Microwave-Assisted versus Conventional Isolation of Glucosinolate Degradation Products from Lunaria annua L. and Their Cytotoxic Activity

Glucosinolates (GSLs) from Lunaria annua L. seeds were analyzed qualitatively and quantitatively by their desulfo counterparts using UHPLC-DAD-MS/MS technique and by their volatile breakdown products, isothiocyanates (ITCs), using GC-MS technique. GSL breakdown products were obtained by conventional techniques (hydrodistillation in a Clevenger type apparatus (HD), CH2Cl2 extraction after myrosinase hydrolysis (EXT) for 24 h) as well as by modern techniques, microwave-assisted distillation (MAD) and microwave hydrodiffusion and gravity (MHG). Seven GSLs were identified as follows: isopropyl GSL (1), sec-butyl GSL (2), 5-(methylsulfinyl)pentyl GSL (3), 6-(methylsulfinyl)hexyl GSL (4), 5-(methylsulfanyl)pentyl GSL (5), 6-(methylsulfanyl)hexyl GSL (6), and benzyl GSL (7). Additionally, pent-4-enyl- and hex-5-enyl ITCs were detected in the volatile extracts. However, their corresponding GSLs were not detected using UHPLC-DAD-MS/MS. Thus, they are suggested to be formed during GC-MS analysis via thermolysis of 5-(methylsulfinyl)pentyl- and 6-(methylsulfinyl)hexyl ITCs, respectively. Volatile isolates were tested for their cytotoxic activity using MTT assay. EXT and MHG showed the best cytotoxic activity against human lung cancer cell line A549 during an incubation time of 72 h (IC50 18.8, and 33.5 μg/mL, respectively), and against breast cancer cell line MDA-MB-231 after 48 h (IC50 6.0 and 11.8 μg/mL, respectively). These activities can be attributed to the ITCs originating from 3 and 4.


Introduction
Lunaria annua L. (annual honesty) belongs to the Brassicaceae family, which is, without exception, characterized by the presence of glucosinolates (GSLs) [1]. Native to Southeast Europe and Southwest Asia, L. annua is a biennial plant which flowers in mid-summer, developing its distinctive silver-dollar sized silicles that become translucent with maturity.
From the beginning, seeds of L. annua have served as a source of isopropyl isothiocyanate (ITC), originating from glucoputranjivin (1) degradation. Kjaer (1959) first isolated 1 in the form of its crystalline tetraacetate [2]. In addition to 1, using thin-layer chromatography, Danielak and Borkowski (1969) also

HPLC-DAD Analysis of Desulfoglucosinolates
Analysis was performed on HPLC-DAD (Ultimate 3000, Thermo Fischer Scientific, USA) using Hypersil GOLD column (5 µm, 250 mm × 4.0 mm, Thermo Fischer Scientific, USA). Each extract was analyzed twice. A gradient consisting of solvent A (H 2 O) and solvent B (acetonitrile:H 2 O 30:70 v/v) was applied at a flow rate of 0.8 mL/min as follows: 0.14 min 96% A and 4% B, 28.14 min 14% A and 86% B, 32.14 min 14% A and 86% B, 34.14 min 5% A and 95% B, 47.14 min 5% A and 95% B, 48.14 min 96% A and 4% B, and 56.14 min 96% A and 4% B. The column temperature was held at 30 • C and the injection volume was 20 µL. For confirmation of the peaks from HPLC-DAD, individual peaks were subjected to UHPLC-MS/MS analysis.
The amount of GSLs was quantified using a calibration curve of pure desulfosinigrin solution (range from 0.14 to 1.4 mM) and RPFs for each individual dGSL [13]. RPF values for quantification of dGSLs were as following: RPF 1.0 for 1, 4, and 6 [14]; RPF 1.07 and 0.95 for 3 and 7, respectively [15]; and arbitrary RPF 1.0 for 2 and 5.

UHPLC-MS/MS Analysis
Analysis was performed on UHPLC-MS/MS (Ultimate 3000RS with TSQ Quantis MS/MS detector, Thermo Fischer Scientific, USA) using Hypersil GOLD column (3.0 µm, 3.0 × 100 mm, Thermo Fischer Scientific, USA). A gradient consisting of solvent A (50 µM NaCl in H 2 O) and solvent B (acetonitrile/H 2 O 30:70 v/v) was applied at a flow rate of 0.5 mL/min as follows: 0.14 min 96% A and 4% B, 7.84 min 14% A and 86% B, 8.96 min 14% A and 86% B;,9.52 min 5% A and 95% B, 13.16 min 5% A and 95% B, 13.44 min 96% A and 4% B, and 15.68 min 96% A and 4% B. The column temperature was held at 40 • C and the injection volume was 5 µL. The electrospray interface was H-ESI source operating with a capillary voltage of 3.5 kV at 350 • C. The system was operated in the positive ion electrospray mode.

Conventional Isolation
The volatiles were isolated by the following two approaches: Hydrodistillation in Clevenger type apparatus for 2.5 h using 50 g of ground seeds (HD), and CH 2 Cl 2 extraction after endogenous and exogenous hydrolysis by myrosinase (1 to 2 units) for 24 h at 27 • C using 10 g of ground seeds (EXT), as described previously [12,16].

Microwave-Assisted Isolation
Seeds were hydrated for 2 h prior to the isolation process. A Milestone 'ETHOS X' microwave laboratory oven (1900 W maximum) was used for microwave-assisted isolation. This is a multimode microwave reactor of 2.45 GHz. Temperature is monitored by an external infrared sensor.
Microwave-assisted distillation (MAD): A typical experiment is conducted at atmospheric pressure with 100 g of matrix, during 35 min at 500 W (98 • C). The distillation process started after 10 min. The distillate was collected in a side-tube using pentane trap, dried over anhydrous sodium sulphate, and stored at −20 • C, until analysis ( Figure 1a).
Microwave hydrodiffusion and gravity (MHG): A typical experiment is conducted at atmospheric pressure with 100 g of matrix, during 15 min at 500 W (98 • C). The water extract was collected and extracted by CH 2 Cl 2 , dried over anhydrous sodium sulphate and concentrated using an automated sample concentrator (VLM GmbH, Germany) to a volume of 1 mL. The sample was stored at −20 • C, until analysis (Figure 1b).

GC-MS Analysis
All isolates were analyzed by GC-MS (model 3900-2100T; Varian Inc., Lake Forest, CA, USA) using a VF-5MS column (30 m × 0.25 mm i.d., coating thickness 0.25 μm) programmed at 60 °C isothermal for 3 min, increased to 246 °C at a rate of 3 °C/min and held isothermal for 25 min, as previously described [12]. The analyses were carried out in duplicate.
Individual peaks of volatiles were identified by comparing their retention indices and mass spectra with those of authentic samples, as well as by computer matching against the Wiley 7 spectral database and comparison of the mass spectra with literature data [17][18][19]. The analyses were run in duplicate and the percentages in Table 2 were calculated as the mean value of component percentages on a VF-5MS column.

Cell Viability Assay (MTT)
MTT spectrophotometric assay was performed on a microplate photometer, model HiPo MPP-96 (BioSan, Riga, Latvia), as previously described [10,12]. The cells (A549 and MDA-MB-231) were treated with L. annua volatile isolates (HD, EXT and MHG) at concentrations of 1, 5, 10, 50, and 100 µ g/mL in a complete medium (in triplicate) for 4, 24, 48, and 72 h. In order to determine IC50 more accurately, both cell lines were treated with volatile isolates at additional concentrations. The A549 cell line was treated with 20, 30, and 40 µ g/mL concentrations of EXT and MHG. The MDA-MB-231 cell line was treated with 2.5, 7.5, and 30 μg/mL concentrations of EXT and with 20, 30, and 40 µ g/mL concentrations of MHG. After treatment with the isolated compounds, the cells were incubated with 0.5 g MTT/L at 37 °C for 2 h, followed by removal of the medium. Dimethyl sulfoxide (DMSO) was added and incubated for another 10 min at 37 °C while shaking. The degree of formazan formation, an indicator of living and metabolically active cells, was measured at 570 nm. The data were calculated in relation to the untreated control (100%) from three independent measurements. The calculation of IC50 values was performed using GraphPad Prism software version 7.0. The criteria used to categorize the activity against the tested cell lines was based on IC50 values as follows: 20

GC-MS Analysis
All isolates were analyzed by GC-MS (model 3900-2100T; Varian Inc., Lake Forest, CA, USA) using a VF-5MS column (30 m × 0.25 mm i.d., coating thickness 0.25 µm) programmed at 60 • C isothermal for 3 min, increased to 246 • C at a rate of 3 • C/min and held isothermal for 25 min, as previously described [12]. The analyses were carried out in duplicate.
Individual peaks of volatiles were identified by comparing their retention indices and mass spectra with those of authentic samples, as well as by computer matching against the Wiley 7 spectral database and comparison of the mass spectra with literature data [17][18][19]. The analyses were run in duplicate and the percentages in Table 2 were calculated as the mean value of component percentages on a VF-5MS column.

Cell Viability Assay (MTT)
MTT spectrophotometric assay was performed on a microplate photometer, model HiPo MPP-96 (BioSan, Riga, Latvia), as previously described [10,12]. The cells (A549 and MDA-MB-231) were treated with L. annua volatile isolates (HD, EXT and MHG) at concentrations of 1, 5, 10, 50, and 100 µg/mL in a complete medium (in triplicate) for 4, 24, 48, and 72 h. In order to determine IC 50 more accurately, both cell lines were treated with volatile isolates at additional concentrations. The A549 cell line was treated with 20, 30, and 40 µg/mL concentrations of EXT and MHG. The MDA-MB-231 cell line was treated with 2.5, 7.5, and 30 µg/mL concentrations of EXT and with 20, 30, and 40 µg/mL concentrations of MHG. After treatment with the isolated compounds, the cells were incubated with 0.5 g MTT/L at 37 • C for 2 h, followed by removal of the medium. Dimethyl sulfoxide (DMSO) was added and incubated for another 10 min at 37 • C while shaking. The degree of formazan formation, an indicator of living and metabolically active cells, was measured at 570 nm. The data were calculated in relation to the untreated control (100%) from three independent measurements. The calculation of IC 50 values was performed using GraphPad Prism software version 7.0. The criteria used to categorize the activity against the tested cell lines was based on IC 50 values as follows: 20 µg/mL = highly active, 21 to 200 µg/mL = moderately active, 201 to 500 µg/mL = weakly active, and >501 µg/mL = inactive [20].

Glucosinolates and Volatile Constituents
GSLs of Lunaria annua seed were qualitatively analyzed by UHPLC-MS/MS and quantified by HPLC-DAD of their desulfo counterparts (Table 1, Figure 2, Figures S1 and S2). GSLs were subjected to enzymatic and thermal degradation. Enzymatic degradation was performed by endogenous and exogenous myrosinase, whereas thermal degradation was assessed using conventional (hydrodistillation) and up-to-date (MAD and MHG) techniques. Isothiocyanates (ITCs), the main volatiles that originate from degradation of GSLs, were identified by GC-MS. (Table 2). [M + Na] + , sodium adducts of desulfoglucosinolate; DW, dry weight; tr < 0.1 µmol/g DW, value is the mean ± standard error (n = 2).

Glucosinolates and Volatile Constituents
GSLs of Lunaria annua seed were qualitatively analyzed by UHPLC-MS/MS and quantified by HPLC-DAD of their desulfo counterparts (Table 1, Figures 2, S1, and S2). GSLs were subjected to enzymatic and thermal degradation. Enzymatic degradation was performed by endogenous and exogenous myrosinase, whereas thermal degradation was assessed using conventional (hydrodistillation) and up-to-date (MAD and MHG) techniques. Isothiocyanates (ITCs), the main volatiles that originate from degradation of GSLs, were identified by GC-MS. (Table 2).
Seven GSLs were detected by UHPLC-DAD-MS/MS analyses. The structures of the corresponding GSLs are shown in Figure 2. [M + Na] + , sodium adducts of desulfoglucosinolate; DW, dry weight; tr < 0.1 μmol/g DW, value is the mean ± standard error (n = 2).   Generally, the yields of the volatiles obtained by conventional techniques were much higher than those obtained by microwave-assisted isolation ( Table 2). HD (100 °C, 2.30 h) which uses large quantities of water and energy in comparison to MAD and MHG, enabled degradation of GSLs and isolation of mostly low molecular and less polar ITCs (from 1, 2, and 5-7). Volatiles obtained by MAD (500 W, 98 °C, 30 min) showed a similar profile of GSL degradation products as HD, albeit with 152fold less yield (3.80 and 578.90 μg/g, respectively). GC-MS analysis revealed that the extraction of enzymatically formed GSLs breakdown volatiles enabled isolation of sulfoxide-bearing aliphatic ITCs, namely 5-(methylsulfinyl)pentyl ITC and 6-(methylsulfinyl)hexyl ITC. Those ITCs were not isolated by HD or MAD, probably due to their higher polarity.

Cytotoxic Activity
Hydrodistillate (HD), CH2Cl2 extract after myrosinase hydrolysis (EXT), and microwave hydrodiffusion and gravity (MHG) were tested against human lung cancer cell line A549 and breast cancer cell line MDA-MB-231 for their cytotoxic activity (Figure 4) (Table S1). The criteria used to categorize the activity against the tested cell lines was based on IC50 values as described under Materials and Methods.  Generally, the yields of the volatiles obtained by conventional techniques w than those obtained by microwave-assisted isolation ( Table 2). HD (100 °C, 2.30 h) quantities of water and energy in comparison to MAD and MHG, enabled degrada isolation of mostly low molecular and less polar ITCs (from 1, 2, and 5-7). Volatiles o (500 W, 98 °C, 30 min) showed a similar profile of GSL degradation products as HD fold less yield (3.80 and 578.90 μg/g, respectively). GC-MS analysis revealed that enzymatically formed GSLs breakdown volatiles enabled isolation of sulfoxide-ITCs, namely 5-(methylsulfinyl)pentyl ITC and 6-(methylsulfinyl)hexyl ITC. Tho isolated by HD or MAD, probably due to their higher polarity.

Cytotoxic Activity
Hydrodistillate (HD), CH2Cl2 extract after myrosinase hydrolysis (EXT), hydrodiffusion and gravity (MHG) were tested against human lung cancer cell line cancer cell line MDA-MB-231 for their cytotoxic activity (Figure 4) (Table S1). Th categorize the activity against the tested cell lines was based on IC50 values as Materials and Methods. Next to dGSL sodium adduct, MS 2 showed characteristic fragments m/z = 185 or 219 corresponding to [C 6 H 10 O 5 + Na] + and [thioGlc + Na] + , respectively ( Figure S2). Two branched GSLs 1, and 2 originated from Val and Ile biosynthetic pathway, respectively. The most abundant peak observed at t R = 3.80 min showed characteristic dGSL sodium adduct m/z = 304 which was identified as desulfoglucoputranjivin (d1), with 14.60 µmol/g of DW. Isopropyl ITC originating from 1 was found in all volatile samples obtained by different isolation methods, ranging from 35% to 92%. dGSL sodium adduct of 2 m/z = 318 was detected at t R = 5.34 min. GSL 2 and its regioisomer isobutyl GSL, both known to exist in plants, cannot be distinguished by MS 2 spectra [21]. However, MS spectra of the corresponding ITCs, having the same odd molecular ion m/z = 115 and characteristic m/z = 72 [CH 2 NCS] + , can be easily discriminated as sec-butyl ITC delivers a strong m/z 86 fragment in difference to isobutyl ITC [17] that enabled us to ascertain the structure of 2.
Generally, the yields of the volatiles obtained by conventional techniques were much higher than those obtained by microwave-assisted isolation ( Table 2). HD (100 • C, 2.30 h) which uses large quantities of water and energy in comparison to MAD and MHG, enabled degradation of GSLs and isolation of mostly low molecular and less polar ITCs (from 1, 2, and 5-7). Volatiles obtained by MAD (500 W, 98 • C, 30 min) showed a similar profile of GSL degradation products as HD, albeit with 152-fold less yield (3.80 and 578.90 µg/g, respectively). GC-MS analysis revealed that the extraction of enzymatically formed GSLs breakdown volatiles enabled isolation of sulfoxide-bearing aliphatic ITCs, namely 5-(methylsulfinyl)pentyl ITC and 6-(methylsulfinyl)hexyl ITC. Those ITCs were not isolated by HD or MAD, probably due to their higher polarity.

Cytotoxic Activity
Hydrodistillate (HD), CH 2 Cl 2 extract after myrosinase hydrolysis (EXT), and microwave hydrodiffusion and gravity (MHG) were tested against human lung cancer cell line A549 and breast cancer cell line MDA-MB-231 for their cytotoxic activity (Figure 4) (Table S1). The criteria used to categorize the activity against the tested cell lines was based on IC 50 values as described under Materials and Methods.
According to the IC 50 , the best cytotoxic activities against A549 cell line were observed for EXT and MHG. EXT showed the best activity after 72 h with IC 50 of 39.15 µg/mL, while MHG showed the best activity after 48 h with IC 50 of 26.79 µg/mL and both can be considered as moderately active. HD showed the lowest activity (IC 50 not reached at 100 µg/mL).
When tested against the MDA-MB-231 cell line, the best cytotoxic activities were shown after 72 h. EXT showed to be highly active with IC 50 11.27, while MHG showed 2.3-fold lower activity, having IC 50 25.83 µg/mL.  Table S1.
According to the IC50, the best cytotoxic activities against A549 cell line were observed for EXT and MHG. EXT showed the best activity after 72 h with IC50 of 39.15 μg/mL, while MHG showed the best activity after 48 h with IC50 of 26.79 μg/mL and both can be considered as moderately active. HD showed the lowest activity (IC50 not reached at 100 μg/mL).
When tested against the MDA-MB-231 cell line, the best cytotoxic activities were shown after 72 h. EXT showed to be highly active with IC50 11.27, while MHG showed 2.3-fold lower activity, having IC50 25.83 μg/mL.
In all volatile isolates, isopropyl ITC was found in highest percentage. HD isolate, having 92% of isopropyl ITC, showed the lowest activity on both cell lines tested. In contrast to EXT and MHG, HD did not contain either 5-(methylsulfinyl)pentyl-or 6-(methylsulfinyl)hexyl ITCs. Therefore, the high cytotoxic activities of EXT and MHG can be attributed to the presence of those two ITCs.
In recent years, 6-(methylsulfinyl)hexyl ITC has been widely studied as a major bioactive compound found in Japanese spice wasabi (Wasabia japonica (Miq.) Matsum.). In vivo and in vitro  Table S1.
In all volatile isolates, isopropyl ITC was found in highest percentage. HD isolate, having 92% of isopropyl ITC, showed the lowest activity on both cell lines tested. In contrast to EXT and MHG, HD did not contain either 5-(methylsulfinyl)pentyl-or 6-(methylsulfinyl)hexyl ITCs. Therefore, the high cytotoxic activities of EXT and MHG can be attributed to the presence of those two ITCs.
In recent years, 6-(methylsulfinyl)hexyl ITC has been widely studied as a major bioactive compound found in Japanese spice wasabi (Wasabia japonica (Miq.) Matsum.). In vivo and in vitro studies have demonstrated that 6-(methylsulfinyl)hexyl ITC has several biological properties regulated through diverse multiple pathways. Nomura et al. studied the effect of 6-(methylsulfinyl)hexyl ITC against several human cancer cell lines (HCC panel) and found the most effective suppression in low concentration towards breast cancer MDA-MB-231 and melanoma LOX-IMVI cell lines. They also compared the activity of similar compounds found in food, i.e., 4-(methylsulfinyl)butyl ITC (sulforaphane, from broccoli) and artificially synthesized ones, namely 2-(methylsulfinyl)ethyl and 8-(methylsulfinyl)octyl ITCs. All tested compounds suppressed cancer cell lines, however 6-(methylsulfinyl)hexyl ITC was the most effective at lower concentrations [24]. Fuke et al. also demonstrated the inhibitory effect of 6-(methylsulfinyl)hexyl ITC in human breast cancer cell line through promotion of apoptosis by inhibiting NF-κB and its control of the PI3K/AKT pathway [25].
Morroni et al. studied the neuroprotective effects of 6-(methylsulfinyl)hexyl ITC in a Parkinson s disease mouse model and found that the decrease of apoptotic cell death and the activation of glutathione-dependent antioxidant systems could be underlying a mechanism of beneficial ITC effect [26]. Further investigation of the neuroprotective effect on IMR-32 human neuroblastoma cells was conducted by Trio et al., who concluded that 6-(methylsulfinyl)hexyl ITC can exert a neuroprotective effect by activating the Nrf2-mediated oxidative stress response pathway [27]. Likewise, 6-(methylsulfinyl)hexyl ITC showed inhibitory effect on colon carcinogenesis through a p53-independent mitochondrial dysfunction pathway [28]. Inhibition of viability, accompanied by the features of mitotic arrest and apoptosis of human pancreatic cancer cell lines PANC-1 and BxPC-3 were also observed for 6-(methylsulfinyl)hexyl ITC [29]. The latest study showed that 6-(methylsulfinyl)hexyl ITC inhibited the viability of human chronic myelogenous leukemia K562 cells, as well as suggesting its anti-leukemia activity [30].
These findings are in accordance with observed high cytotoxic activities of L. annua EXT and MHG against human lung cancer cell line A549 and breast cancer cell line MDA-MB-231 which can be associated with the presence of 5-(methylsulfinyl)pentyl ITC and 6-(methylsulfinyl)hexyl ITC.

Conclusions
The volatiles obtained from L. annua seeds by CH 2 Cl 2 extraction after myrosinase hydrolysis showed high cytotoxic effect on the tested tumor cell lines. Such activity can be associated with the ITCs, which are present in large proportion. It can be suggested that the ω-(methylsulfinyl)alkyl ITCs are responsible for the observed high activities against the tested cancer cell lines, making them potential candidates for anticancer treatment studies. The compounds obtained by microwave hydrodiffusion and gravity were consistent with the ones obtained by the extraction after the enzyme hydrolysis. However, the yields obtained using the microwave techniques were much lower. Therefore, it is necessary to optimize time and power in order to produce the compounds of interest with improved yields.