Journal of Chromatography B, 941 (2013) 62–68
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Determination of gibberellins in soybean using tertiary amine labeling and capillary electrophoresis coupled with electrochemiluminescence detection Guimei Zhu, Shihua Long, Hao Sun, Wen Luo, Xia Li, Zaibin Hao ∗ College of Chemistry and Bioengineering, Guilin University of Technology, Guilin 541004, China
a r t i c l e
i n f o
Article history: Received 15 April 2013 Accepted 2 October 2013 Available online 11 October 2013 Keywords: Capillary electrophoresis Electrochemiluminescence Gibberellin A3 Soybean Tertiary amine labeling
a b s t r a c t A novel sensitive method based on tertiary amine labeling for the analysis of gibberellins (GAs) by capillary electrophoresis (CE) coupled with electrochemiluminescence (ECL) detection was proposed. GA3 was tagged with 2-(2-aminoethyl)-1-methylpyrrolidine (AEMP) using N, N -dicyclohexylcarbodiimide (DCC) and 3,4-dihydro-3-hydroxy-4-oxo-1,2,3-benzotriazine (HOOBt) as coupling agents in acetonitrile to produce GA3 -AEMP-derivative. The GA3 -AEMP-derivative was injected into CE by electrokinetic injection and detected by Ru(bpy)3 2+ -based ECL. The parameters affecting derivatization, detection and separation such as concentration of reactants, detection potential, pH and concentration of separation buffer, were investigated in detail. Under optimum conditions, the linear concentration range for GA3 was from 2.0 × 10−7 to 1.28 × 10−4 M with a correlation coefficient of 0.9997. The detection limit was 8 × 10−8 M (S/N = 3). The relative standard deviations of migration time, peak intensity and peak area for nine continuous injections of 2.0 × 10−5 M GA3 -AEMP-derivative were 1.0%, 2.1% and 4.2%, respectively. The developed approach was successfully applied to the determination of total GAs in the stem, leaf and seed of soybean (Glycine max [L.] Merr.) with recoveries in the range from 89.6% to 99.3%. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Gibberellins (GAs), one of the five ‘classic’ plant hormones, are a large family of tetracyclic diterpenoid carboxylic acids which consist of 136 members from GA1 to GA136 and usually present in plants, fungi and bacteria [1,2]. GAs play an important role in controlling plant growth and development by regulating seed germination, stem elongation, leaf expansion, floral initiation, sex expression, fruit development, dormancy, etc. [3–11]. Since gibberellin A3 (GA3 ), which is one of the most active GAs, can promote plant growth and development, it is considered as a high-value industrial product and widely used to improve agricultural production and quality [12]. However, it could threaten the health and safety of consumers as result of extensive use [13]. On the other hand, nowadays, the determination of plant hormones involving GAs has become one of the bottlenecks for their study because of the extremely low concentrations (generally at ng/g and even pg/g levels) in plants and the lack of sensitive detection techniques. Since GAs have little ultraviolet (UV) absorption, no fluorescence and no other distinguishing chemical characteristics essential to a specific chemical assay, quantitative analysis of GAs
∗ Corresponding author. Tel.: +86 15977348610. E-mail address: [email protected] (Z. Hao). 1570-0232/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jchromb.2013.10.004
in the complicated plant matrixes is always very difficult [13,14]. As a result, it is extremely important to establish a more simple, sensitive and effective method for the detection of GAs. Until now, many methods have been developed for GAs, including enzyme-linked immunosorbent assay (ELISA) [15], bioassay [16], fluorescence [17], biosensor [18], high performance liquid chromatography (HPLC) [13,19–22], gas chromatography (GC) [23], capillary electrophoresis (CE) [24,25]. In recent years, HPLC, GC and CE in conjunction with different detectors have attracted much attention of researchers. Although HPLC and GC coupled with mass spectrometry are the most powerful and sensitive analytical methods for the analysis of GAs, they also suffer from drawbacks such as time-consuming analysis, expensive equipments. However, CE coupled with electrochemiluminescence (ECL) detection possesses the combined advantages of high efficiency, good resolution, short analysis time, less sample volume, high sensitivity, good selectivity, wide linear range, simple operation and low equipment cost. Ru(bpy)3 2+ -based ECL has received considerable attention among a number of ECL systems, owing to its low background signal, excellent chemical stability and high ECL efficiency in aqueous and non-aqueous solvents [26–28]. It has been used for determination of drugs, alkaloids, herbicides and other analytes [26,29–35]. Those analytes containing a tertiary amine group can theoretically initiate the Ru(bpy)3 2+ -based ECL reaction, which makes the analysis of them possible. In addition, tertiary amine labeling technique has
G. Zhu et al. / J. Chromatogr. B 941 (2013) 62–68
63
been developed to extend ECL method for the detection of analytes which are not ECL-active [35–38]. In this work, a sensitive CE-ECL method based on 2-(2aminoethyl)-1-methylpyrrolidine (AEMP) labeling was proposed for the analysis of GAs for the first time. The developed method was applied to determination of total GAs in the stem, leaf and seed of soybean (Glycine max [L.] Merr.) after the crude GAs extracts had been labeled with AEMP. The factors influencing derivatization, detection and separation were investigated.
125 ± 5 m with the aid of an optical microscope. 10 mM phosphate buffer at pH 5.7 was used as separation buffer. Analytes were electrokinetically injected into CE at 10 kV for 10 s. The separation voltage was 15 kV. The PMT voltage was set at −800 V to collect the ECL signal and the detection potential was fixed at 1.20 V. A 250 L of 5 mM Ru(bpy)3 2+ in 50 mM phosphate buffer at pH 7.0 was placed in the ECL detection cell for ECL measurement. The Ru(bpy)3 2+ -phosphate solution was refreshed every 3 h so as to obtain good reproducibility.
2. Materials and methods
2.3. Synthesis of AEMP-derivative of GA3
2.1.1. Reagents and chemicals
As shown in Fig. 1, the synthesis of AEMP-derivative of GA3 was carried out as described by Yin et al. [35] and Morita et al. [37] with some modifications. The process was as follows: GA3 (4.2 mg), DCC (2.6 mg) and HOOBt (1.1 mg) were dissolved in 2.0 mL of acetonitrile, followed by being shaken for 10 min, and then 1.5 L of AEMP was added into the mixture. The resultant mixture was incubated for 24 h at room temperature. The reaction mixture was filtered and the filtrate was evaporated to dryness in vacuo. The residue was extracted into ethyl acetate. The organic layer was dried with anhydrous Na2 SO4 and filtered. Finally, the filtrate was evaporated in vacuo. The prepared product was used as the standard to validate and quantify GAs in samples.
GA3 (purity 99%) was purchased from Acros Organics (Geel, Belgium, NJ, USA). GA4 and GA7 were bought from J&K Chemical Ltd. (Beijing, China). Tris(2,2 -bipyridyl) ruthenium(II) chloride hexahydrate (Ru(bpy)3 Cl2 ·6H2 O) was obtained from Sigma–Aldrich Inc. (St. Louis, MO, USA). 2-(2-aminoethyl)-1-methylpyrrolidine (AEMP), N,N -dicyclohexylcarbodiimide (DCC) and 3,4-dihydro-3hydroxy-4-oxo-1,2,3-benzotriazine (HOOBt) from Aladdin reagent Co., Ltd. (Shanghai, China), were used for the derivatization of GA3 . Acetonitrile and methanol, HPLC grade, were acquired from Fisher Scientific Co. (Fair Lawn, NJ, USA). All other reagents and chemicals, analytical grade, were bought from Xilong Chemical Co., Ltd. (Guangdong, China). Except otherwise stated, all CE solutions were dissolved in ultrapure water processed with Molecular Lab Water Purifier (Molecular Instrument Shanghai Co., Ltd., Shanghai, China) and stored in the refrigerator at 4 ◦ C before use. The working standard solutions and sample solutions were freshly prepared by precisely diluting stock solution with proper buffer. The required sample solutions and buffer solutions were filtered through 0.22 m membrane filters (Shanghai Xinya Purification Material Factory, Shanghai, China) prior to CE-ECL analysis. 2.2. Apparatus All separations and analyses were performed with a CE-ECL detection system (Model MPI-B, Xi’an Remex Analysis Instrument Co., Ltd., Xi’an, China), including a high-voltage power supplier for electrokinetic injection and separation, an electrochemical potentiostat, a multifunctional chemiluminescence detector and a data collection analyzer. A three-electrode system used in chemiluminescence detection system was composed of a 500 m diameter Pt disk as the working electrode, an Ag/AgCl electrode as the reference electrode and a Pt wire as the auxiliary electrode. The surface of the working electrode was polished in sequence with 1.0, 0.3 and 0.05 m alpha alumina powder (CH Instruments, Inc., Austin, TX, USA) and then was cleaned with ultrapure water in a KQ5200DB ultrasonic cleaner (Kun Shan Ultrasonic Instruments Co., Ltd., Jiangsu, China) before use. A photomultiplier tube (PMT) was used to collect the ECL signal. The PMT signal was amplified and recorded using the MPI-B software. Cleanert C18-SPE cartridges (200 mg/3 mL, Bonnal-Agela Technologies, Tianjin, China) and a rotary evaporator (Model RE-52A, Shanghai Yarong Biochemical Instrument Factory, Shanghai, China) were used for sample pretreatment. An uncoated fused silica capillary (50 m i.d., 375 m o.d.) of 42 cm length (Yongnian Reafine Chromatography Ltd., Hebei, China) was used as electrophoretic separation channel. The new capillary was filled with 0.1 M NaOH overnight before first use. The capillary was rinsed with 0.1 M NaOH, ultrapure water and the corresponding separation buffer for 10 min, respectively, to maintain an active and reproducible inner surface prior to a series of analyses. During the entire experiment, the distance between the working electrode and the outlet of the capillary was fixed at
2.4. Sample pretreatment The samples of stem, leaf and seed were collected from the soybean (Glycine max [L.] Merr.) cultivar in the laboratory of corresponding author. All samples were ground to a homogenous powder. Subsequently, 2.00 g of powdered plant material was accurately weighed and extracted with 30 mL of 80% methanol in the dark at 4 ◦ C for 12 h. The extract was filtered and the filtrate was transferred into a flask. The extraction was repeated again for 2 h to increase the extraction yield and the extracted methanol solutions were combined. The combined extracts were evaporated using a rotary evaporator at 45 ◦ C. However, the above crude extract contained many interfering components. Liquid–liquid extraction (LLE) combined with solid-phase extraction (SPE) was applied to clean up the sample and enrich GAs in this experiment. The pH value of residues was adjusted to 3 with acetic acid. The residues were further extracted with 30 mL of ethyl acetate. The extraction was repeated two additional times with 20 mL and 10 mL of ethyl acetate, respectively. The ethyl acetate extracts were combined and evaporated to nearly dryness using a rotary evaporator at 45 ◦ C. The dried extract was then dissolved in 10% methanol and centrifuged at 10,000 rpm for 5 min at 4 ◦ C. The supernatant was further purified using SPE with the Cleanert C18-SPE cartridge, in which 1 mL of methanol, 1 mL of water and 5 mL of 40% methanol were used as purificant, activator and eluent, respectively. The eluates were evaporated to dryness under the same conditions. The dried residues were dissolved in acetonitrile and mixed with 128 L of 10.0 × 10−2 M DCC, 165 L of 4.0 × 10−2 M HOOBt and 318 L of 3.6 × 10−2 M AEMP for AEMP labeling in the same manner. The derivatives were dissolved in methanol to 1 mL. Then 0.125 mL of derivatization solution was diluted with separation buffer to 10 mL and filtered through a 0.22 m membrane filter for CE-ECL detection. 3. Results and discussion 3.1. ECL behavior of Ru(bpy)3 2+ in the presence of GA3 -AEMP-derivative In the CE-ECL system, the ECL intensity depends upon the rate of the light-emitting chemical reaction which is in turn dependent
64
G. Zhu et al. / J. Chromatogr. B 941 (2013) 62–68
Fig. 1. Labeling of GA3 with AEMP using DCC and HOOBt as coupling agents.
on the potential applied to the electrode [26,33]. Fig. 2A shows cyclic voltammograms of 5 mM Ru(bpy)3 2+ in 50 mM phosphate buffer (pH 7.0) before and after adding GA3 , GA3 -AEMP-derivative and AEMP, respectively. The oxidation peak currents of GA3 , GA3 -AEMP-derivative and AEMP could be distinguished from the background current. The corresponding ECL curves were illustrated in Fig. 2B. Compared with the background curve (curve a in Fig. 2B), curve b in Fig. 2B revealed that there was a little increase in the ECL intensity in the presence of GA3 . As expected, there was an obvious increase of the ECL intensity when GA3 was tagged with AEMP (curve c in Fig. 2B). Nevertheless, the ECL intensity of GA3 -AEMP-derivative was still lower than that of AEMP (curve d in Fig. 2B). The observations indicated that GA3 -AEMP-derivative could react with Ru(bpy)3 2+ in the electrochemiluminescence process and enhanced the emitted light intensity. It was due to the AEMP derivative conserved the intensity of the original molecule. The results suggested that the proposed CE-ECL method could be used for the sensitive determination of GA3 . 3.2. Labeling of GA3 with AEMP Generally speaking, the ECL intensity produced by tertiary amines is the highest, followed by secondary and finally primary amines in the ECL system based on Ru(bpy)3 2+ [28,34]. As
a result, the detection sensitivity of compounds can be enhanced by changing their structures or derivatizing with different ECLactive species into tertiary amines. Given that GA3 contains a carboxylic acid group, the potential derivatization reagents containing both a primary amine group and a tertiary amine group will be an excellent choice. It is because many approaches can be applied to couple these two groups with high efficiency under mild condition. AEMP was found to be an ECL-active compound [35–37]. Therefore, AEMP was chosen as the derivatization reagent for labeling of GA3 in this study. In addition, two common systems, namely 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC)/N-hydroxysuccinimide (NHS) system and DCC/HOOBt system, can be used as coupling agents to label a carboxylic acid group with a primary amine group in organic solvents. Nevertheless, the EDC/NHS system will result in some ECL-active species, which could interfere with the ECL detection of GA3 [35]. Hence, the DCC/HOOBt system was adopted. Fig. 1 displays the molecular structure of AEMP and the process of synthesis. It is necessary to select a proper solvent for derivatization because GA3 -AEMP-derivative should be synthesized in a suitable organic solvent, in which reactants dissolve well and GA3 is labeled with AEMP with high efficiency. Given that dioxane is highly toxic, acetonitrile, ethyl acetate, chloroform and DMSO were used as candidates for derivatization solvents. The results indicated that the highest derivatization efficiency was achieved by using acetonitrile as the derivatization solvent. Consequently, acetonitrile was selected as the optimum derivatization solvent. The concentration of DCC, HOOBt and AEMP also greatly influences the coupling of GA3 to AEMP. Fig. 3 shows the effect of the concentration of DCC, HOOBt and AEMP on the ECL peak area of 2 × 10−5 M GA3-AEMP-derivtive. The maximum peak area was obtained at 6.4 mM for DCC, 3.3 mM for HOOBt, 5.8 mM for AEMP. Therefore, 6.4 mM DCC, 3.3 mM HOOBt and 5.8 mM AEMP were chosen as their optimum concentrations for AEMP labeling. 3.3. Selection of solvent of GA3 -AEMP-derivative
Fig. 2. Cyclic voltammograms (A) and their corresponding ECL curves (B). (a) 5 mM Ru(bpy)3 2+ + 50 mM phosphate buffer (pH 7.0); (b) 5 mM Ru(bpy)3 2+ + 50 mM phosphate buffer (pH 7.0) + 2 × 10−5 M GA3 ; (c) 5 mM Ru(bpy)3 2+ + 50 mM phosphate buffer (pH 7.0) + 2 × 10−5 M GA3 -AEMP-derivative; (d) 5 mM Ru(bpy)3 2+ + 50 mM phosphate buffer (pH 7.0) + 2 × 10−5 M AEMP; scan rate, 100 mV/s.
The sample should be dissolved well in an appropriate solvent. In previous works, acetonitrile was used as an effective additive in CE separation because it can provide good resolution and repeatability for determination of similar compounds [39–41]. Nonetheless, high concentrations of acetonitrile will lead to bubble formation and thus may decrease CE separation. In this work, acetonitrile and methanol were used as candidates for solvents of GA3 -AEMP-derivative. At last, methanol was found to give both higher ECL intensity and better ability to dissolve
G. Zhu et al. / J. Chromatogr. B 941 (2013) 62–68
65
Fig. 3. Effect of concentration of DCC (A), HOOBt (B) and AEMP (C) on the ECL peak area of 2 × 10−5 M GA3 -AEMP-derivative. Detection conditions: in ECL cell, 50 mM phosphate buffer (pH 7.5) containing 5 mM Ru(bpy)3 2+ ; detection potential, 1.20 V; separation buffer, 10 mM phosphate buffer (pH 5.7); separation voltage, 15 kV; injection voltage, 10 kV; injection time, 10 s; sample solvent, acetonitrile.
GA3 -AEMP-derivative. Moreover, better peak shape was obtained by using methanol as sample solvent. So methanol was used as the solvent of GA3 -AEMP-derivative. Furthermore, even though the methanol solution of GA3 -AEMP-derivative was stored in the dark for 30 days in the refrigerator at 4 ◦ C, the CE-ECL analysis was nearly identical to that of freshly prepared solution. The results showed that the GA3 derivative was stable.
varied from 6.0 to 7.0 and decreased at higher pH values. Moreover, high pH value resulted in enhancement of background noise and thus led to less sensitive and stable analysis. The results were similar to previous works [35,37,38,41]. Finally, 50 mM phosphate buffer at pH 7.0 was used as the detection buffer in the experiment.
3.4. Optimization of CE-ECL detection conditions
3.5.1. Effect of pH and concentration of separation buffer The composition of separation buffer impacts on the ECL intensity and the separation efficiency by influencing the zeta potential
3.4.1. Effect of detection potential The electrochemiluminescence reaction depends on the potential applied to the electrode, so the relationship between ECL intensity and detection potential was investigated in the potential range of 1.00–1.30 V. It was found that the ECL intensity first increased and then decreased with the increase of detection potential. The maximum ECL intensity was obtained at a potential of 1.20 V. Therefore, 1.20 V was chosen as the detection potential in the following experiments. 3.4.2. Effect of concentration and pH of detection buffer The concentration of detection buffer is crucial for the ECL detection. Over the detection buffer concentration range of 20–60 mM, the maximum ECL intensity was achieved at 50 mM. So the phosphate buffer concentration was set at 50 mM. The buffer pH value in ECL detection cell affects the sensitivity and stability of determination of analytes because the reaction process relates to the deprotonation. The effect of pH value on ECL intensity was studied in the range of 6.0–9.0 by fixing the concentration of the detection buffer at 50 mM. As displayed in Fig. 4, the ECL intensity increased when the pH value of detection buffer
3.5. Optimization of separation conditions
Fig. 4. Effect of detection buffer pH on ECL intensity of 2 × 10−5 M GA3 -AEMPderivative. Detection conditions: sample solvent, methanol; other conditions were the same as those in Fig. 3.
66
G. Zhu et al. / J. Chromatogr. B 941 (2013) 62–68
Fig. 5. Effect of separation buffer pH on ECL intensity of 2 × 10−5 M GA3 -AEMPderivative. Detection conditions: in ECL cell, 50 mM phosphate buffer (pH 7.0) containing 5 mM Ru(bpy)3 2+ ; sample solvent, methanol; other conditions were the same as those in Fig. 3.
(), the electroosmotic flow (EOF) and the charge of analytes. 10 mM phosphate buffer was tested over the pH range of 4.0–8.0 (Fig. 5). The ECL intensity of GA3 -AEMP-derivative increased rapidly in the pH range of 4.0–5.7, but the intensity decreased when the pH value was higher than 5.7. At the same time, the migration time was shortened because of the increasing EOF. Therefore, the phosphate buffer at pH 5.7 was chosen for further experiments. The concentration of separation buffer is also one of the most crucial factors in the CE-ECL detection. The ECL intensity first ascended and then descended with the increase of buffer concentration in the rage of 5–20 mM. The maximum ECL intensity was observed when the concentration of phosphate buffer was 10 mM. Thereby, 10 mM phosphate buffer at pH 5.7 was used as separation buffer for the subsequent study. 3.5.2. Effect of separation voltage It is essential to optimize the separation voltage because of its influence on the migration time of analytes and the number of theoretical plates (N) by driving samples through the capillary. The ECL intensity kept increasing when the separation voltage was augmented from 8 to 18 kV in this investigation. The ECL response obviously increased when the separation voltage was changed from 8 to 15 kV, and the increasing trend became slower above 15 kV. In the meantime, the migration time decreased and the level of baseline noise increased with increasing separation voltage due to the increase of Joule heating in the capillary. Furthermore, the fast flow rate might reduce the concentration of Ru(bpy)3 3+ at the working electrode surface, which in turn will cause lower efficiency of light emission. Taking into consideration ECL intensity, migration time, baseline noise and Joule heating, a voltage of 15 kV was selected for separation. 3.5.3. Effect of injection voltage and injection time The injection parameters are major elements affecting the ECL intensity and N, too. N is calculated by the equation: N = 2 5.54(tm /W1/2 ) , where tm and W1/2 are the migration time and the width at half–maximum peak height, respectively. In general, a higher injection voltage and a longer injection time can cause stronger ECL signal but lower separation efficiency [32,33,42]. Their effects on the ECL intensity and N were studied by altering injection voltage when injection time was invariable (Fig. 6A) and changing injection time while injection voltage was kept constant (Fig. 6B). As predicted, both a higher injection voltage and a longer injection time gave rise to higher ECL intensity but lower N and broadening
Fig. 6. Effects of injection voltage (A) and injection time (B) on ECL intensity (a) and N (b) of 2 × 10−5 M GA3 -AEMP-derivative. Detection conditions: in ECL cell, 50 mM phosphate buffer (pH 7.0) containing 5 mM Ru(bpy)3 2+ ; sample solvent, methanol; other conditions were the same as those in Fig. 3.
of ECL peak as a result of the dispersion of analytes. Moreover, the ECL intensity slowly increased when injection voltage was higher than 10 kV and injection time exceeded 12 s. This was due to more analyses were involved in ECL reaction using a higher injection voltage or a longer injection time. Given that an electrokinetic injection duration of 10 s at 10 kV could offer better peak shape than that of duration of 12 s, 10 kV and 10 s were used as the optimal injection voltage and injection time, respectively. 3.6. Linearity, detection limit and reproducibility Under the optimized conditions, the linear range, detection limit and reproducibility were studied. The calibration curve was linear between 2.0 × 10−7 and 1.28 × 10−4 M for GA3 . The calibration equation and correlation coefficient were: y = 702.07(10−7 M) + 1905.7 and r = 0.9997 in terms of peak area response as a function of GA3 concentration. The limit of detection for GA3 was 8 × 10−8 M (S/N = 3), which is higher than those obtained using biosensor [18], HPLC–MS [20] and CE-LIF [25], but lower than that obtained using CE-MS [24]. Furthermore, the linear range of this method is wider than that of the above methods. The relative standard deviations of migration time, peak intensity and peak area for nine continuous injections of 2.0 × 10−5 M GA3 -AEMP-derivative were 1.0%, 2.1% and 4.2%, respectively. 3.7. Interference study Many molecules in plant tissues may interfere with the detection of GAs. Six amino acids, tyrosine (Tyr), proline (Pro),
G. Zhu et al. / J. Chromatogr. B 941 (2013) 62–68
67
Table 1 Analytical results and recoveries of GAs in samples (n = 3). Sample
GAs content (103 g/kg)
Added (103 g/kg)
Found (103 g/kg)
Recovery (%)
RSD (%)
Stem Leaf Seed
2.2637 0.2983 0.3370
3.8000 3.8000 3.8000
6.0375 3.7035 4.0502
99.3 89.6 97.7
0.7 3.5 0.4
methionine (Met), cysteine (Cys), glycine (Gly) and histidine (His) were selected to assess the potential interference for GAs in the proposed method. These amino acids contain a carboxylic group, which enables them to be coupled to AEMP. 4 mg of individual Tyr, Pro, Met, Cys, Gly and His were added to the standard solution to study their effects on the determination of GAs. Nevertheless, no interference was observed because they are hardly dissolved in acetonitrile and high soluble in aqueous solution. Even if some species are extracted from plant tissues, they are hardly labeled with AEMP and detected by CE-ECL. Therefore, the tertiary amine labeling contributed to the high selectivity of this method for GAs. At the same time, GA4 and GA7 were added to the standard solution to study the selectivity of this method for GA3 . However, interference was observed with 10 M of individual GA4 and GA7 on the determination of GA3 and the peaks of GA3 , GA4 and GA7 after derivatization could not be successfully separated due to highly similar structures. It may because they could not be separated from each other under the proposed separation conditions. Therefore, the proposed method could be applied to the analysis of total GAs. Nevertheless, the researches of Ge [24] and Chen [25] exhibited the potential for the separation and simultaneous determination of multiple GAs using CE-ECL method by optimizing separation conditions. 3.8. Application
of the standard derivative of GA3 with those of extracts and spiked extracts as described in Fig. 7. The contents of GAs in the stem, leaf and seed of soybean (Glycine max [L.] Merr.) were 0.181, 0.024, 0.027 mg/g on an average, respectively. Recovery experiments were carried out by adding GA3 standard to the known real extracts of samples, followed by being labeled with AEMP and analyzed by the established method (Table 1). The recovery of GAs was in the range of 89.6%–99.3%. 4. Conclusions This is the first report to describe the determination of GAs by CE-ECL. All relevant operational parameters have been optimized. Under the optimum conditions, the developed CE-ECL method exhibits excellent performance with wide linear range, fast separation, high sensitivity and low detection limit. The proposed method was successfully applied to the detection of GAs in real samples. It is a valuable addition to existing analytical tools for GAs analysis. Acknowledgments This work was supported by the National Natural Science Foundation of China (Nos. 31060193 and 31200269) and the Natural Science Foundation of Zhejiang Province (No. LQ21C02004). References
The proposed CE-ECL method was applied to the analysis of total content of GAs in the extracts of stem, leaf and seed from soybean (Glycine max [L.] Merr.) under the optimized conditions. In the present work, the extracts of GAs in samples were determined after being coupled to AEMP. Identification of the GAs derivative of the extracts was confirmed by comparing the electropherogram
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]
−5
Fig. 7. CE-ECL electropherograms of 2 × 10 M GA3 -AEMP-derivative spiked with 2 × 10−5 M AEMP (A), soybean leaf (B) and soybean leaf spiked with 3.8 × 103 g/kg GA3 (C). Detection conditions: in ECL cell, 50 mM phosphate buffer (pH 7.0) containing 5 mM Ru(bpy)3 2+ ; detection potential, 1.20 V; separation buffer, 10 mM phosphate buffer (pH 5.7); separation voltage, 15 kV; injection voltage, 10 kV; injection time, 10 s; sample solvent, methanol.
[25] [26] [27] [28] [29] [30] [31] [32]
A. Piotrowska, A. Bajguz, Phytochemistry 72 (2011) 2097. X. Shan, J. Yan, D. Xie, Curr. Opin. Plant Biol. 15 (2012) 84. T. Reynolds, P. Thompson, Physiol. Plant. 28 (2006) 516. J. Zeevaart, D.A. Gage, M. Talon, Proc. Natl. Acad. Sci. USA 90 (1993) 7401. A. Maggio, G. Barbieri, G. Raimondi, S. Pascale, J. Plant Growth Regul. 29 (2009) 63. H. Magome, S. Yamaguchi, A. Hanada, Y. Kamiya, K. Oda, Plant J. 37 (2004) 720. R.G. Sharp, M.A. Else, W.J. Davies, R.W. Cameron, Sci. Hortic. 124 (2010) 231. T.D. Thomas, Euphytica 164 (2008) 317. J.C. Serrani, O. Ruiz-Rivero, M. Fos, J.L. García-Martínez, Plant J. 56 (2008) 922. A.K. Subbaraj, K.A. Funnell, D.J. Woolley, J. Plant Growth Regul. 29 (2010) 487. M. Iqbal, M. Ashraf, Environ. Exp. Bot. 86 (2013) 76. L. Sun, Y. Zhao, X. Nie, Y. Ling, X. Chu, D. Shang, Y. Dong, Chem. Res. Chin. Univ. 28 (2012) 797. W. Xie, C. Han, Z. Zheng, X. Chen, Y. Qian, H. Ding, L. Shi, C. Lv, Food Chem. 127 (2011) 890. C. Rodrigues, L.P.d.S. Vandenberghe, J. de Oliveira, C.R. Soccol, Crit. Rev. Biotechnol. 32 (2012) 263. G.K. Wang, M. Zhang, J.F. Gong, Q.F. Guo, Y.N. Feng, W. Wang, Plant Cell Rep. 31 (2012) 2215. L. Kinsman, N. Pinfield, A. Stobart, Planta 127 (1975) 149. Z. Jiang, X. Liu, A. Liang, Anal. Chim. Acta 346 (1997) 249. J. Li, S. Li, X. Wei, H. Tao, H. Pan, Anal. Chem. 84 (2012) 9951. V. Gupta, M. Kumar, H. Brahmbhatt, C.R. Reddy, A. Seth, B. Jha, Plant Physiol. Biochem. 49 (2011) 1259. Z. Zhang, W. Tan, Y. Hu, G. Li, S. Zan, Analyst 137 (2012) 968. J. Zhang, P. Du, Se Pu 29 (2011) 1133. L. Wang, Q. Wu, C. Duan, D. Wu, Y. Guan, Se Pu 29 (2011) 923. P.S. Blake, J.M. Taylor, W.E. Finch-Savage, Plant Growth Regul. 37 (2002) 119. L. Ge, C.Y. Peh, J.W. Yong, S.N. Tan, L. Hua, E.S. Ong, J. Chromatogr. A 1159 (2007) 242. H. Chen, X.F. Guo, H.S. Zhang, H. Wang, J. Chromatogr. B 879 (2011) 1802. Y. Gao, Y. Xu, B. Han, J. Li, Q. Xiang, Talanta 80 (2009) 448. X.B. Yin, E. Wang, Anal. Chim. Acta 533 (2005) 113. M. Su, W. Wei, S. Liu, Anal. Chim. Acta 704 (2011) 16. J. Yin, Y. Xu, J. Li, E. Wang, Talanta 75 (2008) 38. B. Han, Y. Du, E. Wang, Microchem. J. 89 (2008) 137. B. Yuan, C. Zheng, H. Teng, T. You, J. Chromatogr. A 1217 (2010) 171. Y. Wang, L. Xiao, M. Cheng, J. Chromatogr. A 1218 (2011) 9115.
68 [33] [34] [35] [36] [37]
G. Zhu et al. / J. Chromatogr. B 941 (2013) 62–68 B. Deng, L. Ye, H. Yin, Y. Liu, S. Hu, B. Li, J. Chromatogr. B 879 (2011) 927. B. Deng, Y. Liu, H. Yin, X. Ning, H. Lu, L. Ye, Q. Xu, Talanta 91 (2012) 128. X.B. Yin, J.M. Guo, W. Wei, J. Chromatogr. A 1217 (2010) 1399. X.B. Yin, Y. Du, X. Yang, E. Wang, J. Chromatogr. A 1091 (2005) 158. H. Morita, M. Konishi, Anal. Chem. 74 (2002) 1584.
[38] [39] [40] [41] [42]
H. Morita, M. Konishi, Anal. Chem. 75 (2003) 940. S. Bean, G. Lookhart, J. Bietz, J. Agric. Food Chem. 48 (2000) 318. Y. Walbroehl, J.W. Jorgenson, Anal. Chem. 58 (1986) 479. C. Yu, H. Du, T. You, Talanta 83 (2011) 1376. B. Deng, Y. Kang, X. Li, Q. Xu, J. Chromatogr. B 857 (2007) 136.
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Determination of gibberellins in soybean using tertiary amine labeling and capillary electrophoresis coupled with electrochemiluminescence detection Guimei Zhu, Shihua Long, Hao Sun, Wen Luo, Xia Li, Zaibin Hao ∗ College of Chemistry and Bioengineering, Guilin University of Technology, Guilin 541004, China
a r t i c l e
i n f o
Article history: Received 15 April 2013 Accepted 2 October 2013 Available online 11 October 2013 Keywords: Capillary electrophoresis Electrochemiluminescence Gibberellin A3 Soybean Tertiary amine labeling
a b s t r a c t A novel sensitive method based on tertiary amine labeling for the analysis of gibberellins (GAs) by capillary electrophoresis (CE) coupled with electrochemiluminescence (ECL) detection was proposed. GA3 was tagged with 2-(2-aminoethyl)-1-methylpyrrolidine (AEMP) using N, N -dicyclohexylcarbodiimide (DCC) and 3,4-dihydro-3-hydroxy-4-oxo-1,2,3-benzotriazine (HOOBt) as coupling agents in acetonitrile to produce GA3 -AEMP-derivative. The GA3 -AEMP-derivative was injected into CE by electrokinetic injection and detected by Ru(bpy)3 2+ -based ECL. The parameters affecting derivatization, detection and separation such as concentration of reactants, detection potential, pH and concentration of separation buffer, were investigated in detail. Under optimum conditions, the linear concentration range for GA3 was from 2.0 × 10−7 to 1.28 × 10−4 M with a correlation coefficient of 0.9997. The detection limit was 8 × 10−8 M (S/N = 3). The relative standard deviations of migration time, peak intensity and peak area for nine continuous injections of 2.0 × 10−5 M GA3 -AEMP-derivative were 1.0%, 2.1% and 4.2%, respectively. The developed approach was successfully applied to the determination of total GAs in the stem, leaf and seed of soybean (Glycine max [L.] Merr.) with recoveries in the range from 89.6% to 99.3%. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Gibberellins (GAs), one of the five ‘classic’ plant hormones, are a large family of tetracyclic diterpenoid carboxylic acids which consist of 136 members from GA1 to GA136 and usually present in plants, fungi and bacteria [1,2]. GAs play an important role in controlling plant growth and development by regulating seed germination, stem elongation, leaf expansion, floral initiation, sex expression, fruit development, dormancy, etc. [3–11]. Since gibberellin A3 (GA3 ), which is one of the most active GAs, can promote plant growth and development, it is considered as a high-value industrial product and widely used to improve agricultural production and quality [12]. However, it could threaten the health and safety of consumers as result of extensive use [13]. On the other hand, nowadays, the determination of plant hormones involving GAs has become one of the bottlenecks for their study because of the extremely low concentrations (generally at ng/g and even pg/g levels) in plants and the lack of sensitive detection techniques. Since GAs have little ultraviolet (UV) absorption, no fluorescence and no other distinguishing chemical characteristics essential to a specific chemical assay, quantitative analysis of GAs
∗ Corresponding author. Tel.: +86 15977348610. E-mail address: [email protected] (Z. Hao). 1570-0232/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jchromb.2013.10.004
in the complicated plant matrixes is always very difficult [13,14]. As a result, it is extremely important to establish a more simple, sensitive and effective method for the detection of GAs. Until now, many methods have been developed for GAs, including enzyme-linked immunosorbent assay (ELISA) [15], bioassay [16], fluorescence [17], biosensor [18], high performance liquid chromatography (HPLC) [13,19–22], gas chromatography (GC) [23], capillary electrophoresis (CE) [24,25]. In recent years, HPLC, GC and CE in conjunction with different detectors have attracted much attention of researchers. Although HPLC and GC coupled with mass spectrometry are the most powerful and sensitive analytical methods for the analysis of GAs, they also suffer from drawbacks such as time-consuming analysis, expensive equipments. However, CE coupled with electrochemiluminescence (ECL) detection possesses the combined advantages of high efficiency, good resolution, short analysis time, less sample volume, high sensitivity, good selectivity, wide linear range, simple operation and low equipment cost. Ru(bpy)3 2+ -based ECL has received considerable attention among a number of ECL systems, owing to its low background signal, excellent chemical stability and high ECL efficiency in aqueous and non-aqueous solvents [26–28]. It has been used for determination of drugs, alkaloids, herbicides and other analytes [26,29–35]. Those analytes containing a tertiary amine group can theoretically initiate the Ru(bpy)3 2+ -based ECL reaction, which makes the analysis of them possible. In addition, tertiary amine labeling technique has
G. Zhu et al. / J. Chromatogr. B 941 (2013) 62–68
63
been developed to extend ECL method for the detection of analytes which are not ECL-active [35–38]. In this work, a sensitive CE-ECL method based on 2-(2aminoethyl)-1-methylpyrrolidine (AEMP) labeling was proposed for the analysis of GAs for the first time. The developed method was applied to determination of total GAs in the stem, leaf and seed of soybean (Glycine max [L.] Merr.) after the crude GAs extracts had been labeled with AEMP. The factors influencing derivatization, detection and separation were investigated.
125 ± 5 m with the aid of an optical microscope. 10 mM phosphate buffer at pH 5.7 was used as separation buffer. Analytes were electrokinetically injected into CE at 10 kV for 10 s. The separation voltage was 15 kV. The PMT voltage was set at −800 V to collect the ECL signal and the detection potential was fixed at 1.20 V. A 250 L of 5 mM Ru(bpy)3 2+ in 50 mM phosphate buffer at pH 7.0 was placed in the ECL detection cell for ECL measurement. The Ru(bpy)3 2+ -phosphate solution was refreshed every 3 h so as to obtain good reproducibility.
2. Materials and methods
2.3. Synthesis of AEMP-derivative of GA3
2.1.1. Reagents and chemicals
As shown in Fig. 1, the synthesis of AEMP-derivative of GA3 was carried out as described by Yin et al. [35] and Morita et al. [37] with some modifications. The process was as follows: GA3 (4.2 mg), DCC (2.6 mg) and HOOBt (1.1 mg) were dissolved in 2.0 mL of acetonitrile, followed by being shaken for 10 min, and then 1.5 L of AEMP was added into the mixture. The resultant mixture was incubated for 24 h at room temperature. The reaction mixture was filtered and the filtrate was evaporated to dryness in vacuo. The residue was extracted into ethyl acetate. The organic layer was dried with anhydrous Na2 SO4 and filtered. Finally, the filtrate was evaporated in vacuo. The prepared product was used as the standard to validate and quantify GAs in samples.
GA3 (purity 99%) was purchased from Acros Organics (Geel, Belgium, NJ, USA). GA4 and GA7 were bought from J&K Chemical Ltd. (Beijing, China). Tris(2,2 -bipyridyl) ruthenium(II) chloride hexahydrate (Ru(bpy)3 Cl2 ·6H2 O) was obtained from Sigma–Aldrich Inc. (St. Louis, MO, USA). 2-(2-aminoethyl)-1-methylpyrrolidine (AEMP), N,N -dicyclohexylcarbodiimide (DCC) and 3,4-dihydro-3hydroxy-4-oxo-1,2,3-benzotriazine (HOOBt) from Aladdin reagent Co., Ltd. (Shanghai, China), were used for the derivatization of GA3 . Acetonitrile and methanol, HPLC grade, were acquired from Fisher Scientific Co. (Fair Lawn, NJ, USA). All other reagents and chemicals, analytical grade, were bought from Xilong Chemical Co., Ltd. (Guangdong, China). Except otherwise stated, all CE solutions were dissolved in ultrapure water processed with Molecular Lab Water Purifier (Molecular Instrument Shanghai Co., Ltd., Shanghai, China) and stored in the refrigerator at 4 ◦ C before use. The working standard solutions and sample solutions were freshly prepared by precisely diluting stock solution with proper buffer. The required sample solutions and buffer solutions were filtered through 0.22 m membrane filters (Shanghai Xinya Purification Material Factory, Shanghai, China) prior to CE-ECL analysis. 2.2. Apparatus All separations and analyses were performed with a CE-ECL detection system (Model MPI-B, Xi’an Remex Analysis Instrument Co., Ltd., Xi’an, China), including a high-voltage power supplier for electrokinetic injection and separation, an electrochemical potentiostat, a multifunctional chemiluminescence detector and a data collection analyzer. A three-electrode system used in chemiluminescence detection system was composed of a 500 m diameter Pt disk as the working electrode, an Ag/AgCl electrode as the reference electrode and a Pt wire as the auxiliary electrode. The surface of the working electrode was polished in sequence with 1.0, 0.3 and 0.05 m alpha alumina powder (CH Instruments, Inc., Austin, TX, USA) and then was cleaned with ultrapure water in a KQ5200DB ultrasonic cleaner (Kun Shan Ultrasonic Instruments Co., Ltd., Jiangsu, China) before use. A photomultiplier tube (PMT) was used to collect the ECL signal. The PMT signal was amplified and recorded using the MPI-B software. Cleanert C18-SPE cartridges (200 mg/3 mL, Bonnal-Agela Technologies, Tianjin, China) and a rotary evaporator (Model RE-52A, Shanghai Yarong Biochemical Instrument Factory, Shanghai, China) were used for sample pretreatment. An uncoated fused silica capillary (50 m i.d., 375 m o.d.) of 42 cm length (Yongnian Reafine Chromatography Ltd., Hebei, China) was used as electrophoretic separation channel. The new capillary was filled with 0.1 M NaOH overnight before first use. The capillary was rinsed with 0.1 M NaOH, ultrapure water and the corresponding separation buffer for 10 min, respectively, to maintain an active and reproducible inner surface prior to a series of analyses. During the entire experiment, the distance between the working electrode and the outlet of the capillary was fixed at
2.4. Sample pretreatment The samples of stem, leaf and seed were collected from the soybean (Glycine max [L.] Merr.) cultivar in the laboratory of corresponding author. All samples were ground to a homogenous powder. Subsequently, 2.00 g of powdered plant material was accurately weighed and extracted with 30 mL of 80% methanol in the dark at 4 ◦ C for 12 h. The extract was filtered and the filtrate was transferred into a flask. The extraction was repeated again for 2 h to increase the extraction yield and the extracted methanol solutions were combined. The combined extracts were evaporated using a rotary evaporator at 45 ◦ C. However, the above crude extract contained many interfering components. Liquid–liquid extraction (LLE) combined with solid-phase extraction (SPE) was applied to clean up the sample and enrich GAs in this experiment. The pH value of residues was adjusted to 3 with acetic acid. The residues were further extracted with 30 mL of ethyl acetate. The extraction was repeated two additional times with 20 mL and 10 mL of ethyl acetate, respectively. The ethyl acetate extracts were combined and evaporated to nearly dryness using a rotary evaporator at 45 ◦ C. The dried extract was then dissolved in 10% methanol and centrifuged at 10,000 rpm for 5 min at 4 ◦ C. The supernatant was further purified using SPE with the Cleanert C18-SPE cartridge, in which 1 mL of methanol, 1 mL of water and 5 mL of 40% methanol were used as purificant, activator and eluent, respectively. The eluates were evaporated to dryness under the same conditions. The dried residues were dissolved in acetonitrile and mixed with 128 L of 10.0 × 10−2 M DCC, 165 L of 4.0 × 10−2 M HOOBt and 318 L of 3.6 × 10−2 M AEMP for AEMP labeling in the same manner. The derivatives were dissolved in methanol to 1 mL. Then 0.125 mL of derivatization solution was diluted with separation buffer to 10 mL and filtered through a 0.22 m membrane filter for CE-ECL detection. 3. Results and discussion 3.1. ECL behavior of Ru(bpy)3 2+ in the presence of GA3 -AEMP-derivative In the CE-ECL system, the ECL intensity depends upon the rate of the light-emitting chemical reaction which is in turn dependent
64
G. Zhu et al. / J. Chromatogr. B 941 (2013) 62–68
Fig. 1. Labeling of GA3 with AEMP using DCC and HOOBt as coupling agents.
on the potential applied to the electrode [26,33]. Fig. 2A shows cyclic voltammograms of 5 mM Ru(bpy)3 2+ in 50 mM phosphate buffer (pH 7.0) before and after adding GA3 , GA3 -AEMP-derivative and AEMP, respectively. The oxidation peak currents of GA3 , GA3 -AEMP-derivative and AEMP could be distinguished from the background current. The corresponding ECL curves were illustrated in Fig. 2B. Compared with the background curve (curve a in Fig. 2B), curve b in Fig. 2B revealed that there was a little increase in the ECL intensity in the presence of GA3 . As expected, there was an obvious increase of the ECL intensity when GA3 was tagged with AEMP (curve c in Fig. 2B). Nevertheless, the ECL intensity of GA3 -AEMP-derivative was still lower than that of AEMP (curve d in Fig. 2B). The observations indicated that GA3 -AEMP-derivative could react with Ru(bpy)3 2+ in the electrochemiluminescence process and enhanced the emitted light intensity. It was due to the AEMP derivative conserved the intensity of the original molecule. The results suggested that the proposed CE-ECL method could be used for the sensitive determination of GA3 . 3.2. Labeling of GA3 with AEMP Generally speaking, the ECL intensity produced by tertiary amines is the highest, followed by secondary and finally primary amines in the ECL system based on Ru(bpy)3 2+ [28,34]. As
a result, the detection sensitivity of compounds can be enhanced by changing their structures or derivatizing with different ECLactive species into tertiary amines. Given that GA3 contains a carboxylic acid group, the potential derivatization reagents containing both a primary amine group and a tertiary amine group will be an excellent choice. It is because many approaches can be applied to couple these two groups with high efficiency under mild condition. AEMP was found to be an ECL-active compound [35–37]. Therefore, AEMP was chosen as the derivatization reagent for labeling of GA3 in this study. In addition, two common systems, namely 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC)/N-hydroxysuccinimide (NHS) system and DCC/HOOBt system, can be used as coupling agents to label a carboxylic acid group with a primary amine group in organic solvents. Nevertheless, the EDC/NHS system will result in some ECL-active species, which could interfere with the ECL detection of GA3 [35]. Hence, the DCC/HOOBt system was adopted. Fig. 1 displays the molecular structure of AEMP and the process of synthesis. It is necessary to select a proper solvent for derivatization because GA3 -AEMP-derivative should be synthesized in a suitable organic solvent, in which reactants dissolve well and GA3 is labeled with AEMP with high efficiency. Given that dioxane is highly toxic, acetonitrile, ethyl acetate, chloroform and DMSO were used as candidates for derivatization solvents. The results indicated that the highest derivatization efficiency was achieved by using acetonitrile as the derivatization solvent. Consequently, acetonitrile was selected as the optimum derivatization solvent. The concentration of DCC, HOOBt and AEMP also greatly influences the coupling of GA3 to AEMP. Fig. 3 shows the effect of the concentration of DCC, HOOBt and AEMP on the ECL peak area of 2 × 10−5 M GA3-AEMP-derivtive. The maximum peak area was obtained at 6.4 mM for DCC, 3.3 mM for HOOBt, 5.8 mM for AEMP. Therefore, 6.4 mM DCC, 3.3 mM HOOBt and 5.8 mM AEMP were chosen as their optimum concentrations for AEMP labeling. 3.3. Selection of solvent of GA3 -AEMP-derivative
Fig. 2. Cyclic voltammograms (A) and their corresponding ECL curves (B). (a) 5 mM Ru(bpy)3 2+ + 50 mM phosphate buffer (pH 7.0); (b) 5 mM Ru(bpy)3 2+ + 50 mM phosphate buffer (pH 7.0) + 2 × 10−5 M GA3 ; (c) 5 mM Ru(bpy)3 2+ + 50 mM phosphate buffer (pH 7.0) + 2 × 10−5 M GA3 -AEMP-derivative; (d) 5 mM Ru(bpy)3 2+ + 50 mM phosphate buffer (pH 7.0) + 2 × 10−5 M AEMP; scan rate, 100 mV/s.
The sample should be dissolved well in an appropriate solvent. In previous works, acetonitrile was used as an effective additive in CE separation because it can provide good resolution and repeatability for determination of similar compounds [39–41]. Nonetheless, high concentrations of acetonitrile will lead to bubble formation and thus may decrease CE separation. In this work, acetonitrile and methanol were used as candidates for solvents of GA3 -AEMP-derivative. At last, methanol was found to give both higher ECL intensity and better ability to dissolve
G. Zhu et al. / J. Chromatogr. B 941 (2013) 62–68
65
Fig. 3. Effect of concentration of DCC (A), HOOBt (B) and AEMP (C) on the ECL peak area of 2 × 10−5 M GA3 -AEMP-derivative. Detection conditions: in ECL cell, 50 mM phosphate buffer (pH 7.5) containing 5 mM Ru(bpy)3 2+ ; detection potential, 1.20 V; separation buffer, 10 mM phosphate buffer (pH 5.7); separation voltage, 15 kV; injection voltage, 10 kV; injection time, 10 s; sample solvent, acetonitrile.
GA3 -AEMP-derivative. Moreover, better peak shape was obtained by using methanol as sample solvent. So methanol was used as the solvent of GA3 -AEMP-derivative. Furthermore, even though the methanol solution of GA3 -AEMP-derivative was stored in the dark for 30 days in the refrigerator at 4 ◦ C, the CE-ECL analysis was nearly identical to that of freshly prepared solution. The results showed that the GA3 derivative was stable.
varied from 6.0 to 7.0 and decreased at higher pH values. Moreover, high pH value resulted in enhancement of background noise and thus led to less sensitive and stable analysis. The results were similar to previous works [35,37,38,41]. Finally, 50 mM phosphate buffer at pH 7.0 was used as the detection buffer in the experiment.
3.4. Optimization of CE-ECL detection conditions
3.5.1. Effect of pH and concentration of separation buffer The composition of separation buffer impacts on the ECL intensity and the separation efficiency by influencing the zeta potential
3.4.1. Effect of detection potential The electrochemiluminescence reaction depends on the potential applied to the electrode, so the relationship between ECL intensity and detection potential was investigated in the potential range of 1.00–1.30 V. It was found that the ECL intensity first increased and then decreased with the increase of detection potential. The maximum ECL intensity was obtained at a potential of 1.20 V. Therefore, 1.20 V was chosen as the detection potential in the following experiments. 3.4.2. Effect of concentration and pH of detection buffer The concentration of detection buffer is crucial for the ECL detection. Over the detection buffer concentration range of 20–60 mM, the maximum ECL intensity was achieved at 50 mM. So the phosphate buffer concentration was set at 50 mM. The buffer pH value in ECL detection cell affects the sensitivity and stability of determination of analytes because the reaction process relates to the deprotonation. The effect of pH value on ECL intensity was studied in the range of 6.0–9.0 by fixing the concentration of the detection buffer at 50 mM. As displayed in Fig. 4, the ECL intensity increased when the pH value of detection buffer
3.5. Optimization of separation conditions
Fig. 4. Effect of detection buffer pH on ECL intensity of 2 × 10−5 M GA3 -AEMPderivative. Detection conditions: sample solvent, methanol; other conditions were the same as those in Fig. 3.
66
G. Zhu et al. / J. Chromatogr. B 941 (2013) 62–68
Fig. 5. Effect of separation buffer pH on ECL intensity of 2 × 10−5 M GA3 -AEMPderivative. Detection conditions: in ECL cell, 50 mM phosphate buffer (pH 7.0) containing 5 mM Ru(bpy)3 2+ ; sample solvent, methanol; other conditions were the same as those in Fig. 3.
(), the electroosmotic flow (EOF) and the charge of analytes. 10 mM phosphate buffer was tested over the pH range of 4.0–8.0 (Fig. 5). The ECL intensity of GA3 -AEMP-derivative increased rapidly in the pH range of 4.0–5.7, but the intensity decreased when the pH value was higher than 5.7. At the same time, the migration time was shortened because of the increasing EOF. Therefore, the phosphate buffer at pH 5.7 was chosen for further experiments. The concentration of separation buffer is also one of the most crucial factors in the CE-ECL detection. The ECL intensity first ascended and then descended with the increase of buffer concentration in the rage of 5–20 mM. The maximum ECL intensity was observed when the concentration of phosphate buffer was 10 mM. Thereby, 10 mM phosphate buffer at pH 5.7 was used as separation buffer for the subsequent study. 3.5.2. Effect of separation voltage It is essential to optimize the separation voltage because of its influence on the migration time of analytes and the number of theoretical plates (N) by driving samples through the capillary. The ECL intensity kept increasing when the separation voltage was augmented from 8 to 18 kV in this investigation. The ECL response obviously increased when the separation voltage was changed from 8 to 15 kV, and the increasing trend became slower above 15 kV. In the meantime, the migration time decreased and the level of baseline noise increased with increasing separation voltage due to the increase of Joule heating in the capillary. Furthermore, the fast flow rate might reduce the concentration of Ru(bpy)3 3+ at the working electrode surface, which in turn will cause lower efficiency of light emission. Taking into consideration ECL intensity, migration time, baseline noise and Joule heating, a voltage of 15 kV was selected for separation. 3.5.3. Effect of injection voltage and injection time The injection parameters are major elements affecting the ECL intensity and N, too. N is calculated by the equation: N = 2 5.54(tm /W1/2 ) , where tm and W1/2 are the migration time and the width at half–maximum peak height, respectively. In general, a higher injection voltage and a longer injection time can cause stronger ECL signal but lower separation efficiency [32,33,42]. Their effects on the ECL intensity and N were studied by altering injection voltage when injection time was invariable (Fig. 6A) and changing injection time while injection voltage was kept constant (Fig. 6B). As predicted, both a higher injection voltage and a longer injection time gave rise to higher ECL intensity but lower N and broadening
Fig. 6. Effects of injection voltage (A) and injection time (B) on ECL intensity (a) and N (b) of 2 × 10−5 M GA3 -AEMP-derivative. Detection conditions: in ECL cell, 50 mM phosphate buffer (pH 7.0) containing 5 mM Ru(bpy)3 2+ ; sample solvent, methanol; other conditions were the same as those in Fig. 3.
of ECL peak as a result of the dispersion of analytes. Moreover, the ECL intensity slowly increased when injection voltage was higher than 10 kV and injection time exceeded 12 s. This was due to more analyses were involved in ECL reaction using a higher injection voltage or a longer injection time. Given that an electrokinetic injection duration of 10 s at 10 kV could offer better peak shape than that of duration of 12 s, 10 kV and 10 s were used as the optimal injection voltage and injection time, respectively. 3.6. Linearity, detection limit and reproducibility Under the optimized conditions, the linear range, detection limit and reproducibility were studied. The calibration curve was linear between 2.0 × 10−7 and 1.28 × 10−4 M for GA3 . The calibration equation and correlation coefficient were: y = 702.07(10−7 M) + 1905.7 and r = 0.9997 in terms of peak area response as a function of GA3 concentration. The limit of detection for GA3 was 8 × 10−8 M (S/N = 3), which is higher than those obtained using biosensor [18], HPLC–MS [20] and CE-LIF [25], but lower than that obtained using CE-MS [24]. Furthermore, the linear range of this method is wider than that of the above methods. The relative standard deviations of migration time, peak intensity and peak area for nine continuous injections of 2.0 × 10−5 M GA3 -AEMP-derivative were 1.0%, 2.1% and 4.2%, respectively. 3.7. Interference study Many molecules in plant tissues may interfere with the detection of GAs. Six amino acids, tyrosine (Tyr), proline (Pro),
G. Zhu et al. / J. Chromatogr. B 941 (2013) 62–68
67
Table 1 Analytical results and recoveries of GAs in samples (n = 3). Sample
GAs content (103 g/kg)
Added (103 g/kg)
Found (103 g/kg)
Recovery (%)
RSD (%)
Stem Leaf Seed
2.2637 0.2983 0.3370
3.8000 3.8000 3.8000
6.0375 3.7035 4.0502
99.3 89.6 97.7
0.7 3.5 0.4
methionine (Met), cysteine (Cys), glycine (Gly) and histidine (His) were selected to assess the potential interference for GAs in the proposed method. These amino acids contain a carboxylic group, which enables them to be coupled to AEMP. 4 mg of individual Tyr, Pro, Met, Cys, Gly and His were added to the standard solution to study their effects on the determination of GAs. Nevertheless, no interference was observed because they are hardly dissolved in acetonitrile and high soluble in aqueous solution. Even if some species are extracted from plant tissues, they are hardly labeled with AEMP and detected by CE-ECL. Therefore, the tertiary amine labeling contributed to the high selectivity of this method for GAs. At the same time, GA4 and GA7 were added to the standard solution to study the selectivity of this method for GA3 . However, interference was observed with 10 M of individual GA4 and GA7 on the determination of GA3 and the peaks of GA3 , GA4 and GA7 after derivatization could not be successfully separated due to highly similar structures. It may because they could not be separated from each other under the proposed separation conditions. Therefore, the proposed method could be applied to the analysis of total GAs. Nevertheless, the researches of Ge [24] and Chen [25] exhibited the potential for the separation and simultaneous determination of multiple GAs using CE-ECL method by optimizing separation conditions. 3.8. Application
of the standard derivative of GA3 with those of extracts and spiked extracts as described in Fig. 7. The contents of GAs in the stem, leaf and seed of soybean (Glycine max [L.] Merr.) were 0.181, 0.024, 0.027 mg/g on an average, respectively. Recovery experiments were carried out by adding GA3 standard to the known real extracts of samples, followed by being labeled with AEMP and analyzed by the established method (Table 1). The recovery of GAs was in the range of 89.6%–99.3%. 4. Conclusions This is the first report to describe the determination of GAs by CE-ECL. All relevant operational parameters have been optimized. Under the optimum conditions, the developed CE-ECL method exhibits excellent performance with wide linear range, fast separation, high sensitivity and low detection limit. The proposed method was successfully applied to the detection of GAs in real samples. It is a valuable addition to existing analytical tools for GAs analysis. Acknowledgments This work was supported by the National Natural Science Foundation of China (Nos. 31060193 and 31200269) and the Natural Science Foundation of Zhejiang Province (No. LQ21C02004). References
The proposed CE-ECL method was applied to the analysis of total content of GAs in the extracts of stem, leaf and seed from soybean (Glycine max [L.] Merr.) under the optimized conditions. In the present work, the extracts of GAs in samples were determined after being coupled to AEMP. Identification of the GAs derivative of the extracts was confirmed by comparing the electropherogram
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]
−5
Fig. 7. CE-ECL electropherograms of 2 × 10 M GA3 -AEMP-derivative spiked with 2 × 10−5 M AEMP (A), soybean leaf (B) and soybean leaf spiked with 3.8 × 103 g/kg GA3 (C). Detection conditions: in ECL cell, 50 mM phosphate buffer (pH 7.0) containing 5 mM Ru(bpy)3 2+ ; detection potential, 1.20 V; separation buffer, 10 mM phosphate buffer (pH 5.7); separation voltage, 15 kV; injection voltage, 10 kV; injection time, 10 s; sample solvent, methanol.
[25] [26] [27] [28] [29] [30] [31] [32]
A. Piotrowska, A. Bajguz, Phytochemistry 72 (2011) 2097. X. Shan, J. Yan, D. Xie, Curr. Opin. Plant Biol. 15 (2012) 84. T. Reynolds, P. Thompson, Physiol. Plant. 28 (2006) 516. J. Zeevaart, D.A. Gage, M. Talon, Proc. Natl. Acad. Sci. USA 90 (1993) 7401. A. Maggio, G. Barbieri, G. Raimondi, S. Pascale, J. Plant Growth Regul. 29 (2009) 63. H. Magome, S. Yamaguchi, A. Hanada, Y. Kamiya, K. Oda, Plant J. 37 (2004) 720. R.G. Sharp, M.A. Else, W.J. Davies, R.W. Cameron, Sci. Hortic. 124 (2010) 231. T.D. Thomas, Euphytica 164 (2008) 317. J.C. Serrani, O. Ruiz-Rivero, M. Fos, J.L. García-Martínez, Plant J. 56 (2008) 922. A.K. Subbaraj, K.A. Funnell, D.J. Woolley, J. Plant Growth Regul. 29 (2010) 487. M. Iqbal, M. Ashraf, Environ. Exp. Bot. 86 (2013) 76. L. Sun, Y. Zhao, X. Nie, Y. Ling, X. Chu, D. Shang, Y. Dong, Chem. Res. Chin. Univ. 28 (2012) 797. W. Xie, C. Han, Z. Zheng, X. Chen, Y. Qian, H. Ding, L. Shi, C. Lv, Food Chem. 127 (2011) 890. C. Rodrigues, L.P.d.S. Vandenberghe, J. de Oliveira, C.R. Soccol, Crit. Rev. Biotechnol. 32 (2012) 263. G.K. Wang, M. Zhang, J.F. Gong, Q.F. Guo, Y.N. Feng, W. Wang, Plant Cell Rep. 31 (2012) 2215. L. Kinsman, N. Pinfield, A. Stobart, Planta 127 (1975) 149. Z. Jiang, X. Liu, A. Liang, Anal. Chim. Acta 346 (1997) 249. J. Li, S. Li, X. Wei, H. Tao, H. Pan, Anal. Chem. 84 (2012) 9951. V. Gupta, M. Kumar, H. Brahmbhatt, C.R. Reddy, A. Seth, B. Jha, Plant Physiol. Biochem. 49 (2011) 1259. Z. Zhang, W. Tan, Y. Hu, G. Li, S. Zan, Analyst 137 (2012) 968. J. Zhang, P. Du, Se Pu 29 (2011) 1133. L. Wang, Q. Wu, C. Duan, D. Wu, Y. Guan, Se Pu 29 (2011) 923. P.S. Blake, J.M. Taylor, W.E. Finch-Savage, Plant Growth Regul. 37 (2002) 119. L. Ge, C.Y. Peh, J.W. Yong, S.N. Tan, L. Hua, E.S. Ong, J. Chromatogr. A 1159 (2007) 242. H. Chen, X.F. Guo, H.S. Zhang, H. Wang, J. Chromatogr. B 879 (2011) 1802. Y. Gao, Y. Xu, B. Han, J. Li, Q. Xiang, Talanta 80 (2009) 448. X.B. Yin, E. Wang, Anal. Chim. Acta 533 (2005) 113. M. Su, W. Wei, S. Liu, Anal. Chim. Acta 704 (2011) 16. J. Yin, Y. Xu, J. Li, E. Wang, Talanta 75 (2008) 38. B. Han, Y. Du, E. Wang, Microchem. J. 89 (2008) 137. B. Yuan, C. Zheng, H. Teng, T. You, J. Chromatogr. A 1217 (2010) 171. Y. Wang, L. Xiao, M. Cheng, J. Chromatogr. A 1218 (2011) 9115.
68 [33] [34] [35] [36] [37]
G. Zhu et al. / J. Chromatogr. B 941 (2013) 62–68 B. Deng, L. Ye, H. Yin, Y. Liu, S. Hu, B. Li, J. Chromatogr. B 879 (2011) 927. B. Deng, Y. Liu, H. Yin, X. Ning, H. Lu, L. Ye, Q. Xu, Talanta 91 (2012) 128. X.B. Yin, J.M. Guo, W. Wei, J. Chromatogr. A 1217 (2010) 1399. X.B. Yin, Y. Du, X. Yang, E. Wang, J. Chromatogr. A 1091 (2005) 158. H. Morita, M. Konishi, Anal. Chem. 74 (2002) 1584.
[38] [39] [40] [41] [42]
H. Morita, M. Konishi, Anal. Chem. 75 (2003) 940. S. Bean, G. Lookhart, J. Bietz, J. Agric. Food Chem. 48 (2000) 318. Y. Walbroehl, J.W. Jorgenson, Anal. Chem. 58 (1986) 479. C. Yu, H. Du, T. You, Talanta 83 (2011) 1376. B. Deng, Y. Kang, X. Li, Q. Xu, J. Chromatogr. B 857 (2007) 136.
