Introduction

Brassica oleracea var. gongylodes L. (Kohlrabi) belongs to the family of Brassicacea. It is a biennial vegetable, with a bulbous stem at the plant basis [1]. Stems of kohlrabi may be green or purple red with flesh that is always white and it is harvested in northwestern areas of Europe and is consumed in North America, India, China, Vietnam, Korea, Europe, Mediterranean, and some zones of Asia. Kohlrabi is popular as a diet food thanks to its high fiber, cellulose and pectin amounts; it is also popular for its nutritional values with high levels of potassium and vitamin C and low amounts of fat [2]. The stem part of this plant is commonly utilized as a cooked vegetable, but the raw grated stem forms a component of winter salads [3]. In Tunisia, the green kohlrabi stem is consumed as a cooked or pickled vegetable. This species contains a considerable amount of phenolic compounds, which support its antioxidant, anticarcinogenic and anti-hyperglycemic potentials [4,5,6]. Park et al. [7] demonstrated that the total glucosinolates and phenylpropanoids levels in the flesh of green Kohlrabi were higher than those in the skin. Consumption of kohlrabi aids people to be healthy by reducing free radicals as it provides high concentrations of vitamin C and protects against prostate and colon cancer by providing the body with important phytochemicals such as isothiocyanates and indole-3-carbinol [8].

However, researches focusing on the bioactivities, metabolic profile and nutritional value of B. oleracea var. gongylodes are limited, and only a few have been published. To our knowledge, no previous report has established phytochemical profiling and antioxidant activity of leaves, flesh and peel of B. oleracea var. gongylodes growing in Tunisia. Therefore, this comparative study focused on profiling and quantifying phenolic compounds, dietary fibers, chlorophylls, minerals, proteins, soluble sugars and fatty acids of leaves, flesh and peel of B. oleracea var. gongylodes. Furthermore, in vitro antioxidant and antibacterial properties of aqueous extracts of B. oleracea var. gongylodes morphological parts were evaluated and compared to the corresponding phytochemical profile.

Materials and methods

Plant material and extraction

Brassica oleracea var. gongylodes was collected in December 2017 (in a vegetative stage) from Menzel Bourguiba area in northern of Tunisia and was botanically identified by the botanist Fethia Skhiri (High Institute of Biotechnology, University of Monastir, Tunisia). The Plant was separated into leaves, flesh and peel and air-dried in the shadow for 15 days. Voucher specimen (number 12/2017/BOG) was deposited at the Herbarium of the Laboratory of Botany, Faculty of Pharmacy, University of Monastir, Tunisia.

An aqueous extract from leaves, flesh and peel was prepared by putting dried material (10 g) in boiling distilled water (100 mL) [9]. The mixture was boiled for 30 min and then filtered. The obtained extract was lyophilized and the different extract yields were determined.

Phenolic content determination

Total phenolic content was measured by a Folin-Ciocalteu method [10] and results are reported as mg of gallic acid equivalent per 100 g of dry weight (mg GAE/100 g DW). Total flavonoid and flavonol contents of extracts were evaluated by the aluminum chloride method [11] and were presented as mg catechin equivalent per 100 g of dry weight (mg CE/100 g DW). Total tannins were evaluated by the method of Julkunen-Titto [12] and tannin content is reported as mg catechin equivalent per 100 g of dry weight (mg CE/100 g DW). Analysis was done in three repetitions.

Phenolic compounds identification and ascorbic acid content determination by high performance liquid chromatography (HPLC)

The identification of phenolic compounds in aqueous extracts from leaves, flesh and peel was carried out using HPLC system as described by Bettaieb Rebey et al. [13]. Phenols were identified according to their retention times and the spectral characteristics of their peaks against those of standards and by spiking the sample with standards.

Ascorbic acid content was determined using HPLC system, according to the method described by Lo Scalzo et al. [14]. Thus, ascorbic acid concentration was calculated from the experimental peak area by analytical interpolation in a standard calibration curve, and was expressed as mg per 100 g of sample dry weight (mg/100 g DW).

Proximate composition and mineral analysis

Samples were analyzed for content of water (WC), crude protein (CP), nitrogen (N), neutral detergent fibers (NDF), acid detergent fibers (ADF) and acid detergent lignin (ADL). WC was determined after drying the sample at 45 °C to a constant weight and results were expressed as percentage (%) of water content. The Kjeldahl method [15] was used to determine nitrogen content and CP was calculated as N × 6.25. The contents of NDF, ADF and ADL were determined according to the methods of Van Soest et al. [16] using an ANKOM 220 Fiber Analyzer (ANKOM Technology Corporation, NY, USA). Hemicellulose was calculated as NDF–ADF and cellulose as ADF–ADL [17]. Mineral content was determined using the official analytical methods [18]. Phosphorus (P) was determined by molybdovanadophosphoric acid method described by Kitson and Mellon [19].

Chlorophyll content determination

Chlorophylls were extracted with 80% acetone (v/v) [20]. Chlorophyll a, chlorophyll b and total chlorophyll concentrations were determined according to the equations of Arnon [20].

Soluble sugar profile determination

Soluble carbohydrates of B. oleracea var gongylodes leaves, flesh and peel were extracted according to the method described by Bartolozzi et al. [21]. Identification of individual soluble sugars was carried out using a gas chromatograph equipped with a flame ionization detector (FID) and a HP-5MS column (30 m × 0.25 mm), and was achieved by mean of relative retention times, in comparison to that of standards [21]. The contents of soluble sugars were expressed as g per 100 g of fresh weight (FW) (g/100 g FW).

Fatty acid analysis by gas chromatography

Fatty acids were transformed into their methyl esters as described by Dhibi et al. [22]. Fatty acid methyl esters (FAMEs) were separated and quantified by gas chromatography, according to their percentage area, obtained by integration of the peaks and were identified by comparing their retention times with respect to pure standard FAMEs and analyzed under the same conditions [22]. Results were expressed as a percentage of individual fatty acid in the lipid fraction [22].

Antioxidant activity assays

Three antioxidant activity assays were performed for different extracts: ABTS and DPPH radical scavenging and reducing power assays. In ABTS assay [23], an aqueous solution of ABTS (7 mM) was mixed with a potassium persulfate solution (2.45 mM). Then, 990 µL of the ABTS radical solution (at absorbance of ≈ 0.70 at 734 nm) was added to 10 µL of extract. The absorbance was determined at 734 nm. In DPPH assay [24], DPPH (950 µL, 10−4 M) was mixed with diluted extract (50 μL). After 30 min, the mixture absorbance was read at 515 nm. The anti-radical activity was determined by the formula: % inhibition of radical ABTS or DPPH = [(Acontrol – Asample)/Acontrol] × 100; Acontrol is control reaction absorbance, and Asample is sample test absorbance. Reducing power was evaluated by the capacity to convert Fe3+ into Fe2+ [25], measuring the absorbance at 690 nm. In all tests, the results were presented as EC50 values (sample concentration providing 50% of antiradical activity or 0.5 of absorbance in the reducing power assay) in mg/mL of extract. Trolox was used as a positive control and absorbance was measured in three replications.

Antibacterial activity

The antibacterial activity of aqueous extracts from leaves, flesh and peel of green kohlrabi, was determined using the agar disk diffusion and micro-dilution methods adopted by Hlila et al. [26], to determine inhibition zone diameters (IZD in mm), Minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC). Activity was tested against Gram positive: Staphylococcus aureus ATCC 6538, Bacillus subtilis ATCC 6633, Bacillus anthracis (clinical strain) and Gram negative: Escherichia coli ATCC 8739, Pseudomonas aeruginosa ATCC 9027 and Salmonella enterica CIP 8039 bacteria. These strains were provided by the Laboratory of Transmissible Diseases and Biological Active Substances (Faculty of Pharmacy of Monastir, Tunisia). Gentamicin was used as a positive reference standard. All tests were performed in triplicate.

Statistical analysis

The means and standard deviation of data were calculated from independent experiments. Data analysis was carried out using one-way ANOVA analyses followed by multi-range post hoc of Duncan’s test and (p < 0.05) or less was considered significant.

Results and discussion

Yields and phenolic contents

The yield percentages of aqueous extractions from Brassica oleracea var. gongylodes leaves, flesh and peel are shown in Table 1. Aqueous extract from leaves (33.4%; p < 0.05) presented the significantly highest yield, followed by peel (27.52%) and flesh (27.48%) extracts. As shown in Table 1, phenol, flavonoid, flavonol and tannin contents varied significantly with the organ type. Among all samples, peel aqueous extract had significantly the highest levels of total phenols (6736.89 mg GAE/100 g DW; p < 0.05), flavonoids (1288,91 mg CE/100 g DW; p > 0.05), flavonols (504,18 mg CE/100 g DW; p > 0.05) and tannins (363,99 mg CE/100 g DW; p > 0.05).

Table 1 Yields, total phenol, flavonoid, flavonol and tannin contents and antioxidant activity (EC50 mg/mL) of aqueous extracts of B. oleraceae var gongylodes leaves, flesh and peel
Full size table

These results that demonstrated that the peel and leaves of B. oleracea var. gongylodes were richer in phenolic compounds than flesh are in accordance with these reported by Pak et al. [27]. According to Chauhan et al. [28], phytochemicals such as flavonoids and tannins are present in distilled water and methanol extracts of B. oleracea var. gongylodes leaves. Consequently, these results have demonstrated that water is a good solvent in extracting polyphenols and that peel is significantly richer on phenols, flavonoids, flavonols and tannins than flesh and leaves. Secondary metabolites are accumulated in all plant cells but their concentration varies according to the plant parts, particular growth phases, climates and seasons.

Phenolic profile and ascorbic acid content

The aqueous extracts of B. oleracea var. gongylodes parts were analyzed using HPLC method and results are presented in Table 2. To the best of our knowledge this is the first study about the phenolic compounds in aqueous extracts of the studied parts of green kohlrabi grown in Tunisia, performed by HPLC technique. In leaves aqueous extract, one hydroxycinnamic acid (chlorogenic acid) and three flavonoids (catechol, epigallocatechin and epicatechin 3-O-gallate) were identified. Also, catechol and isorhamnetin-3-O-glucoside were detected in flesh and peel, but, gallic acid was present only in flesh with a level of 31.74 mg/100 g DW. Catechol was significantly more concentrated in leaves (89.86 mg/100 g DW; p < 0.05) compared to peel (59.99 mg/100 g DW) and flesh (52.53 mg/100 g DW) (Table 2).

Table 2 Phenolic and ascorbic acid composition (mg/100 g DW) of aqueous extracts of B. oleracea var. gongylodes leaves, flesh and peel, using the high performance liquid chromatography
Full size table

The most diverse and widespread polyphenols group in Brassica species are hydroxycinnamic acids and flavonoids [3]. The major common hydroxycinnamic acids in Brassica vegetables are ferulic, p-coumaric and sinapic acids, frequently found in conjugation with other hydroxycinnamic acids or sugar [29]. A study reported the presence of 4-hydroxybenzoic, caffeic, p-coumaric, benzoic and trans-cinnamic acids, quercetin, and kaempferol in the skin and flesh of B. oleracea var. gongylodes [7]. These reports are inconsistent with our study results, where we noted the absence of p-coumaric acid in aqueous extracts from different parts of kohlrabi. Leaves and peel contained chlorogenic acid with the highest amount in leaves aqueous extract (106.362 mg/100 g DW; p < 0.05). Significant amounts of chlorogenic acids were reported in leafy Brassica species, like cabbage, kale and Brussels sprouts [30].

Only aqueous extracts from flesh and peel contained flavonol, isorhamnetin-3-O-glucoside with the greatest amount in the flesh. Isorhamnetin-3-O-glucoside is identified here for the first time in B. oleracea var. gongylodes flesh and peel, in disagreement with previous reports in Brassica species [31]. These reports mentioned that isorhamnetin derivatives were absent in B. oleracea and present in B. rapa group (Romani et al. 2006). Nevertheless, isorhamnetin, quercetin and kaempferol are the main flavonols in Brassica crops and are frequently found as O-glycosides [32, 33]. The quantification of isorhamnetin is recommended for the differentiation of varieties [34], as well as for investigations on cooking conditions [35].

The peel ascorbic acid concentration (233.15 mg/100 g DW; p < 0.05) was significantly higher than that of leaves (212.741 mg/100 g DW) and flesh (118.5 mg/100 g DW) extracts (Table 2). These results are consistent with data published by Cha et al. [36], who reported that peel was richer in vitamin C (402.74 mg/100 g) than flesh (231.36 mg/100 g). Thus, vitamin C can be found reduced (ascorbic acid) or oxidized (dehydroascorbic acid); the sum of these is total vitamin C and the variation of acid ascorbic and vitamin C probably arises from different factors such as variations in post-harvest handling conditions and seasonal cultivar selection. Vitamin C is essential for collagen and carnitine biosynthesis, and has antioxidant and anticarcinogenic activities [37].

Proximate, dietary fiber, chlorophyll and mineral composition

The proximate, dietary fiber, chlorophyll and mineral composition values are shown in Table 3. Water content (87.28%; p < 0.05), crude protein (18.69%; p < 0.05) and nitrogen (2.99%; p < 0.05) were significantly higher in leaves than in flesh and peel. In this context, Cha et al. [36] demonstrated that the crude protein content of flesh and peel of B. oleracea var. gongylodes were 16.63 and 20.45% of dry weight, respectively and that moisture varied from 5.58% (peel) to 16.52% (flesh).

Table 3 Proximate, dietary fiber, chlorophyll and mineral composition of leaves, flesh and peel of B. oleracea var. gongylodes
Full size table

All plant parts contained similar content of NDF (36.86–38.35%; p < 0.05), but ADF was more concentrated in flesh (31.53%) and peel (30.92%). Flesh was richer in ADL (14.88%; p < 0.05) than peel and leaves. A significantly high amount of hemicellulose (11.49%; p < 0.05) was present in leaves. Cellulose was elevated in peel (22.02%; p < 0.05).

No reports on the dietary fiber composition of leaves of B. oleracea var. gongylodes have been presented. Nevertheless, Cha et al. [36] reported that peel of kohlrabi is richer in total dietary fiber (19.58%) than flesh (1.44%). NDF had the highest share in dietary fiber, followed by ADF, which includes lignin and cellulose. Cellulose fibers are virtually undigested in the gastrointestinal tract; however, they aid intestine peristalsis, as similar as ADL [38]. Thus, dietary fibers have advantageous effects on body function and human health and can only be provided by plants. Indeed, the consumption of dietary fibers is related to a reduced incidence of frequent disorders such as obesity, diabetes, cancer and cardiovascular diseases [39].

The mineral elements are expressed in mg/100 g DW and their contents are given in Table 3. The leaves, flesh and peel proved to be a source of potassium (1410.34–1660.21 mg/100 g DW; respectively) and sodium (37.0–100.0 mg/100 g DW). Comparing potassium contents, the flesh showed the highest amount (1660.21 mg/100 g DW; p < 0.05). On the contrary, leaves exhibited greater levels of sodium (100.0 mg/100 g DW; p < 0.05) and phosphorus (80.0 mg/100 g DW) than flesh and peel. In terms of microelements, Fe, Mn, Zn, Ni, Cu, Cr and Co were detected in the leaves, flesh and peel of Kohlrabi (Table 3). Fe is the most abundant microelement and cobalt is the least (Table 3).

No work on the mineral composition of green kohlrabi leaves has been published. However, Cha et al. [36] mentioned that peel contained an elevated amount of sodium (115.6 mg/100 g DW) and that flesh and peel of Kohlrabi were richer in potassium and contained 366.0 and 4440.0 mg/100 g of K, respectively. Also, Cha et al. [36] reported the presence of Cu and Mn in Kohlrabi peel, but its absence in flesh sample. Briefly, the same authors detected eight minerals (K, Na, Ca, Mg, Fe, Cu, Mn and Zn) in peel and six minerals (K, Na, Ca, Mg, Fe and Zn) in the flesh of B. oleracea var. gongylodes cultivated in Korea [36]. Our present study identified four minerals (P, Ni, Cr and Co) which have not yet been reported in the literature for this plant material.

The results of the present investigation of chlorophyll a, chlorophyll b and total chlorophyll contents are shown in Table 3. Chlorophyll a (89.1 mg/100 g), chlorophyll b (30.23 mg/100 g) and total chlorophyll (113.14 mg/100 g) concentrations of leaves were significantly higher (p < 0.05) than those of the peel and flesh of B. oleracea var. gongylodes.

Kohlrabi leaves contained a significantly elevated level of chlorophylls such as a majority of leafy vegetables. Chlorophylls are green pigment photoreceptors present in all photosynthetic organisms and absorb solar light energy and provide mechanisms for its utilization in photosynthetic reactions. Chlorophyll pigments are classified among the least stable natural pigments. Some researchers suggest that these pigments contribute to antioxidant, antiviral and antimutagenic activities [40].

Soluble sugars composition

The sugar composition of different parts of B. oleracea var. gongylodes was presented in Table 4. Leaves were poor in sugar (0.828 g/100 g DW; p < 0.05) compared with flesh (1.934 g/100 g DW) and peel (1.983 g/100 g DW). Sucrose (1.001–1.189 g/100 g DW) was the main sugar in flesh and peel, but arabinose (0.001–0.003 g/100 g DW) was the least in all raw materials.

Table 4 Soluble sugars composition (g/100 g DW) of leaves, flesh and peel of B. oleracea var. gongylodes
Full size table

The total sugar content (0.828–1.983 g/100 g DW) reported here was lower than that quantified by Cha et al. [36] in flesh and peel of Kohlrabi from Korea (314.94–419.15 g/100 g). These authors demonstrated that the main sugars presented in the flesh and peel were fructose (152.23 g/100 g) and glucose (182.23 g/100 g), respectively [36]. In the present study, six sugars (sucrose, inositol, mannitol, raffinose, rhamnose and arabinose) were identified for the first time in B. oleracea var. gongylodes cultivated in Tunisia. Sugars are important chemical compounds with nutritional value for humans and constitute the main energy source in vegetarian diets.

Fatty acids composition

The fatty acids analysis of the studied parts of green kohlrabi indicated the presence of 17 components (Table 5). Linoleic acid was the major fatty acid with a concentration of 24.66% in peel, 23.07% in leaves and 22.00% in flesh, followed by palmitic (20.89–24.49%), and oleic (18.68–20.98%) acids. Nevertheless, the relative percentages of each quantified fatty acid always showed statistically significant differences between both plant parts. The unsaturated fatty acids were predominant (51.67–57.96%) compared to saturated fatty acids (42.04–48.33%). On the contrary, it was reported that the fatty acid profile of kohlrabi was dominated by saturated fatty acids in peel (82.2%) and flesh (72.99%) [36]. Furthermore, the same authors reported that palmitic acid was the dominant fatty acid (36.04%) in peel, followed by heneicosanoic acid (11.85%). However, the most abundant fatty acid in flesh was heneicosanoic acid (21.71%), followed by arachidic acid (18.65%) and linoleic acid (13.56%) [36]. Nevertheless, caproic, lauric, pentadecylic, margiric, eicosenoic and tricosylic acids were identified here for the first time in green kohlrabi grown in Tunisia.

Table 5 Fatty acids composition (relative percentage) of B. oleracea var. gongylodes leaves, flesh and peel
Full size table

Antioxidant activity

The antioxidant activity of the tested samples was determined using the DPPH, ABTS and reducing power assays and was compared to that of Trolox. The aqueous extracts of different plant parts exhibited remarkable antioxidant activities (Table 1). The best effect was observed in the aqueous extract of leaves with EC50 of 0.26, 0.32 and 0.46 mg/mL, by DPPH, ABTS and reducing power assays, respectively. This might be explained by the high level of hydrogen-donating constituents in the leaves extract, especially phenolic compounds. Besides, these activities are slightly lower than a Trolox with EC50 ranging from 0.072 to 0.153 mg/mL.

Aqueous extracts from leaves were more active than those from flesh and peel. Our results are in agreement with those reported by Pak et al. [27], who demonstrated that water and ethanol extracts of leaves were more active than those of flesh and peel, by DPPH and reducing power tests. Hassan et al. [6] proved the existence of a considerable amount of phenolic compounds in B. caulorapa var. gongylodes that is supporting its potential role as antioxidant and anti-carcinogenic agent. B. oleracea var. gongylodes juices have been reported to exhibit potent antioxidant activity conferred by high total phenolic content [5]. Chun et al. [41] demonstrated the high correlation between antioxidant potential and phenol and flavonoid contents for different varieties of B. oleracea (green cabbage, red cabbage, Savoy cabbage and Napa cabbage) using an ABTS radical scavenging assay.

In fact, our results indicate that aqueous extracts of B. oleracea var. gongylodes leaves, flesh and peel, which are rich in phenolics and flavonoids, have an antioxidant activity. Generally, high antioxidant activity is probably due to the combined action of the polyphenol compounds present in variable levels and their high hydrogen atom donating abilities. Moreover, from the present study results, it is evident that the antioxidant activities of B. oleracea var. gongylodes, are associated to various phenolic compounds such as chlorogenic, gallic and ascorbic acids and catechol, epigallocatechin, epicatechin 3-O-gallate and isorhamnetin-3-O-glucoside present in different plant parts.

Antibacterial activity

Antibacterial activity of aqueous extracts of B. oleracea var. gongylodes parts was assessed in vitro against a panel of Gram+ and Gram− bacteria. According to the results presented in Table 6, the samples exhibited an antibacterial activity against some tested bacteria. Staphylococcus aureus appeared the most sensitive towards extracts (IZD range from 6.16 to 11.33 mm, and MIC ranged from 0.625 to 5 mg/mL). Additionally, the leaves extract was the most active against S. aureus (MIC = 0.625 mg/mL and MBC = 2.5 mg/mL), but this antibacterial activity was lower than that of Gentamicin (Table 6).

Table 6 Antibacterial activity (IZD; MIC and MBC) of B. oleracea var. gongylodes leaves, flesh and peel aqueous extracts
Full size table

According to literature investigation, no reports have been found out on the antimicrobial activity of B. oleracea var gongylodes. Nevertheless, it is well known that Brassica vegetables have been shown to possess antimicrobial properties [42]. It has been reported that cauliflower juice inhibits the growth of E. coli, Salmonella enteritidis, Listeria monocytogenes and all foodborne pathogens found in a great variety of foods [42]. Most phenolic compounds are biologically active constituents of plant origin having antibacterial, antiviral and anticancerous properties. In our work, we have found interesting results regarding the antibacterial potential of the aqueous extracts of B. oleracea var gongylodes and its phenolic analysis has demonstrated the phenols and flavonoids richness. Thus, phenolic compounds present in the different plant parts of B. oleracea var gongylodes can contribute to its antibacterial activity.

Conclusion

The aqueous extracts of leaves, flesh and peel of B. oleracea var. gongylodes cultivated in Tunisia showed statistically significant differences in the total phenols, flavonoids, flavonols and tannins contents. These compounds were more concentrated in the aqueous extract from peel. The leaves contained phenolic acid (chlorogenic acid) and flavonoids (catechol, epigallocatechin and epicathechin-3-O-gallate). Accordingly, we suggest using water as the extractor to isolate phenolic acid and flavonoid constituents from green kohlrabi leaves. Flavonol, isorhamnetin-3-O-glucoside was present only in flesh and peel. The peel and leaves were richer in ascorbic acid than flesh, which is a frequently consumed part of kohlrabi. Water, crude protein, nitrogen, hemicellulose, chlorophylls, sodium, phosphorus and microelements (Fe, Zn, Mn and Ni) were more condensed in leaves. However, flesh was the richest in potassium. Flesh and peel contained equal content of soluble sugars and the sucrose, which is the main sugar, was more concentrated in peel. Thus, peel had the highest linoleic and palmitic acids level and oleic acid amount was elevated in leaves. Regarding the bioactivities and summarized data recorded by the three antioxidant assays, it appears that the aqueous extract of leaves was the most active. Different plant parts exhibited antibacterial activity against some studied bacteria and leaves aqueous extract was the most active against S. aureus. Finally, these results suggest that the flesh, and specially the peel and leaves of B. oleracea var. gongylodes, are a potential source of high value components for nutritional and pharmaceutical purposes, as well as functional foods and nutraceutical applications.