Comparison of the Essential Oil Composition of Selected Impatiens Species and Its Antioxidant Activities

The present paper describes the chemical composition of the essential oils obtained by hydrodistillation from four Impatiens species, Impatiens glandulifera Royle, I. parviflora DC., I. balsamina L. and I. noli-tangere L. The GC and GC-MS methods resulted in identification of 226 volatile compounds comprising from 61.7%–88.2% of the total amount. The essential oils differed significantly in their composition. Fifteen compounds were shared among the essential oils of all investigated Impatiens species. The majority of these constituents was linalool (0.7%–15.1%), hexanal (0.2%–5.3%) and benzaldehyde (0.1%–10.2%). Moreover, the antioxidant activity of the essential oils was investigated using different methods. The chemical composition of the essential oils and its antioxidant evaluation are reported for the first time from the investigated taxon.


Introduction
The genus Impatiens L. (Balsaminaceae) includes about 850 species, which occur mainly in tropical and subtropical climate zones, in particular in parts of the Old World, such as tropical Africa, India, southern China and the southwestern part of Asia. Some species also occur in Japan, Europe, Russia and North America [1,2]. Three species, Impatiens glandulifera Royle (Himalayan balsam), I. noli-tangere L. (touch-me-not balsam) and I. parviflora DC. (small balsam), occur in Central Europe, and they are perennial plants that grow in riparian zones along rivers on humid soils and in wet woodlands [3,4]. I. glandulifera and I. parviflora are among the invasive plants originally native to Asia that are rapidly spreading across Europe. In Poland, these are two of the top 20 invasive alien plants [2,5]. As I. parviflora is an extremely invasive plant in Europe, its relation with other plants [6] and with soil yeast complexes was recently investigated [7]. Furthermore, the different extracts from I. glandulifera, I. noli-tangere and I. parviflora showed a strong allelopathic effect on the seed germination of Leucosinapis alba and Brassica napus [4].
Because of the rich and varied composition, numerous studies have been conducted to investigate the feasibility of a medicinal use of members of Impatiens. Among the representatives of the genus Impatiens L., some species have been used since a very long time in Asian and American medicine. In traditional therapeutic systems, I. balsamina L. has been the most popular species. Depending on the type of ailment, the dried herb has been used in the form of compresses, directly on the skin, or as a tea prepared by pouring hot water on dried leaves [14]. Moreover, it has been applied in Chinese traditional medicine to treat rheumatism, against fractures, swelling, contusions, beriberi disease and for its anticancer properties [16][17][18]. It has been also used to alleviate parturient and puerperal pain [14]. I. balsamina flowers have been used as a remedy against the effects of snake bites [19]. I. parviflora has been used in the treatment of warts [20]. Flowers of I. glandulifera are used in Bach flower remedies, which causes sedation, relaxes and helps to balance the emotional state, and they are recommended for psychological problems and pain [21]. The rhizomes of Impatiens pritzellii Hook. f. var. hupehensis Hook. f. [22] and whole plant of Impatiens textori MIQ [23] have been also used in Chinese medicine. In our previous study, we confirmed that the extracts from species of Impatiens, especially I. balfourii Hook. f., I. glandulifera and I. parviflora, contained significant amounts of phenols and flavonoids and have interesting multidirectional biological activity, such as antimicrobial and antioxidant abilities [11].
Based on the significance of these plants from an ecological perspective and no available reports on the essential oils of Impatiens species, the aim of the present study was to investigate the essential oil composition of herb and root of the two most invasive in Poland Impatiens species, I. glandulifera Royle and I. parviflora DC. For comparison, herb oils of two other species, I. balsamina L. and I. noli-tangere L., were included. The antioxidant activity of the essential oils of these species was also evaluated.

Chemical Composition of Essential Oils from Impatiens L.
Six Impatiens essential oils were obtained by steam distillation from air-dried herbs and roots. All of them were yellowish and fragrant. The highest yield of essential oil (w/w relative to dry material weight) was observed for the herb of I. parviflora (0.24%) and I. glandulifera (0.22%), while for the other materials, the yield amounted to 0.10%-0.16%. The chemical composition was analyzed by the GC-MS method, which resulted in identification of 54-94 volatile compounds comprising from 61.7%-88.2% of the total volume in individual oils. In total of 226 compounds was identified. All identified compounds in Impatiens oils are listed in Table 1.
The essential oils differed significantly in their chemical composition. What is more, all investigated oils contained specific constituents that distinguished the essential oils from each other. However, some similarities in the qualitative composition can be observed.
In the essential oil of I. parviflora herb, seventy compounds were identified amounting to 88.2%, and among them, (E)-hex-3-en-1-ol (16.8%), linalool (15.1%) and benzaldehyde (10.2%) were the most prominent. The number of identified components in the oil of I. parviflora roots was eighty nine (86.3% of the total oil), and the major compounds were citronellol (17.9%), geranial (12.8%) and linalool (4.9%). Despite significant differences in the content of major constituents, both herb and root oils contained the same aliphatic saturated and unsaturated alcohols, aldehydes and ketones C6-C16, which were their common distinctive features and constituted 42.8% and 26.1%, respectively. Isomeric heptadienals and octadienones that were identified in these oils were only rarely identified in the remaining investigated oils.
The yield of essential oil obtained from I. balsamina herb was 0.1%. Eighty components were identified, representing 84.2% of this oil. The major constituent was hexahydrofarnesyl acetone (13.4%). Pronounced contents of ionones and damascones (15.8%), as well as fatty acids C6-C16 (9.5%) and alkanes (5.9%) were characteristic features of this oil. The main member of the first group was (E)-β-ionone (5.7%), and of the second group, dodecanoic acid (4.1%), ionones and damascones, which occur in a variety of essential oils, are degradation products of carotenoids and have the same C13 carbon skeleton, but differ in the site of oxygenation. (E)-β-Ionone and its epoxide were also found in other investigated oils, however in smaller amounts. Among 54 identified constituents, which accounted for 61.7% of the total essential oil from I. noli-tangere herb, the main compounds were (Z)-hex-3-enol (9.5%), linalool (6.5%) and benzaldehyde (4.7%).
Fifteen compounds were shared among the essential oils of all investigated Impatiens species. The majority of these constituents was linalool (0.7%-15.1%), hexanal (0.2%-5.3%) and benzaldehyde (0.1%-10.2%). Linalool is a naturally occurring monoterpene constituent found in more than 200 oils obtained from herbs, leaves, flowers and wood. This compound has many proven activities and is present in several remedies used in traditional medicine for sedative purposes. Moreover, linalool revealed antimicrobial, anti-inflammatory, antihyperalgesic and analgesic properties [24]. Chang and Shen investigated the cytotoxic activity of linalool. This study suggested good inhibitory effects against breast, colorectal and liver cancer cells [25].
According to the only available report on the composition of Impatiens volatiles, 42 components were characterized in the n-hexane extract of I. bicolor growing in Pakistan. The major ones were fatty acid methyl esters, such as trans-methyl 13-octadecenoate, methyl heptadecanoate, methyl octadecanoate, methyl docosanoate, methyl tetracosanoate, and methyl eicosanoate and aliphatic hydrocarbons [9].
Compounds of these two groups were identified in the essential oils of all investigated Impatiens species. However, their content was significantly lower.
The invasive ability of some vigorous nonnative plants was thought to be associated with the competitive ability of the invasive species or a release from natural enemies. The allelopathic activity of invasive species also has recently been reported as a significant factor that negatively influences species biodiversity and ecosystem succession [26]. Among the allelochemicals, essential oils and their individual components belong to the most investigated [27]. Oxygenated monoterpenes were proven to possess high phytotoxic activity that inhibits the seed germination and seedling growth of many plants [28]. Terpinene-4-ol, which is a minor constituent of each investigated oil, appeared to be the most active of the 47 monoterpenes against Lactuca sativa, and the linalool, citronellol and geranial, major constituents of I. glandulifera and/or I. parviflora, revealed a pronounced phytotoxic effect [29].

Chemometric Analysis
The main constituents common for all tested essential oils (hexanal, heptanal, benzaldehyde, phenylacetaldehyde, nonanal, linalool, terpinen-4-ol, α-terpineol, geranylacetone, β-ionone epoxide, (E)-β-ionone, methyl palmitate, phytol, tricosane, pentacosane) were compared with hierarchical cluster analysis with Euclidean distance as the similarity measure. The so-called "heatmap" with corresponding dendrograms is presented in Figure 1. Benzaldehyde and linalool form distinct cluster with different values than the rest of the analyzed main constituents. However, Euclidean distance does not uncover any distinct cluster among plant material samples, besides a distinct difference of I. parviflora herb (IPH) compared to the other samples.

Chemometric Analysis
The main constituents common for all tested essential oils (hexanal, heptanal, benzaldehyde, phenylacetaldehyde, nonanal, linalool, terpinen-4-ol, α-terpineol, geranylacetone, β-ionone epoxide, (E)-β-ionone, methyl palmitate, phytol, tricosane, pentacosane) were compared with hierarchical cluster analysis with Euclidean distance as the similarity measure. The so-called "heatmap" with corresponding dendrograms is presented in Figure 1. Benzaldehyde and linalool form distinct cluster with different values than the rest of the analyzed main constituents. However, Euclidean distance does not uncover any distinct cluster among plant material samples, besides a distinct difference of I. parviflora herb (IPH) compared to the other samples.  To the compare correlation between constituents (regardless of the absolute concentrations), scaled principal component analysis was carried out. Forty eight-point-four percent and 32.4% of variance was explained by the first two PCs, respectively (Figure 2). Analyzing the loading vectors (Figure 3), it can be concluded that the compounds form three intercorrelated groups: 1.
I. glandulifera roots (IGR) and I. glandulifera herb (IGH) contain the high concentration of group (1), whereas other material samples have smaller concentrations of them, and the differences arelocated mainly along the PC1 axis.
I. glandulifera roots (IGR) and I. glandulifera herb (IGH) contain the high concentration of group (1), whereas other material samples have smaller concentrations of them, and the differences are located mainly along the PC1 axis.

Antioxidant Activity
In the present study, the antioxidant activities of the essential oils from herb and roots of Impatiens species were determined using two different methods. The free radical scavenging activity of essential oils was evaluated by the DPPH method in comparison with that of ascorbic acid, at different concentrations. Radical scavenging activity was expressed as the amount of antioxidants necessary to decrease the initial DPPH absorbance by 50% (EC50). The highest antiradical activity was detected for herb oils of I. glandulifera (3.96 ± 0.03 µg/mL) and I. noli-tangere (4.76 ± 0.05 µg/mL), whereas the lowest was detected for herb oil of I. balsamina ( Table 2). The EC50 value of ascorbic acids to scavenge hydroxyl radicals was 2.05 ± 0.01 µg/mL. Our results are comparable to those obtained by Nisar and co-authors for the hexane extract of I. bicolor. The EC50 values obtained for different fractions ranged from 23.22 ± 0.75 µg/mL-59.00 ± 2.01 µg/mL, while the value for ascorbic acid was 7.80 ± 0.14 µg/mL [9].
The inhibition of linoleic acid peroxidation revealed low capacities of the essential oils of Impatiens species in comparison to the DPPH test ( Table 2). The most active essential oils from herb and roots of I. glandulifera showed even up to six-times higher IC50 values than the lipophilic antioxidant BHT.

Antioxidant Activity
In the present study, the antioxidant activities of the essential oils from herb and roots of Impatiens species were determined using two different methods. The free radical scavenging activity of essential oils was evaluated by the DPPH method in comparison with that of ascorbic acid, at different concentrations. Radical scavenging activity was expressed as the amount of antioxidants necessary to decrease the initial DPPH absorbance by 50% (EC 50 ). The highest antiradical activity was detected for herb oils of I. glandulifera (3.96 ± 0.03 µg/mL) and I. noli-tangere (4.76 ± 0.05 µg/mL), whereas the lowest was detected for herb oil of I. balsamina ( Table 2). The EC 50 value of ascorbic acids to scavenge hydroxyl radicals was 2.05 ± 0.01 µg/mL. Our results are comparable to those obtained by Nisar and co-authors for the hexane extract of I. bicolor. The EC 50 values obtained for different fractions ranged from 23.22 ± 0.75 µg/mL-59.00 ± 2.01 µg/mL, while the value for ascorbic acid was 7.80 ± 0.14 µg/mL [9].
The inhibition of linoleic acid peroxidation revealed low capacities of the essential oils of Impatiens species in comparison to the DPPH test ( Table 2). The most active essential oils from herb and roots of I. glandulifera showed even up to six-times higher IC 50 values than the lipophilic antioxidant BHT.

Reagents and Materials
Aerial parts and roots of plants were collected during July-September 2014. I. balsamina L. All chemical reagents used in the experiment were purchased from various commercial suppliers and were of the highest purity available. 2,2-diphenyl-1-picrylhydrazyl (DPPH), ascorbic acid, 2,6-di-tert-butyl-4-methylphenol (BHT) and linoleic acid (LA) were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Isolation Procedure
The essential oils (EOs) of 100 g of dried herb or roots of Impatiens species were obtained by hydrodistillation for 3 h using the Clevenger-type apparatus, according to European Pharmacopoeia 5.0. Next, the EOs were trapped in 2 mL of freshly-rectified diethyl ether. After distillation, the organic layers were collected and dried over anhydrous magnesium sulfate. After filtration and additional washing of diethyl ether, the solvent was evaporated at room temperature, and the residues were weighed. The oil yields were calculated on the basis of the dry weight of plant material and according to the formula [30]: where W1 is the net weight of oils (g) and W2 is the total weight of dry samples (g).

GC-MS Analysis
The chemical composition of the essential oils was analyzed using GC-MS on a Trace GC Ultra apparatus (Thermo Electron Corporation, Milan, Italy) with FID and the MS DSQ II detector after dilution in diethyl ether (10 µL in 1 mL). A simultaneous GC-FID and MS analysis was performed using an MS-FID splitter (SGE, Analytical Science, Austin, TX, USA). Operating conditions: apolar capillary column Rtx-1ms (Restek), 60 m × 0.25 mm i.d., film thickness 0.25 µm; temperature program, 50-310 • C at 2 • C/min; SSL (splitless) injector temperature 280 • C; FID temperature 300 • C; split ratio 1:20; carrier gas helium at a regular pressure 300 kPa. Mass spectra were acquired over the mass range of 30-400 Da, ionization voltage 70 eV; ion source temperature 200 • C. Identification of the components was based on a comparison of their mass spectra and relative retention indices with data stored in computer libraries NIST 98.1, Wiley 8th Ed. and MassFinder 4.1. Retention indices (RI, apolar column) were determined with relation to a homologous series of alkanes (C8-C26) under the same conditions with linear interpolation. Percentages were calculated from the FID response without the use of correction factors.

Free Radical Scavenging Activity
To determine the antioxidant activity of essential oils from Impatiens species, the method based on the reduction of the methanolic solution of colored free radical DPPH· was used. The changes in color from deep-violet to light-yellow were measured at 515 nm in a UV/visible light spectrophotometer (Thermo Evolution 300, Madison, WI, USA). Radical scavenging activity was measured according to the Brand-Williams et al. [31] method with the use of six dilutions of the analytes in methanol. The activity of ascorbic acid was evaluated for comparison. Antioxidant activity was expressed as EC 50 (efficient concentration): the amount of dry extract (µg of DW) needed to obtain 50% activity per 1.0 mL of the initial solution.

Inhibition of Linoleic Acid Peroxidation
The antioxidant activity was also determined as the degree of inhibition on the hemoglobincatalyzed peroxidation of linoleic acid according to the method described in previous studies [32,33] with a slight modification. The hydroxyperoxide formed was assayed according to the ferric thiocyanate method with mixing with 0.02 M FeCl 2 followed by 30% ammonium thiocyanate. The absorbance of the sample (A s ) was measured at 480 nm. The absorbance of blank (A 0 ) was obtained without hemoglobin to the reaction mixture; the absorbance of the control (A 100 ) was determined without the sample added to the mixture. Thus, the antioxidative activity of the sample was calculated according to the formula: Antioxidant activity was expressed as IC 50 (inhibition concentration): the amount of antioxidant needed to decrease the linoleic acid peroxidation by 50%.

Data Analysis
All measurements were performed at least in triplicate and expressed as the means ± standard deviations (±SD). Statistical significance was estimated through Tukey's test for the data obtained from three independent samples of each essential oil in three parallel experiments (n = 9). Besides the classical pairwise correlation check, we applied the scaled principal component analysis. Statistical tests were performed using Statistica 6.0 software (Stat-Soft, Inc., Tulsa, OK, USA).

Conclusions
It is well understood that invasive species produce specific compounds affecting native plants occurring in the same habitat. This phenomenon is known as allelopathy. Identification of chemical constituents produced and emitted by invasive species helps to understand their impact on the local environment.
Taking into account the chemical composition of I. glandulifera and I. parviflora essential oils and previous data on the allelopathic activity of monoterpenoids, it seemed possible that the emission of monoterpenes by herb and root of these two Impatiens species plays a role in their invasive ability. However, this hypothesis needs further research.