Analytica Chimica Acta 807 (2014) 153–158
Contents lists available at ScienceDirect
Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca
Contactless conductivity detection for screening myrosinase substrates by capillary electrophoresis Reine Nehmé ∗ , Hala Nehmé, Grégory Roux, Deimante Cerniauskaite, Philippe Morin, Patrick Rollin, Arnaud Tatibouët Institut de Chimie Organique et Analytique (ICOA), Université d’Orléans, CNRS FR 2708, UMR 7311, Orléans, France
h i g h l i g h t s
g r a p h i c a l
a b s t r a c t
• Myrosinase activity was assessed by CE for the ﬁrst time.
• Contactless conductivity detection C4 D is useful to evaluate myrosinase kinetics. • The developed enzymatic assay is economic, simple and fast (short-end injection). • Km value for several glucosinolates toward myrosinase were determined. • The afﬁnity of myrosinase toward various substrates was studied.
a r t i c l e
i n f o
Article history: Received 10 July 2013 Received in revised form 4 November 2013 Accepted 7 November 2013 Available online 18 November 2013 Keywords: Contactless capacitively coupled conductivity detector (C4 D) Enzyme kinetics Glucosinolates Myrosinase Pre-capillary electrophoresis
a b s t r a c t Myrosinase is a unique enzyme that catalyzes the hydrolysis of glucosinolates (GLS) to isothiocyanate (ITC), glucose and sulfate. Isothiocyanates display a diversiﬁed very interesting biological activity. In this study, capillary electrophoresis (CE) was used for the ﬁrst time for evaluating myrosinase kinetics (maximum velocity Vmax and Michaelis–Menten constant Km ) and to assess the afﬁnity of a variety of substrates toward this enzyme. The pre-capillary approach was chosen since it is very simple to conduct. For this, the enzymatic reaction was performed in a micro-vial. The reaction mixture volume was of only 100 L and the incubation lasted only 5 min at 37 ± 1 ◦ C. Short-end injection of few tens of nanoliters (∼25 nL) of the reaction mixture was performed which decreased analysis time without using any electroosmotic modiﬁer. The sulfate produced was detected and quantiﬁed with a contactless capacitively coupled conductivity detector (C4 D) allowing the evaluation of myrosinase kinetics. This study shows, that capillary electrophoresis with contactless conductivity detection can be very useful for monitoring myrosinase activity. Comparing to the conventional spectrophotometric method (1982), the CE method developed here is simple, automated, economic, rapid (incubation for few minutes) and robust. Results compared very well with those reported in literature using the conventional method. Moreover, the afﬁnity of a variety of natural and synthetic glucosinolates toward this enzyme has been assessed for the ﬁrst time. © 2013 Elsevier B.V. All rights reserved.
Abbreviations: C4 D, contactless capacitively coupled conductivity detector; CPA, corrected peak area; GLS, glucosinolate; Km , Michaelis–Menten constant; Vmax , maximum velocity. ∗ Corresponding author. Tel.: +33 2 38 49 27 75; fax: +33 2 38 41 72 81. E-mail address: [email protected]
(R. Nehmé). 0003-2670/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.aca.2013.11.012
From the “Natural History” by Pliny the Elder (23–79 AD) and the “De Materia Medica” pharmacopeia drawn up by Dioscorides (40–90 AD), the beneﬁcial effects of vegetables of the crucifer family have long been acknowledged. Nowadays, this positive impact is recognized to be due to sulfur-containing secondary
R. Nehmé et al. / Analytica Chimica Acta 807 (2014) 153–158
metabolites known as glucosinolates (GLS). These S-glucopyranosyl thioesters-erroneously called “thioglucosides”–are strikingly biorelevant metabolic markers which occur in all plant families of the order Brassicales-namely in our daily vegetables . All known GLS (ca. 120 molecules) display a remarkable structural homogeneity based on a hydrophilic ␤-d-glucopyrano unit, an Osulfated anomeric (Z)-thiohydroximate function connected to a rather hydrophobic side chain which constitution, depending on plant species, is the sole structural variant (Fig. 2(3)) in which diversiﬁed aliphatic, aryl aliphatic or heterocyclic arrangements can be found. For example for sinigrin, a natural substrate of myrosinase, the side chain is CH2 CH CH2 . GLS are associated in plants with an atypical endogenous glucohydrolase, myrosinase, which assists their hydrolysis by releasing the glucose moiety. The resulting unstable aglycon (thiohydroximate-O-sulfate) then eliminates a sulfate ion through a Lossen-type rearrangement (Fig. 1). Several compounds can be produced, mainly isothiocyanates or nitriles . GLS therefore operate like bio-precursors to produce isothiocyanates which display however, in certain concentrations, high cytotoxicity [see  and references therein]. Since several decades, scientists have put many efforts trying to clarify the unique myrosinase-glucosinolate relationship and the mechanism of action of this enzyme-the only glycohydrolase able to break an anomeric carbon-sulfur bond. The kinetic properties of the myrosinase enzyme are certainly the most important to consider. For this, quite a number of artiﬁcial GLS were synthesized to compare with natural GLS [4,5] and their afﬁnity to myrosinase was evaluated by using a spectrophotometric assay developed in 1980 by Gil and McLeod . This is based on following the glucosinolate decomposition through measuring the decrease in absorbance at 227 nm. In 1982, Palmieri et al.  showed that this assay allows the determination of initial myrosinase reaction rates and can thus be used for steady-state kinetic studies. Sinigrin, which is one of the few commercially available GLS, was used as substrate for myrosinase. The Lineweaver–Burk plot could be obtained and the Michaelis–Menten constant (Km ) value for sinigrin was found to be 0.156 ± 0.004 mM (at pH 7.0). Since then, this method has been used for the evaluation of myrosinase kinetics. For example, Palmieri et al.  used this spectrophotometric assay to evaluate the afﬁnity toward myrosinase (isolated from Sinapis alba of several synthetic GLS belonging to miscellaneous structural classes. By determining Km values, it was shown that the role of the glycon site seems to be crucial, both for selectivity and afﬁnity. Similarly, Durham et al.  had previously tested the hydrolytic activity of myrosinase (isolated from Lepidium sativum L.) toward a range of natural and synthetic O- and S-glycosides. Likewise, Botti et al.  showed that myrosinase (S. alba) is generally a very stable enzyme under different reaction conditions (methanol: water mixtures, storage temperature.). However, the spectrophotometric assay developed by Palmieri et al.  is time- and sample-consuming (incubation time of 4 h and several milliliters of reagents used). Moreover, this method lacks precision and sensitivity since it is based on measuring a decrease in absorbance. The aim of the present study was to develop a novel method for evaluating myrosinase activity in order to overcome these limitations [10–12].
Fig. 2. Michaelis–Menten plots for myrosinase-catalyzed hydrolysis of sinigrin. (1) reaction mixture volume 400 L; incubation at 30 ◦ C, (2) reaction mixture volume 400 L; incubation at 37 ◦ C, and (3) reaction mixture volume 100 L; incubation at 37 ◦ C. Incubation time: 5 min. BGE: histidine/acetic acid (I = 40 mM; pH 4.6). Short-end injection: −50 mbar × 10 s; electrophoretic separation: +20 kV at 25 ◦ C (i = 31 A). Rinse between analyses: 3 min with BGE. Bare-fused silica capillary: 35 cm total length; 19 cm detection length to C4 D detector; 50 m i.d. C4 D detector: frequency medium, voltage 0 dB, gain 100%, offset 152, ﬁlter: frequency 1/3 and cut-off 0.05.
Capillary electrophoresis (CE) was chosen since it is known to be simple, rapid, very economic and eco-friendly . The sulfate produced during the hydrolysis of GLS by myrosinase was on-line detected for quantiﬁcation. The choice of a detection technique for the analysis of small ions like sulfate is limited. The widespread indirect UV detection has a narrow linearity range and requires a long optimization [14–16]. Conductivity detection is an interesting alternative [17,18]. The contactless capacitively coupled conductivity detector (C4 D) introduced in 1998 by Zeeman et al.  and Fracassi da Silva and do Lago  was chosen in this work since it is characterized by a high sensitivity and a large linear range [21–23]. Hauser and co-workers successfully used CE-C4 D to monitor activity of several enzymes: hexokinase, glucose oxidase, alcohol dehydrogenase, and esterase . Schubert-Shi and Hauser determined Michaelis–Menten constants (Vmax and Km ) of the conversion of urea to ammonium by urease using CE-C4 D . They also used this approach to monitor the porcine pancreatic lipase catalyzed enantioselective hydrolysis of amino acid esters . They also determined inhibition concentration at 50% (IC50 ) of three acethylcholinesterase inhibitors . The off-line and online peptic and tryptic digestion of several peptides and proteins was also assessed by CE-C4 D . The validity of the novel CE method was veriﬁed by determining the Km value for sinigrin–used as a reference substrate–and comparing it to the one reported by Palmieri et al. . The developed method was also used to evaluate, for the ﬁrst time, Km for various natural GLS 1–6 (Table 2.1) extracted, respectively, from Sinapis alba (white mustard), Indian cress, turnip, daikon, Iberis amara and Moringa oleifera. In addition,
Fig. 1. Hydrolysis scheme of glucosinolates (GLS) catalyzed by myrosinase.
R. Nehmé et al. / Analytica Chimica Acta 807 (2014) 153–158
Table 1 Screening of myrosinase substrates using capillary electrophoresis. Entry
0.072 ± 0.006 (r2 = 0.9924)
0.125 ± 0.01 (r2 = 0.9955)
0.208 ± 0.025 (r2 = 0.9829)
0.311 ± 0.025 (r2 = 0.9923)
0.464 ± 0.035 (r2 = 0.9955)
2.602 ± 0.189 (r2 = 0.9920)
9.614 ± 1.677 (r2 = 0.9844)
13.420 ± 2.908 (r2 = 0.9809)
16.030 ± 2.843 (r2 = 0.9881)
17.050 ± 3.876 (r2 = 0.9858)
R. Nehmé et al. / Analytica Chimica Acta 807 (2014) 153–158
some unnatural GLS 7–10 (Table 2.1) bearing diverse aglycon moieties were synthesized and evaluated. These substrates were expected to cover a wide range of afﬁnity toward myrosinase, with Km values ranging from few M to few tens of mM. 2. Materials and methods 2.1. Chemicals Glacial acetic acid (AcOH, CH3 CO2 H, purity ≥ 99.99%), lhistidine (l-His, C6 H9 N3 O2 , purity ≥ 99.5%), potassium phosphate dibasic trihydrate (K2 HPO4 • 3H2 O, purity ≥ 99.0%), potassium phosphate monobasic (KH2 PO4 , purity ≥ 99.0%), sodium hydroxide (NaOH, purity ≥ 98%), sodium sulfate (Na2 SO4 , purity ≥ 99.0%), (−) sinigrin hydrate (C10 H16 NO9 S2 • K• xH2 O, ≥ 99.0%) and thioglucosidase from Sinapis alba (white mustard) seed (myrosinase, EC 18.104.22.168, 25 U, ≥100 units g−1 ) were purchased from Sigma–Aldrich (Saint-Quentin Fallavier, France). Bidistilled 18 Mcm water was from Carlo Erba (Val de Reuil, France). All chemicals were used as received. All product bottles were placed in tightly closed plastic bags to preserve them from aerosol contaminations  and then stored at 4 ◦ C. Natural glucosinolates are gracious gifts from Dr R. Iori (ISCI, Bologna, Italy). Unnatural glucosinolates were synthesized by our group. Details of their synthesis and analytical characterization (nuclear magnetic resonance (NMR), high resolution mass spectrometry (HRMS)) are presented in the supporting information. 2.2. Equipment and operating conditions Experiments were carried out on an HP3D CE electrophoretic system (Agilent, Waldbronn, Germany). Agilent software 3D-CE Chemstation (rev B.04.02) was used to pilot the CE. The HP3D CE was equipped with an on-capillary TraceDec capacitively coupled contactless conductivity detector (C4 D) (Innovative Sensor Technologies GmbH, Strasshof, Austria). The following parameters were ﬁxed for the C4 D: frequency medium; voltage 0 dB; gain 100%; offset 152; ﬁlter: frequency 1/3 and cut-off 0.05. The detection signal was acquired with the Tracemon software (Istech, version 0.07a). CE analyses were performed in an uncoated fused-silica capillary of 35 cm total length and 50 m i.d., purchased from Polymicro Technologies (Phoenix, AZ, USA). The effective detection length to the C4 D in short-end injection was 19 cm. New capillaries were conditioned by performing the following rinses: 1 M NaOH (15 min), water (5 min) and BGE (10 min) for equilibration before analyses. Between runs, the capillary was only rinsed with the BGE for 3 min. All rinse cycles were carried out at about 950 mbar. At the end of each working-day, the capillary was rinsed for 10 min with water before storing it over night. Polypropylene ﬂasks, containers, CE vials and polyethylene oleﬁn snap caps were used since they have little extractable material unlike glass. To avoid any inadvertent contamination, nitrile-powder free-gloves were worn during solution preparation and sample handling. To ensure good-quality cleaning, all volumetric equipment, vials and snap caps were rinsed with water (18 M-cm), and then soaked in water for an overnight period. 2.3. Solutions All solutions were prepared with ultra-pure bidistilled water, ﬁltered through a 0.45 m polyvinylidenediﬂuoride (PVDF) MillexHV Syringe Filter (Millipore, Molsheim, France) before use and stored at 4 ◦ C when not in use. Williams et al.  found that these ﬁlters do not introduce signiﬁcant amounts of inorganic impurities (e.g. chloride and sulfate) into solutions, especially into
sample solutions. Phosphate buffer (KH2 PO4 /K2 HPO4 ) was used as an incubation buffer at an ionic strength (I) of 33.1 mM. Its pH was ﬁxed at 7.0 to be at the optimum pH value for the myrosinase to be active . For electrophoretic separation, the background electrolyte was His/AcOH (I = 40 mM; pH 4.6). It was prepared daily and degassed by ultra-sonication before use. The BGE solution in the separation vials was changed every three runs. The ionic composition and parameters of these buffers were given by Phoebus software (Analis, Namur, Belgium). Their pH was measured with a MeterLab PHM201 Portable pH-Meter (Radiometer Analytical, Villeurbanne, France). Stock solutions of each glucosinolate substrate, of the product SO4 2− (100 mM) and of the enzyme myrosinase (5 U mL−1 ) solution were prepared by dissolving the appropriate quantity of salt-accurately weighed-in the incubation phosphate buffer. Working solutions were prepared by appropriate dilution of the corresponding stock solution in the incubation buffer. Substrates must be used at large excess compared to enzyme (about 100 times). For this, myrosinase was used at 0.05 U mL−1 for all enzymatic assays. The response function of the C4 D detector was examined by plotting the relative corrected-peak areas vs. SO4 2− concentrations. The calibration curve was performed over the concentration range 0.015–15 mM SO4 2− . Three independent determinations were performed at each concentration. Repeatability was assessed by injecting the same standard solution (0.035 mM SO4 2− ) six times successively (n = 6). It was expressed as relative standard deviation (RSD) on the migration times and the corrected-peak areas. Corrected-peak area (CPA) represents peak area corrected for migration time. In fact, in CE any ﬂuctuation in migration time will cause error in peak area determination unless this latter is normalized for migration time. 2.4. Enzymatic assays: pre-capillary electrophoresis Myrosinase catalyzes the hydrolysis of glucosinolates (GLS) to isothiocyanate (ITC), glucose and sulfate. The myrosinase activity was determined by following the hydrolysis of the glucosinolate substrate. The SO4 2− produced was detected by the C4 D and quantiﬁed. Unless otherwise stated, the reaction mixture was of 100 L volume. It was done in a micro-vial of the CE instrument auto-sampler. It contained the myrosinase (0.05 U mL−1 ) and its substrate at different concentrations in the incubation phosphate buffer (pH 7.0). The incubation time must be ﬁxed so that the substrate conversion does not exceed 10%. Unless otherwise stated, the mixture was incubated for 5 min at 37 ± 1 ◦ C. The reaction was stopped by heating the mixture with a sand bath at 125 ◦ C ± 5 ◦ C for 5 min. The reaction mixture was then injected into the capillary at the cathode (short-end injection) by applying a pressure of −50 mbar for 10 s (24.6 nL). The electrophoretic separation was performed at 25 ◦ C under +20 kV in the His/AcOH buffer. A blank reaction was conducted according to the same procedure without adding the myrosinase. Moreover, it is important to note that the solutions used to assess the calibration curve for SO4 2− underwent the same treatments in order to avoid any error due to evaporation in the sand bath. The initial reaction rate Vi (mM min−1 ) was ﬁrst calculated as the ratio of the concentration of SO4 2− formed per time interval. The nonlinear curve ﬁtting program PRISM® 5.04 (GraphPad, San Diego, CA, USA) was used to determine Km and Vmax according to equation (1): Vi =
Vmax × [S] Km + [S]
where Vi = reaction rate, Km = Michaelis–Menten constant, Vmax = maximum reaction velocity and [S] = substrate (GLS) concentration.
R. Nehmé et al. / Analytica Chimica Acta 807 (2014) 153–158
3. Results 3.1. Preliminary study Before conducting any enzyme reaction, the CE/C4 D analysis and detection of SO4 2− was optimized. Initially, the background electrolyte His/AcOH at low ionic strength of 10 mM was used. When Na2 SO4 dissolved at 0.035 mM SO4 2− in the incubation phosphate buffer (I = 33.1 mM) was analyzed, SO4 2− was detected without any use of an electroosmotic ﬂow modiﬁer. However, the peak of SO4 2− was distorted and of low intensity. This is certainly due to the low ionic strength of the BGE (10 mM) compared to the incubation buffer in which the analyte were dissolved. Indeed, when Na2 SO4 was dissolved in water, a symmetrical intense peak was obtained for SO4 2− . For this, His/AcOH at higher ionic strength 40 mM was then used. In this case, good results were obtained for the analysis of SO4 2− dissolved in the incubation phosphate buffer. Fast analyses were obtained with SO4 2− detected at 1.02 min. The repeatability on the migration time (RSDtm < 0.5%) and on the corrected-peak area (RSDCPA < 2%) was excellent. Good peak-symmetry was obtained (As = 1.1) and the limit of quantiﬁcation (LOQ) was found to be 15 M (S N−1 = 10). The response function of the C4 D was examined over the concentration range (0.015 mM–15 mM). It was proved to be linear all over this range with r2 ≥ 0.998. A calibration curve was plotted before each kinetic assay to determine the SO4 2− concentration formed during the enzymatic reaction. Then, the myrosinase hydrolysis of the GLS sinigrin was followed. A 5 min incubation time was chosen for myrosinase assays. To be sure that in these conditions the velocity was in the linear domain, the hydrolysis of the sinigrin by the myrosinase was followed by detecting SO4 2− . Indeed, the reaction was proved to be linear up to at least 45 min of incubation time. A range of sinigrin concentration from 0.01 mM to 1.5 mM was used. The kinetic parameters of the myrosinase reaction were determined by nonlinear ﬁtting of the corrected-peak area of SO4 2− as a function of sinigrin concentration (n = 3; 95% conﬁdence interval). First of all, the volume of the reaction mixture was ﬁxed to 400 L and the incubation was done at 30 ◦ C (Fig. 2(1)). Vmax obtained in these conditions was 51100 ± 1321 mM min−1 and Km was 0.205 ± 0.018 mM (r2 = 0.9911). In order to increase the enzymatic activity, the incubation was then realized at 37 ◦ C (Fig. 2(2)). Indeed, Km was decreased to 0.102 ± 0.010 mM (r2 = 0.9860). The Km for sinigrin in similar incubation conditions (37 ◦ C) is reported in literature using a spectrophotometric assay; Km = 0.156 ± 0.004 mM . It can be seen that the Km obtained by the CE/C4 D approach is in the same range. In an attempt to economize the reactants, the incubation volume was then decreased from 400 L to 100 L (vs. ∼ 1500 L in the literature ). Fig. 3 shows the electropherograms obtained in these conditions for monitoring myrosinase hydrolysis of sinigrin at different concentrations. Km obtained (Fig. 2(3)) was similar to the one obtained using the conventional spectrophotometric method [6,30] which conﬁrms the reliability of the results obtained by the CE approach (Km = 0.107 ± 0.011 mM (r2 = 0.9818)). Thus, the Km for sinigrin was successfully determined by CE/C4 D.
Fig. 3. Electropherograms obtained for the myrosinase hydrolysis of sinigrin. Reaction mixture (100 L total volume): incubation buffer KH2 PO4 /K2 HPO4 (I = 33.1 mM; pH 7.0); myrosinase (0.05 U mL−1 ); sinigrin (0.01–1.5 mM). Incubation at 37 ◦ C. For other conditions see Fig. 2.
To conclude, the myrosinase hydrolysis of sinigrin could be followed by CE/C4 D in less than 10 min using 100 L reaction mixture (vs. ∼4 h and ∼1500 L in literature ). This approach is very simple, economic, fast and robust (Table 2). It necessitates two steps: (1) incubation of the reaction mixture in a micro-vial; and (2) electrophoretic separation in the CE capillary for detecting and quantifying the product SO4 2− . 3.2. Kinetic study of myrosinase-Km determination for the screening of myrosinase substrates Ten glucosinolates, six natural and four synthetic, were screened as myrosinase substrates using the CE/C4 D approach developed above. The Michaelis–Menten constant value (Km ) was determined for each which allowed the classiﬁcation of these substrates as shown in Table 1. Natural GLS 1–5 showed a similar pattern of hydrolysis with myrosinase. The Km constants determined were in the same range of value as the Km of sinigrin. Between these ﬁve substrates the aromatic glucosinalbin 1 and glucotropaeolin 2 are slightly more rapidly hydrolyzed than the aliphatic ones 3–5 (including sinigrin). Introducing a more bulky and functionalized aglycon as for glucomoringin 6 (Km = 2.2 mM) with the ␣-l-rhamnopyranoside sugar moiety hampered the hydrolysis by a 37 fold factor compared to the parent structure sinalbine 1. As expected, from previous results observed by Mays et al.  introducing bulkiness resulted in a decrease of myrosinase activity especially with the artiﬁcial glucosinolates. Thus naphthylmethyl GLS 7, carbazole GLS 8 and phenanthrenylmethyl GLS 10 showed a common behavior with Km constants in the same range of value. Contrary to what would have been expected, a slight change of hydroxyl position on the aromatic ring from para (glucosinalbin 1, Km = 72 M) to ortho (orthosinalbin 9, Km = 16 mM) engendered an impressive gap of more than 200 fold. This result is directly linked to the modiﬁed position of
Table 2 Comparison of the CE/C4 D approach with the classical spectrophotometric technique  for following myrosinase activity; application to sinigrin (the natural substrate). Technique
Reaction mixture volume (L)
Enzyme concentration (U mL−1 )
Incubation time (min)
Overall analysis time (min)
Km for Sinigrin (mM)
Photometric  Pre-capillary electrophoresis
0.156 ± 0.004 0.107 ± 0.011 (r2 = 0.9818)
∼Incubation time. Rinse time (3 min) + separation/detection time.
R. Nehmé et al. / Analytica Chimica Acta 807 (2014) 153–158
the hydroxyl group which probably impacts the recognition pattern of the enzymatic site, either through a modiﬁed interaction with the enzyme or through the possible intra-molecular hydrogen bond formation with the C-2 hydroxyl of the glucopyrano unit. The X-Ray analysis of the benzyl aglycon interactions of 3 different inhibitors, with the myrosinase of Sinapis alba showed a clear map of hydrophobic interactions between the phenyl group and the residues Tyr 330, Phe 331 and Phe 371. The ortho position of the hydroxyl group might distort these hydrophobic contacts. Furthermore, modelisation of orthosinalbin 9 showed the possibility to have an intra-molecular hydrogen bond between the phenolic hydroxyl and the sulfate group, thus hampering the interaction with the enzymatic pocket by modifying the ideal conformation of the substrate [32,33].
References            
This study shows the use of capillary electrophoresis with contactless conductivity detection for the monitoring of the myrosinase activity. This enzymatic system was not yet studied by capillary electrophoresis. Myrosinase kinetics, maximum velocity Vmax and Michaelis–Menten constant Km were evaluated for a variety of natural and synthetic glucosinolates not yet studied by conventional methods. This allowed, for the ﬁrst time, the classiﬁcation of these substrates according to their afﬁnity for myrosinase. Results compared very well with the few results reported using the conventional method. Comparing to the conventional method, the CE/C4 D method developed in this work is simple, automated, economic (few tens of microliters of reactants used) and rapid (incubation for few minutes). Several novelties in our structure activity relation (SAR) studies were found especially regarding to the position of the hydroxyl on the aromatic ring. This novel CE method can be expanded to assess the afﬁnity of different substrates and even inhibitors against myrosinase.
                   
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.aca.2013.11.012.
J.W. Fahey, A.T. Zalcmann, P. Talalay, Phytochemistry 56 (2001) 5. V. Gil, J. McLeod, Phytochemistry 19 (1980) 2547. M. Traka, R. Mithen, Phytochem. Rev. 8 (2009) 269. ˇ D. Cerniauskaitë, J. Rousseau, A. Saˇckus, P. Rollin, A. Tatibouët, Eur. J. Org. Chem. 2011 (2011) 2293. P. Rollin, A. Tatibouët, C. R. Chimie 14 (2011) 194. S. Palmieri, O. Leoni, R. Iori, Anal. Biochem. 123 (1982) 320. S. Palmieri, R. Iori, B. Joseph, P. Rollin, Proceedings of the Glucosinolate Colloquium, Ardon 13–14 January, 1993, p. 57. P.L. Durham, J.E. Poulton, Z. Naturforsch, J. Biol. Sci. 45 (1990) 173. M.G. Botti, M.G. Taylor, N.P. Botting, J. Biol. Chem. 270 (1995) 20530. H. Nehmé, R. Nehmé, P. Laﬁte, S. Routier, P. Morin, Anal. Chim. Acta 722 (2012) 127. H. Nehmé, R. Nehmé, P. Laﬁte, S. Routier, P. Morin, J. Sep. Sci. 36 (2013) 2151. H. Nehmé, R. Nehmé, P. Laﬁte, S. Routier, P. Morin, J. Chromatogr. A 1314 (2013) 298. J.P. Landers, Introduction to Capillary Electrophoresis in Capillary and Microchip Electrophoresis and Associated Microtechniques, CRC Press Taylor & Francis Group, New York, 2008. K. Altria, M. Kelly, B. Clark, Chromatographia 43 (1996) 153. P. Jandik, W.R. Jones, J. Chromatogr. 546 (1991) 431. W. Beck, H. Engelhardt, Chromatographia 33 (1992) 313. X. Huang, R.N. Zare, Anal. Chem. 63 (1991) 2193. R.C. Williams, R. Boucher, J. Brown, J.R. Scull, J. Walker, D. Paolini, J. Pharm. Biomed. Anal. 16 (1997) 469. A.J. Zemann, E. Schnell, D. Volgger, G.K. Bonn, Anal. Chem. 70 (1998) 563. J.A. Fracassi da Silva, C.L. do Lago, Anal. Chem. 70 (1998) 4339. P. Kubán, P. Kubán, V. Kubán, Electrophoresis 24 (2003) 1935. R. Nehmé, A. Lascaux, R. Delépée, B. Claude, P. Morin, Anal. Chim. Acta 663 (2010) 190. V. Unterholzner, M. Macka, P.R. Haddad, A. Zemann, Analyst 127 (2002) 715. ˇ P.C. Hauser, Electrophoresis 28 (2007) 4690. A. Schuchert-Shi, P. Kubán, A. Schuchert-Shi, P.C. Hauser, Anal. Biochem. 376 (2008) 262. A. Schuchert-Shi, P.C. Hauser, Chirality 22 (2010) 331. A. Schuchert-Shi, P.C. Hauser, Electrophoresis 30 (2009) 3442. A. Schuchert-Shi, P.C. Hauser, Anal. Biochem. 387 (2009) 202. C. Haber, R.J. Van Saun, W.R. Jones, Anal. Chem. 70 (1998) 2261. A. Hochkoeppler, S. Palmieri, Biotechnol. Progr. 8 (1992) 91. J.R. Mays, R.L. Weller Roska, S. Sarfaraz, H. Mukhtar, S.R. Rajski, Chem. Bio. Chem. 9 (2008) 729. A. Bourderioux, M. Lefoix, D. Gueyrard, A. Tatibouët, S. Cottaz, S. Arzt, W.P. Burmeister, P. Rollin, Org. Biomol. Chem. 3 (2005) 1872. ˇ A. Besle, X. Brazzolotto, A. Tatibouët, D. Cerniauskaite, E. Gallienne, P. Rollin, W.P. Burmeister, Acta Cryst. 66 (2010) 152.