3 to 3.0 mg/g of DW in garlic (A. sativum L.), onion (A. cepa L) and leek (Allium porrum L) (Yamazaki et al. 2011). Yamazaki et al. (2011 determined the cycloalliin content in seven Allium vegetables, such as garlic, onion, Chinese chives, rakkyo and leek, to be 0.7 to 4.0 mg/g of DW. Therefore, the amount of methiin and cycloalliin in hooker chive was relatively predominant compared with other Allium vegetables. Several studies have indicated various biological activities of methiin, such as anti-oxidative and anti-diabetic properties and having a hypolipidemic effect (Kumari et al. 1995;Kumari & Augusti 2002;Dini et al. 2008). Cycloalliin has been reported to exhibit biological properties, including fibrinolytic activity, reduction of serum triglycerols and induction of quinine reductase in hepatic cells (Agarwal et al. 1977;Xiao & Parkin 2002;Yanagita et al. 2003). Considering the biological activities of methiin and cycloalliin, hooker chive has promising potential as nutraceutical ingredient.

Identification of volatile organosulphur compounds
Solid-phase micro extraction (SPME) was combined with GC-MS to efficiently investigate the volatile compounds of hooker chive. SPME minimises preparation time, solvent use and equipment requirements. It has been widely used in combination with GC and GC/MS, and has been successfully applied to the extraction of volatile compounds from diverse food sample (Kataoka et al. 2000). Therefore, we identified volatile compounds extracted by SPME. The identification of each volatile compound was confirmed by comparing its mass spectral data and retention time with the Wiley 275&7N mass spectral database. Table S2 lists the volatile compounds, their peak area and rIs on a DB-WAX column identified in the roots and leaves of hooker chive. A total of 42 volatile compounds were identified by SPME, including 21 sulphides, 9 aldehydes, 2 vinyldithiins, 3 ketones and 7 miscellaneous compounds. There were significant differences in the compositions of the volatile compounds between the roots and leaves and between the domestic and imported samples. Aldehydes can be produced by the decarboxylation of amino acids and deamination by mono amine oxidase (Gatfield 1999). The levels of aldehyde, 2-butenal and hexanal in the domestic and imported roots were higher than those in leaves. The area values of 2-methyl-2-butenal and 2-methyl-2-pentanal in the roots were significantly greater than those in the leaves. These compounds can be formed by an aldol condensation and a subsequent dehydration of two molecules of propanol (Boelens et al. 1971). However, 2-methyl butanal, 3-methyl butanal and nonanal could be found in the leaves. The predominant volatile group in the roots and leaves was sulfides, containing 83.1 to 91.2 area%. The main sulphides in the roots were allyl methyl sulphides, such as allyl methylmonosulphide, allyl methyl disulphide, allyl methyl trisulphide and allyl sulphides, including allyl monosulphide, allyl disulphide and allyl trisulphide, whereas these compounds were observed rarely in the leaves. These compounds are related to allicin, which decomposed in the oven of a gas chromatograph, at moderate temperatures. Brodnitz et al. (1971) reported that the allicin is primarily transformed to allyl sulphides, including allyl monosulphide, allyl disulphide and allyltrisulphide. As an intermediate, allyl mercaptan and allyl alcohol were formed during the transformation of allicin (Block 1992). In addition, vinyldithiins are known to be major degradation products of allicin and were generated as artefacts during the gas chromatography analysis of allicin. Therefore, the identification of these compounds in the roots indicated the presence of allicin. Furthermore, these compounds produce a characteristic garlic-like odour. otherwise, dimethyl sulphides, such as dimethyl disulphide and dimethyltrisulphide, and methyl propenyl disulphide isomers were the major volatiles present in the leaves. In particular, the dimethyl sulphides have a characteristic odour commonly described as cabbage and beet like when cooked (Parliment et al. 1997). Additionally, the hydrolysis of the S-methyll-cysteinesulphoxides(ACSos), isoalliin, cycloalliin and methiin, and most of the ACSos resulted in the production of pyruvic acid, ammonia and dimethyl sulphides. In addition, the propenyl-containing sulphides, such as methyl propenyl sulphide isomers and propenyl propylsulphideisomers possess the flavour of cooked onions (Brodnitz et al. 1971). overall, the composition of the volatiles of the imported and domestic samples varied, i.e. the volatiles were influenced by the amount of organosulphur compounds, the cultivation and storage conditions. Therefore, we suggest that our data on volatile compounds may be useful as chemical markers for the quality control of hooker chive.

Antioxidant activities
An antioxidant, which can quench reactive free radicals, can prevent the oxidation of other molecules and may, therefore, have health-promoting effects in the prevention of diseases (Kim et al. 2003). The DPPH radical is one of the most commonly used substrates for the rapid evaluation of antioxidant capacity because of its stability in radical form and the simplicity of the 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical assay. In the DPPH radical assay, the investigated extracts act as hydrogen atoms or electron donors in the transformation of the DPPH radical into its reduced, stable, purple-coloured radical and subsequently, a yellow colour indicating 50% reduction (Bozin et al. 2008). Therefore, we determined the total antioxidant activity as evaluated by DPPH radical scavenging activity.
The ability of the examined hooker chive to scavenge radicals is shown in Table S3. The following IC 50 values were found: 9.59 mg/mL for leaf extracts, 35.11 mg/mL for root extracts from imported sample and 38.54 mg/mL for root extracts from domestic sample. The majority of antioxidant capacities from plants are derived from phenolic compounds. Similar to the results of total phenolic contents, the leaves showed the highest antioxidant capacity.
Ferric reducing antioxidant power assay (FrAP), which provides reliable and reproducible results, measures the ability of an antioxidant to reduce the ferric tripyridyltriazine (Fe 3+ -TPTZ) complex to a blue-coloured ferrous tripyridyltriazine (Fe 2+ -TPTZ) complex by antioxidants and chemical reductants (Benzie & Strain 1996). In this study, FrAP activity of hooker chive obtained is shown in Table S3. Importantly, the activity of leaf extracts (2.36) was significantly higher (p < 0.05) than the activity of root extracts (Cheongju: 0.39 and Myanmar: 0.44).
In addition, a highly significant correlation was observed between the total phenolic contents and antioxidant activities involving DPPH radical scavenging activity (R = 0.998) and FrAP activity (R = 0.999). Although antioxidant capacity significantly correlated with phenolic compounds, this effect did not always relate with the presence of large quantities of phenolics. The antioxidant activities in Allium tissue extracts have been of particular interest, because the thiosulfinate or related organosulphur compounds are primarily responsible for the observed antioxidant effects (Kourounakis & rekka 1991;rekka & Kourounakis 1994;Prasad et al. 1995). In this study, good correlation between antioxidant activity and cycloalliin concentration in hooker chive was observed with a high significance level (R = 0.992). Therefore, the further study regarding the antioxidant functions of organosulphur compounds using various antioxidant activity assays would be performed.

Conclusion
This study is the first to report the comprehensive organosulphur compound profiles of A. hookeri, as obtained by an HPLC-PDA system and GC-MS. our data show that the non-volatile organosulphur compounds of hooker chive were mainly composed of methiin and cycloalliin. The volatile compounds were mainly allyl sulphides and propenyl containing sulphides compound. Additionally, antioxidant activities were evaluated via DPPH radical scavenging assay and FArP activity. As a result, the antioxidant activities of hooker chive were significantly powerful. These results correlated with total phenolic and cycloalliin. However, the antioxidant function of cycloalliin will be researched in this regard.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This research was supported by the Korea Food research Institute [grant number E0143033672].