Nova Biotechnol Chim (2019) 18(1): 37-43 DOI: 10.2478/nbec-2019-0005  Corresponding author: katarina.hrobonova@stuba.sk Nova Biotechnologica et Chimica Application of solid-phase extraction for isolation of coumarins from wine samples Katarína Hroboňová, Andrea Špačková and Martina Ondáková Institute of Analytical Chemistry, Faculty of Chemical and Food Technology, Slovak University of Technology in Bratislava, Radlinského 9, SK-812 37 Bratislava, Slovak Republic Article info Article history: Received: 10th May 2019 Accepted: 20th June 2019 Keywords: HPLC Simple coumarins Solid-phase extraction Wine Abstract Coumarins can be as a result wine storage and aging in wood drums and they can affect organoleptic characteristics of wine. The aim of this work was to determine the content of coumarins in wine samples originated from Slovak Tokaj wine region. The HPLC method with high specificity, accuracy, precision, and recovery was proposed. SPE sorbents of C18 type, styrene-divinylbenzene copolymer and molecularly imprinted polymers were compared for extraction of six coumarins, coumarin, aesculin, scoparone, scopoletin, 4-methylumbelliferone, and herniarin. Higher recoveries (above 89 %; except aesculin – recoveries higher from 68 %, RSDs less than 6 %) were obtained with selective polymeric sorbent laboratory prepared by molecularly imprinted technology. The results showed that content of coumarins in wine samples are in ng.mL-1 concentration levels and depend on the age and wine puttony.  University of SS. Cyril and Methodius in Trnava Introduction Coumarins, a class of compounds that contain a 1,2-benzopyrone skeleton, occur as secondary metabolites mainly originated from shikimic acid pathway. These compounds are widely distributed in the seeds, roots and leaves of many plants of Apiaceae, Rutaceae, Asteraceae and Fabaceae families. The source of coumarins (glycosylated or aglycone forms) in wine could be wood of drums in which it is matured. Wood releases the coumarins and other compounds (e.g. phenolic acids) into the wine, which contribute to wine taste (e.g. glycosides are bitter, aglycones are acidic). The content of coumarins in wood changes through the degradation of the wood lignin during the heat treatment of the barrel and also depends on the type of the wood (Ribéreau-Gayon et al. 2006; De Rosso et al. 2009; Canas 2017). The main coumarins in wine are aesculetin (6,7-dihydroxy- coumarin), aesculin (6,7-dihydroxy-coumarin-6- glucoside), scopoletin (7-methoxy-6-hydroxy- coumarin), and scopolin (scopoletin-7-glucoside). In some wines, 4-hydroxycoumarin and umbelliferone (7-hydroxycoumarin) were identified (Liberatore et al. 2010; Canas 2017). Coumarins can be used for evaluation of wine, as a possible markers of storage of wines in wood barrels. In addition, the coumarins exhibit a wide range of biological effects, e.g. anti-inflammatory, antithrombotic, antimicrobial, antifungal, antiviral (including anti-HIV), anticonvulsant, antioxidant, and antitumor (Kubrak et al. 2017). Wine is very complex matrices, often components of very low concentrations are of interest (concentrations of ng.mL-1 for coumarins; Salagoity-Auguste et al. 1987). Therefore, it is necessary to use sample preparation and analytical Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 12:01 UTC mailto:katarina.hrobonova@stuba.sk Nova Biotechnol Chim (2019) 18(1): 37-43 38 Fig. 1. Names, structural formulas and pKa values of simple coumarins. methods for determination of coumarins that are characterised by high selectivity and sensitivity. Sample preparation is an essential part of chemical analyses. The main goal of sample preparation for subsequent HPLC analysis is removal of matrix interferences and to obtain sample aliquot or final analytes extract in solvent which is compatible with the HPLC mobile phase used. Sometimes, depending on detection sensitivity, pre-concentration of target analytes is needed. The most widely and routinely used sample preparation technique (clean-up and pre- concentration) for the extraction of analytes from a complex matrix (including wine samples) is solid-phase extraction (SPE). This technique is rapid, easy to perform, can be automated, and low amount of solvents is handled (compared with liquid-liquid extraction). A broad range of sorbents based on electrostatic, hydrophobic or polar interactions exists for SPE. These traditional SPE sorbents are widely available, well characterized and have high binding capacity, but only a few types (e.g. C18 type) are used in the majority of SPE sample preparations, and they are the solvents in wash and/or elution steps which are varied according to the applications. Sometimes their selectivity is limited and the target analyte often co-elutes with interfering compounds from matrix. There are several publications, which present applications of different sample preparation techniques in the wine analysis (Lorrain et al. 2013), mainly for extraction of selected phenolic compounds. Research in solid-phase extraction techniques is focused on development of sorption materials with increasing selectivity for target analytes. A novel type of SPE sorbets are molecularly imprinted polymers (MIPs) ‒ synthetic tailor-made materials with a pre-defined selectivity for a target analyte or structurally related compounds based on specific interactions between analyte and binding sites on MIP. This technology enables to reduce the presence of interferences during the extraction procedure and obtain higher recoveries even at low concentration levels. MIPs are obtained by polymerising functional and cross-linking monomers around a template molecule, which lead to a highly cross-linked three-dimensional network polymer. After polymerisation, the template molecule is extracted and binding sites having their shape, size, and functionalities complementary to the target analyte are established. The resulting imprinted polymers are stable, robust, and resistant to a wide range of pH, solvents, and temperature. The MIP solid-phase extraction (MISPE) was applied in wine analysis for determination of quercetin (Molinelli et al. 2002), ochratoxin A (Giovannoli et al. 2014), phthalates (Barciela- Alonso et al. 2017), organophosphorous herbicides (Sanagi et al. 2011), catechin (Büyüktuncel et al. 2018), umbelliferone (Machyňáková et al 2017a). In this paper the content of coumarins in wine samples were determined. Different SPE sorbents, C18 bonded silica, styrene-divinylbenzene 7-hydroxycoumarin (umbelliferone) pKa = 8.01 7-methoxycoumarin (herniarin) coumarin 7-methoxy-6- hydroxycoumarin (scopoletin) 6,7-dimethoxycoumarin (scoparone) pKa = 7.13 6,7-dihydroxycoumarin-6-β- D-glucoside (aesculin) pKa = 8.12 4-methyl-7-hydroxycoumarin (4-methylumbelliferone) pKa = 7.80 Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 12:01 UTC Nova Biotechnol Chim (2019) 18(1): 37-43 39 copolymers, and MIP based sorbents have been tested and compared for extraction of simple coumarins (Fig. 1) the most important natural coumarins. HPLC method was used for determination of the extracted analytes. Experimental Chemicals and samples The standards of 7-hydroxycoumarin (99 %), 7-methoxycoumarin (98 %), 4-methyl-7-hydroxy- coumarin (98 %), 6,7-dimethoxycoumarin (98 %), 7-methoxy-6-hydroxycoumarin (98 %), 6,7- dihydroxycoumarin-6-β-D-glucoside (98 %), and coumarin (99 %) used for purpose of this study were purchased from Sigma-Aldrich (St. Louis, USA). Solvents for SPE extraction and preparation of HPLC mobile phases, methanol and ethanol of HPLC gradient grade and acetic acid (99 %) were purchased from Merck (Germany). HPLC grade water was prepared by AquaMax ultra (series 370) water purification system. The Tokaj wine (sample I: puttony 3 (2002), sample II: puttony 4 (2004), sample III: puttony 5 (2002), sample IV: puttony 6 (2003) was a product of Slovak winery. Preparation of standard solutions Stock standard solutions of each compound (aesculin, scopoletin, coumarin, scoparone, 4-methylumbelliferone, herniarin,) were prepared in methanol/1 % acetic acid (20/80; v/v) to reach concentration of 0.1 mg.mL-1 and stored in freeze at –20 °C. The mixed working solutions were prepared by diluting the stock solutions obtaining lower concentrations. Sample preparation The SPE cartridge (C18 Hydra (I), HR-X (II), HR- P (III) (all cartridges purchased from Chromabond, Macherey-Nagel, Germany), AFFINIMIP Phenolic (IV) (purchased from Affinisep, France) and cartridge with 0.1 g of laboratory prepared MIP-umbelliferone adsorbent) was preconditioned with 3 mL of methanol and subsequently with 3 mL of 12 % ethanol. The appropriate volume of sample (5 mL) was passed through cartridge followed by washing with 3 mL of water. Finally, the analytes were eluted with 1 mL of methanol/acetic acid (9/1; v/v). Extract was filtered through a 0.45 µm nylon membrane filter and used for the injection into HPLC. The extraction procedure was repeated three times for each kind of tested sorbent material. For solid- phase extraction the vacuum manifold was used. HPLC analysis A HPLC Agilent Technologies (series 1200) consisting of a binary pump, an autosampler, a column oven, a diode array detector, and fluorescence detector was used. HPLC analyses were performed on a Kinetex C18 (100 mm x 4.6 mm I.D., 5 m particle size) analytical column. The column was kept at 23 °C. An injection volume was 20 µL. The mobile phase consisted of methanol/acetic acid (99/1; v/v) (A) and 1 % aqueous solution of acetic acid (B) was mixed by the gradient program as follows, 0 ‒ 12 min linear gradient of A from 20 to 45 %, 12 ‒ 12.5 min linear gradient of A from 45 to 100 %, 12.5 ‒ 14.5 min 100 % of A, 14.5 ‒ 15 min a reverse gradient of A from 100 to 20 %, and 15 ‒ 19 min 20 % A. The flow rate was 1.0 mL min-1. The diode array detector was operated in the wavelength range of 190 ‒ 400 nm and chromatograms were recorded at 280 nm for coumarin and 330 nm for aesculin, scopoletin, scoparone, 4-methylumbelliferone, herniarin. Fluorescence detector was set at the wavelengths 330 nm (ex) and 450 nm (em) and fluorescence spectra were scanned in the wavelength range of 340 ‒ 500 nm. Results and Discussion Different types of SPE sorbents were selected for purpose of this study, i) alkyl – bonded silica, ii) styrene-divinylbenzene copolymeric sorbents, and iii) selective MIP based polymeric sorbents. In this work, all selected cartridges and laboratory prepared MIP-umbelliferone cartridge were used in the same SPE procedure with optimal extraction conditions included conditioning with methanol and 12 % ethanol, washing with water, and analyte eluting with methanol/acetic acid (9/1; v/v). Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 12:01 UTC Nova Biotechnol Chim (2019) 18(1): 37-43 40 Fig. 2. Recovery values of coumarins on different SPE sorbents. Efficiencies of SPE sorbents evaluated by recovery values were calculated and compared (Fig. 2). Coumarins under study represent less polar type of compounds and from this reason the reversed phase approach was selected for SPE. Reversed phase mode involves a polar or moderately polar sample matrix and a nonpolar adsorbent. Octadecyl silica type of sorbent (I) is most frequently used for isolation of less polar compounds. The retention of analytes from polar matrices is due to the Van der Waals forces between the carbon-hydrogen bonds in the analyte and the functional groups on the silica surface. To elute a compound from adsorbent, a nonpolar solvent is used (the disruption of the forces analyte-packing adsorbent). The hydrophobic (II) and high porous (III) polymeric polystyrene-divinylbenzene sorbents were also used for purpose of the study. The advantage of these adsorbents is higher resistance in whole pH range, which leads to applicability of the SPE methods for acidified samples to. They are suitable for extraction of compounds from complex, food, pharmaceutical or biological samples (II) and from water samples (III). The efficiency values of SPE for sorbents I and II are approximately the same and higher than 89 %, except aesculin, where sorption recovery was significantly lower (less than 64 %). The lover recoveries were obtained for some of compounds for adsorbent III. The best extraction performance of adsorbent II, in comparison with I and III adsorbents, is related to aromatic functional groups and hydrophobic moieties of the sorbent which produce - and hydrophobic interactions for retention of analytes. MIP sorbents, on contrary to the traditional SPE sorbents (e.g. silica-based sorbents) are characterised by higher selectivity and specificity Table 1. Parameters of the MISPE-HPLC methoda. Compound tr [min] Regression equation Concentration range LOD LOQ aesculin 2.51 y = 131.1x – 3.3 3.0 – 100 1.0 3.0 R2 = 0.987 ng.mL-1 ng.mL-1 ng.mL-1 scopoletin 6.50 y = 1219.7x + 13.2 3.0 – 100 1.0 3.0 R2 = 0.998 ng.mL-1 ng.mL-1 ng.mL-1 coumarin 8.38 y = 3650.8x + 5.3 0.5 – 100 0.2 0.5 R2 = 0.995 µg.mL-1 µg.mL-1 µg.mL-1 scoparone 8.98 y = 82.8x + 1.3 10.0 – 100 3.0 10.0 R2 = 0.998 ng.mL-1 ng.mL-1 ng.mL-1 4-methylumbelliferone 9.45 y = 306.3x + 6.2 0.5 – 100 0.2 0.5 R2 = 0.997 ng.mL-1 ng.mL-1 ng.mL-1 herniarin 11.30 y = 25.5x – 0.2 12.0 – 100 4.0 12.0 R 2 = 0.998 ng.mL-1 ng.mL-1 ng.mL-1 a coumarin ‒ parameters for UV spectrophotometric detection; other coumarins – parameters for FL detection. Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 12:01 UTC Nova Biotechnol Chim (2019) 18(1): 37-43 41 Table 2. Accuracy and precision of the MISPE-HPLC method a. Compound Recovery (%) ± SD C1 C2 C3 aesculin 68.7±0.8 70.9±0.9 72.1±0.9 scopoletin 91.4±3.2 94.7±3.8 96.0±3.3 coumarin 90.0±3.1 89.8±2.6 89.0±2.9 scoparone 94.9±1.1 91.3±1.1 95.4±1.4 4-methylumbelliferone 91.3±4.5 95.3±5.1 92.3±4.5 herniarin 95.2±2.3 97.3±2.3 92.3±2.0 a coumarin ‒ C1 1.0 µg.mL-1, C2 5.0 µg.mL-1, C3 50.0 µg.mL-1; other coumarins ‒ C1 0.01 µg.mL-1, C2 0.05 µg.mL-1, C3 0.1 µg.mL-1; n = 6. to target analyte, leading to effective elimination of matrix interferences. MIP selectivity is related Fig. 3. Chromatograms of coumarins mixture (A; line a ‒ UV spectrophotometric detection, b ‒ fluorescence detection), wine sample III extract without (B) and after (C) MISPE extraction. The peaks indicate for aesculin ‒ 1, scopoletin ‒ 2, coumarin ‒ 3, scoparone ‒ 4, 4-methylumbelliferone ‒ 5, herniarin ‒ 6. to recognition cavities in polymeric matrices which are complementary to the target molecule in size, shape, and arrangement of the functional groups. Two types of MIPs were tested for purpose of this study, MIP selective for phenolic compounds (IV) and coumarins (V). Although the sorbent IV was designed for phenolic compounds, the recovery values (higher than 88 %) indicated, that it was suitable for isolation of investigated coumarins. The lowest recovery of aesculin (about 50 %) results from incompatibility of MIP imprinted cavity with shape and functionality of analyte. The laboratory synthesised MIP (V) by imprinting of umbelliferone molecule was selective for template and also for other structural analoges (Fig. 1) (Machyňáková et al. 2017b). This sorbent shows complete retention of all investigated coumarins with highest values of recovery, 92 ‒ 98 %, compared with traditional and polymeric sorbents I-IV. The advantage of MIP based sorbents is alongside increased selectivity also reusability. Sorbent can be reused up to fifty times without losing extraction capacity. The real sample of wines were selected in order to demonstrate the applicability for the MISPE method. Parameters including LOD, LOQ, linear range, correlation coefficient, precision and accuracy were examined to validate the developed MISPE method coupled with HPLC-DAD/FLD. By using HPLC method, acceptable asymmetry (AS ~ 0.9), good peaks resolution (RS  2.1), and analysis time lower than 15 min were obtained (Fig. 3A). The RSDs repeatability of retention times varied from 0.2 to 0.5 %. LODs and LOQs (Table 1) were calculated using 3 sa and 10 sa criteria (sa is the standard deviation of the intercept of calibration curve). The calibration curves were linear in working ranges from LOQ to 100 µg.mL-1 A B C Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 12:01 UTC Nova Biotechnol Chim (2019) 18(1): 37-43 42 for coumarin detected with UV detection and in the LOQ – 100 ng.mL-1 concentration range for other coumarins detected with fluorescence detector. The precision and accuracy (expressed as the recovery) of the developed method was determined for spiked sample at three concentration levels of each analyte (Table 2). The intra-day precision expressed as RSD (%) ranged from 1.2 to 5.4 %. The obtained recoveries ranged from 68.7 to 96.0 %. Two of coumarins, scopoletin and 4-methylumbelliferone, were determined in wine samples using MISPE-HPLC method in ng.mL-1 concentration levels (sample I: 5.40.3 ng.mL-1 of scopoletin, sample II: 9.20.1 ng.mL-1 of scopoletin, sample III: 10.10.5 ng.mL-1 of scopoletin and 12.00.3 ng.mL-1 of 4-methylumbelliferone, sample IV: 17.20.3 ng.mL-1 of scopoletin). Chromatograms of wine extracts (Fig. 3) documented that MISPE was suitable for extraction and pre-concentration of coumarins from wine samples. The results showed that content of coumarins in wine samples depended on the age and wine puttony. Conclusions In this study, the solid-phase extraction with different types of sorbents, C18-silica, styrene- divinylbenzene copolymeric sorbents, and selective MIP based sorbents were compared for isolation and pre-concentration of six coumarins. MIPs displayed higher selectivity and reusability. The obtained recoveries for target analytes ranged from 68.7 to 96.0 %. The results obtained demonstrated the appropriateness of solid-phase extraction with MIP-umbelliferone sorbent in wine analysis. Developed MISPE coupled with HPLC method was applicable for determination of simple coumarins in real samples and exhibited good specificity, accuracy, precision and recovery. The results showed that content of coumarins in wine samples are in ng.mL-1 concentration levels and depend on the age and wine puttony. For this purpose, further research on identification and quantification of coumarins is needed. Alternatively, this study could provide a promising method for the selective separation and determination of coumarins in other samples of beverages and sample extracts too. Acknowledgements This work was supported by the Slovak Research and Development Agency (Grant no. APVV-15-0355). Conflict of Interest The authors declare that they have no conflict of interest. References Barciela-Alonso MC, Otero-Lavandeira N, Bermejo-Barrera P (2017) Solid phase extraction using molecular imprinted polymers for phthalate determination in water and wine samples by HPLC-ESI-MS. Microchem. J. 132: 233-237. Büyüktuncel E, Porgali E, Özkara S (2018) Catechin- molecularly imprinted cryogel for determination of catechin in red wines by HPLC-DAD–Fluorescence detector. Acta Chromatogr. 30: 54-61. Canas S. (2017) Phenolic compounds and related properties of aged wine spirits: Influence of barrel characteristics. A Review. Beverages 3: Article number: 55. De Rosso M, Panighel A, Dalla Vedova A, Stella L, Flamini R (2009) Changes in chemical composition of a red wine aged in acacia, cherry, chestnut, mulberry, and oak wood barrels. J. Agric. Food Chem. 57: 1915-1920. Giovannoli C, Passini C, Di Nardo F, Anfossi L, Baggiani C (2014) Determination of Ochratoxin A in italian red wines by molecularly imprinted solid-phase extraction and HPLC analysis. J. Agric. Food Chem. 62: 5220- 5225. Kubrak T, Podgórski R, Stompor M (2017) Natural and synthetic coumarins and their pharmacological activity. Eur. J. Clin. Exp. Med. 15: 169-175. Liberatore MT, Pati S, Del Nobile MA, La Notte E (2010) Aroma quality improvement of Chardonnay white wine by fermentation and ageing in barrique on lees. Food Res. Int. 43: 996-1002. Lorrain B, Ky K, Pechamat L, Teissedre PL (2013) Evolution of analysis of polyhenols from grapes, wines, and extracts. Molecules 18: 1076-1100. Machyňáková A, Lhotská I, Hroboňová K, Šatínský D (2017a) On-line coupling of molecularly imprinted solid-phase extraction with liquid chromatography for the fast determination of coumarins from complex samples. J. Pharm. Biomed. Anal. 145: 144-150. Machyňáková A, Hroboňová K (2017b) Synthesis and evaluation of molecularly imprinted polymers as sorbents for selective extraction of coumarins. Chromatographia 80: 1015-1024. Molinelli A, Weiss R, Mizaikoff B (2002) Advanced solid- phase extraction using molecularly imprinted polymers for the determination of quercetin in red wine. J. Agric. Food Chem. 50: 1804-1808. Ribéreau-Gayon P, Glories Y, Maujean A, Dubourdieu D (2006) Handbook of Enology, Volume 2: The Chemistry Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 12:01 UTC Nova Biotechnol Chim (2019) 18(1): 37-43 43 of Wine – Stabilization and Treatment, 2nd ed., John Wiley & Sons, Ltd, Chichester, p. 143-144. Salagoity-Auguste MH, Tricard C, Sudral P (1987) Simultaneous determination of aromatic aldehydes and coumarins by high-preform liquid chromatography. Application to wines and brandies stored in oak barrels. J. Chromatogr. 392: 379-387. Sanagi MM, Salleh S, Ibrahim WAW, Naim AA (2011) Determination of organophosporus pesticides using molecular imprinted polymer solid-phase extraction. Malaysian J. Anal. Sci. 15: 175-183. Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 12:01 UTC