- Email: [email protected]
Stable isotope labeling assisted liquid chromatography–electrospray tandem mass spectrometry for quantitative analysis of endogenous gibberellins
Talanta 144 (2015) 341–348
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
Stable isotope labeling assisted liquid chromatography–electrospray tandem mass spectrometry for quantitative analysis of endogenous gibberellins Yan-Hong Hao, Zheng Zhang, Lu Wang, Chao Liu, Ai-Wen Lei, Bi-Feng Yuan, Yu-Qi Feng n Key Laboratory of Analytical Chemistry for Biology and Medicine (Ministry of Education), Department of Chemistry, Wuhan University, Wuhan 430072, PR China
art ic l e i nf o
a b s t r a c t
Article history: Received 4 April 2015 Received in revised form 17 June 2015 Accepted 20 June 2015 Available online 24 June 2015
In the current study, we developed a stable isotope labeling strategy for the absolute quantification of gibberellins (GAs) by high performance liquid chromatography–electrospray tandem mass spectrometry (HPLC–ESI-MS/MS). N,N-dimethyl ethylenediamine (DMED) and its deuterated counterpart d4-DMED were used to derivatize GAs extracted from plant tissue samples and GA standards respectively. The both derivatives of GAs were mixed and then subjected to HPLC–ESI-MS/MS analysis. The absolute quantification of GAs in plant tissues could be achieved by calculating the peak area ratios of DMED labeled GAs/d4-DMED labeled GAs. In the proposed strategy, the derivatization reaction of the labeling reagents with GAs could be completed rapidly (within 5 min) with high efficiency ( 499%) under mild conditions. The resulting derivatives could produce specific fragments in collision induced dissociation (CID), leading to high selectivity in multiple-reaction monitoring (MRM) mode, thus enhanced the reliability of the LC– MS/MS method. Furthermore, the limits of quantitation (LOQs) of GAs were considerably decreased (2– 32 folds) due to incorporating easily ionized moieties into GAs, and the quantification of GAs in plant tissue could be achieved without isotopically labeled GA standards. Good linearity was obtained with correlation coefficients R2 values of 40.99. The limits of detection (LODs) and quantitation (LOQs) ranged from 0.02 to 0.74 pg and 0.07 to 2.45 pg, respectively. Eleven GAs could be successfully determined in spiked sample with 72–128% recoveries and the relative standard deviations (RSDs) were between 1.0% and 13.9%. Finally, the developed method was successfully applied for the detection of GAs in 50 mg (fresh weight) Oryza sativa leaves. & 2015 Elsevier B.V. All rights reserved.
Keywords: Gibberellins HPLC–ESI-MS/MS Stable isotope labeling
1. Introduction Gibberellins (GAs) are a class of acid phytohormones that regulate many aspects of plant growth and development, mainly including stem elongation, germination, flowering and fruit development [1–3]. Elucidation of GAs functions and the molecular mechanism of how GAs control the developmental processes of plants are important for agriculture production. Since the physiological actions of GAs partly depend on its endogenous concentrations, it is important to build a method for accurate quantification of GAs in plant samples. Liquid chromatography–electrospray ionization-tandem mass spectrometry (LC–ESI-MS/MS) has been used for GAs analysis due to its high selectivity and sensitivity [4–10]. However, the ionization efficiency of GAs in negative ion mode is poor, leading to the n
Corresponding author. Fax: þ 86 27 68755595. E-mail address: [email protected] (Y.-Q. Feng).
http://dx.doi.org/10.1016/j.talanta.2015.06.056 0039-9140/& 2015 Elsevier B.V. All rights reserved.
low sensitivity of GAs by LC–ESI-MS/MS. Most of the endogenous GAs in plant tissue exists in an extremely low level. In this case, it is difficult to get a full profile of the endogenous GAs by LC–ESIMS/MS in negative ion mode. Furthermore, the complicated matrix of plant extracts leads to severe ionization suppression of GAs, and the ionization suppression varies widely depending on the coelution. For this reason, isotopically labeled internal standards are essential to correct matrix effects during ionization. However, isotopically labeled internal standards of GAs are often expensive or difficult to obtain. Besides, the commercial available isotopically labeled internal standards of GAs only contain two deuterium atoms in structure, thus the cross interference derived from isotope forms of the native GAs is hard to avoid. Stable isotope labeling has been proved to be an effective strategy for relative quantification of various kinds of analytes including plant hormones in samples with complex matrix [11– 16]. Instead of synthesizing the stable isotope labeled internal standards of the analytes of interest, an alternative strategy is to introduce stable isotopes into analytes in comparative samples by
342
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
Fig. 1. Chemical structures of 11 GAs.
chemical derivatization. The two different labeled samples are mixed together and then subjected to LC–MS analysis. As the isotope labeled derivatives are expected to co-elute during the LC separation procedure, the matrix effects and run-to-run ionization differences are identical for them. Therefore, the peak intensity ratio of the different isotope labeled analyte pair provides relative quantification of analyte in two comparative samples. When one of the samples is a standard solution with a known concentration, absolute quantification can be achieved without the use of individual isotopically labeled internal standards. For detection of analytes of low concentrations, an ideal labeling reagent should contain ionizable functional groups in structure. Thus sensitivity can be improved by incorporating easily ionized moieties into analytes. Up to now, stable isotope labeling method has not been found in application of absolute quantitative analysis of plant hormones including GAs. In this study, N,N-dimethyl ethylenediamine (DMED) and its isotope-labeled counterpart d4-DMED were developed as a pair of isotope mass probes for absolute quantitative analysis of 11 GAs (Fig. 1) in complex plant samples by LC–MS/MS. The derivatization reaction could be completed rapidly with high efficiency under mild conditions, and the resulting GAs derivatives were found to produce specific fragments which impart high selectivity for MRM analysis. The developed stable isotope labeling assisted LC–MS/MS method exhibited high sensitivity and accuracy in the quantification of GAs without using individual isotopically labeled GAs. Finally, the developed method was successfully applied for determination of GAs in plant tissue samples, demonstrating its excellent application potential in plant hormone research.
2. Materials and methods 2.1. Chemicals and reagents GAS standards: GA8, GA29, GA3, GA1, GA6, GA5, GA34, GA51, GA7, GA4 and GA9 were purchased from Olchemim Ltd. (Olomouc, Czech Republic). HPLC-grade acetonitrile (ACN) and methanol (MeOH) were purchased from Merck (Darmstadt, Germany). MilliQ water (Millipore, Bradford, USA) was used in all experiments. Formic acid (FA) and triethylamine (TEA), ethyl ether, 1,2-diaminoethane, ammonium formate, dichloromethane (DCM), methanol (MeOH) and acetic acid were bought from Sinopharm Chemical Reagent (Shanghai, China). Benzylorthochloroformiate, 10% palladium on activated charcoal, NaBH3CN and 2-chloro-1-methylpyridinium iodide (CMPI) were purchased from Aladdin Reagent Co. (Shanghai, China). N,N-dimethyl ethylenediamine (DMED, 98%) and deuterium formaldehyde (d4-HCHO, 20% in water) were purchased from Sigma (St. Louis, MO, USA). C18 SPE cartridges (1 mL, 50 mg) were obtained from Weltech Co. (Wuhan, China). The stock solutions of CMPI (100 mmol/mL), DMED (100 mmol/mL), d4-DMED (100 mmol/mL) and TEA (100 mmol/mL) were prepared by dissolving appropriate amount of CMPI, DMED and TEA in ACN, respectively. For GA8, GA29, GA3, GA1, GA6, GA5, GA34, GA51, GA7, GA4 and GA9, the stock solutions were prepared at the concentration of 10 mg/mL in ACN. All stock solutions were stored at 18 °C. The stock solutions were diluted with ACN to working solutions before analysis. 2.2. Synthesis of d4-DMED The pathway for synthesis of d4-DMED is shown in Fig. 2. 2.2.1. Synthesis of (2-amino-ethyl)-carbamic acid benzyl ester To a solution of 1,2-diaminoethane (30.0 g, 0.50 mmol) in DCM (500 mL) benzylorthochloroformiate (8.53 g, 50 mmol) in DCM
Fig. 2. Synthesis procedure of the d4-DMED reagents.
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
(250 mL) was added drop wise at 0 °C over 8 h. After stirring at ambient temperature over night, the mixture was washed five times with brine (100 mL) and then with water (100 mL). The combined organic phases were dried over MgSO4 and concentrated in vacuo to afford colorless oil, which solidifies in the refrigerator as a white solid (7.58 g, 39 mmol, 78%). 2.2.2. Synthesis of d4-(2-dimethylamino-ethyl)-carbamic acid benzyl ester To a vigorously stirred suspension of (2-amino-ethyl)-carbamic acid benzyl ester (2.22 g, 11.4 mmol) in methanol (30 mL), deuterium formaldehyde (20% in water, 6.3 mL) and NaBH3CN (1.03 g, 16 mmol) were added. After 15 min, acetic acid was added until the pH was neutral and stirring was continued for 24 h. After evaporation of solvents, the mixture was dissolved in DCM (30 mL) and washed by three times with brine (10 mL). After evaporation of the solvent under reduced pressure at 20 °C, the mixture was then purified by silica gel column chromatography (eluent: DCM/ MeOH/TEA, 94/1/5) to afford the expected compound (2.11 g, 9.5 mmol, 83% yield) as an orange oil. 2.2.3. Synthesis of d4-N1,N1-dimethylethane-1, 2-diamine (d4-DMED) To a solution of (2-dimethylamino-ethyl)-carbamic acid benzyl ester (449.1 mg, 2.0 mmol) in methanol (15 mL), 10% palladium on activated charcoal (30 mg) was added. The reaction was performed at 50 °C over 6 h in the present of hydrogen gas (20 atm). The suspension was then filtered to remove Pd–C. 1H-NMR purity: 498% (Fig. S1). In addition, DMED and d4-DMED were mixed equally and subjected to LC–MS analysis. In addition, it could be seen that the peak area ratio of the two compounds was approximate to 1, indicating the high purity of the synthesized reagent (Fig. S2). 2.3. Plant materials Oryza sativa was grown in an artificial environmental chamber greenhouse at 30 °C under 16 h light/8 h dark photoperiods. 10 days old O. sativa leaves were harvested, weighted, immediately
343
frozen in liquid nitrogen, and stored at 80 °C until needed. 2.4. Plant sample pretreatment and stable isotope labeled derivatization The overall procedure for sample pretreatment and stable isotope labeled derivatization is summarized in Fig. 3. The pretreatment of plant samples was performed according to our previously published work [17]. Briefly, Plant materials (50 mg fresh weight) were frozen in liquid nitrogen, ground into powder, and extracted with 1 mL methanol containing 20% water at 4 °C for 12 h. The extract was centrifuged at 12,000 g under 4 °C for 20 min. The supernatant was collected and then passed through a C18 SPE-cartridge (1 mL, 50 mg) which was pre-conditioned with 2 mL methanol and 2 mL methanol containing 20% water. The elution was pooled and dried under nitrogen gas stream, reconstituted in 100 mL 1% FA H2O (v/v), and then extracted with ethyl ether (4 1 mL). The ether phases were combined and then evaporated to dryness under nitrogen gas stream. The standard GAs and the pretreated plant samples were separately dissolved in 100 mL ACN, and the stable isotope labeled derivatization was started by adding 10 mL TEA (10 mmol/mL), 10 mL CMPI (20 mmol/mL). After mixing, 20 mL DMED (20 mmol/mL) was added to the plant extracts and 20 mL d4-DMED (20 mmol/mL) was added to the standards solution for derivatization. The reaction solutions were incubated at 40 °C for 5 min with shaking at 1500 rpm and then evaporated under a stream of nitrogen gas until dryness. At last, the “light” labeled plant samples and “heavy” labeled standard GAs were respectively dissolved in 50 mL ACN/H2O 5/95 (v/v) and mixed together. Then 30 mL of the combined solutions was injected for analysis. 2.5. HPLC–ESI-MS/MS analysis Analysis of GAs and its derivatives were performed on a HPLC– ESI-MS/MS system consisting of a Shimadzu MS-8050 mass spectrometer (Tokyo, Japan) with an electrospray ionization source (Turbo Ionspray), and a Shimadzu LC-20AD HPLC system (Tokyo, Japan), which was equipped with two LC-20AD pumps, a
Fig. 3. Procedure for sample pretreatment and stable isotope labeled derivatization.
344
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
SIL-20A autosampler, a CTO-20AC column thermostat and a DGU20A3 degasser. Data acquisition and processing were performed using LabSolutions software (version 5.53 sp2, Shimadzu, Tokyo, Japan). To illustrate the sensitivity enhancement by chemical labeling, both GAs and its derivatives were analyzed in multiple reaction monitoring (MRM) mode. Optimized parameters for quantification of DMED labeled GAs and the native GA molecules by MRM were listed in Table S1. The HPLC separation was performed on a Shimpack VP-ODS (150 mm 2.0 mm i.d., 5 mm, Shimadzu, Tokyo, Japan) column at 40 °C with a flow rate of 0.2 mL/min. GAs were analyzed in negative ion mode, FA in H2O (0.01%, v/v, solvent A) and ACN (solvent B) were used as mobile phase. Gradient elution was achieved by the following program: 2 min, 10% B, 20 min, 10– 70% B, 1 min, 70–10% B, 10 min, 10% B. DMED/d4-DMED labeled GAs were analyzed in positive ion mode. 5 mmol/L ammonium formate aqueous solution (solvent A) and ACN (solvent B) were chosen for the separation procedure. The gradient elution program was set as follows: 2 min 5% B, 20 min 5–45% B, 20–21 min 45–5% B, 30 min 5% B. The optimal ESI source conditions for GAs and its derivatives were as follows: DL temperature, 250 °C; heat block temperature, 400 °C; nebulizing gas, 2 L/min; drying gas, 10 L/min and heating gas, 10 L/min.
3. Results and discussion 3.1. Chemical derivatization CMPI was found to activate carboxyl groups to form a reactive ester, and then amines attacked the reactive ester rapidly to form an amide bond under slightly alkaline conditions [18–20]. In a preliminary experiment, DMED was added to the mixtures of GA1, GA3, GA4, GA7 and GA9 for derivatization at a molar ratio of 10/1 with CMPI as an activator. As CMPI also reacts with amines, CMPI was added to the GAs containing solution to active the analytes first and then DMED was put in for derivatization. The molar ratio of the derivatization reagent to the activator was fixed at 2:1 to avoid complete DMED consumption by CMPI. The reaction was performed at 40 °C for 1 h, and then the resulting solution was subjected to MS analysis by direct infusion to examine the derivatization products. As expected, in positive ion mode, GA9-DMED (m/z 387.2), GA7-DMED (m/z 401.2), GA4-DMED (m/z 403.2), GA3-DMED (m/z 417.2) and GA1-DMED (m/z 419.2) were distinctly observed on the MS spectrum, while m/z 102.2 and m/z 180.1 represent the parent ions of TEA and CMPI–DMED, respectively. And no apparent peaks of impurities were observed on the MS spectrum. In negative ion mode, m/z 127.1, which resulted from I of CMPI, was the main peak on the MS spectrum. And no apparent peaks of unlabeled GA9 (m/z 315.2), GA7 (m/z 401.2), GA4 (m/z 329.2), GA3 (m/z 331.2) and GA1 (m/z 347.2) were observed (Fig. S3). These experimental results indicate high conversion rates of the target analytes (GAs) to the products (GAs-DMED). 3.2. Optimization of derivatization conditions In order to achieve high derivatization efficiency, derivatization conditions including TEA concentration, reaction temperature, DMED amount and reaction time were optimized. Four bioactive GAs including GA1, GA3, GA4, GA7 (1 ng for each GA) were used for the optimization procedures. The reaction was stopped by immediate freezing at 20 °C for 10 min and then evaporated to dryness under a flow of nitrogen. d4-DMED labeled GAs (1 ng for each GA) were spiked before LC–MS/MS analysis to calibrate matrix effects and signal fluctuations during ionization. TEA concentration was firstly investigated. GAs were dissolved
in 100 μL ACN, followed by addition of 10 μL TEA of different concentrations ranging from 1 to 20 μmol/mL. Then 10 μL CMPI (20 μmol/mL) was added. After mixing, 20 μL DMED (20 μmol/mL) was put in for derivatization. The reaction was incubated at 40 °C for 60 min with shaking at 1500 rpm. Our results showed that the signal of derivatives kept stable when the concentration of TEA increased from 1 mmol/mL to 20 mmol/mL (Fig. 4A). Reaction temperature was then examined. Derivatization was performed under different temperatures from 10 to 60 °C for 60 min with excess amount of DMED as described above. The results showed that temperature had little influence on labeling efficiency (Fig. 4B). To optimize the amount of DMED, the molar ratios of DMED to the total amount of GAs were changed from 20 to 5000. The reaction was kept at 40 °C for 60 min. The results show that 200-fold molar excess of DMED was enough for the efficient labeling of GAs (Fig. 4C). At last, derivatization time was optimized ranging from 5 to 60 min with the TEA concentration, reaction temperature and DMED amount optimized in the previous procedures. The result showed that derivatization of GAs with DMED was very fast and 5 min was enough for efficient reaction (Fig. 4D). Taken together, the optimized derivatization conditions for GAs with DMED were 40 °C for 5 min with a 200/1 molar ratio of DMED/GAs. 10 mL TEA (10 mmol/mL) was added to create slightly alkaline conditions. Under the optimized derivatization conditions, derivatization efficiency was determined by dividing the peak areas of underivatized GAs after derivatization with equal amount of GA standards without derivatization, the ratio was then subtracted from one. The results showed that more than 99% of GAs were consumed (Table 1), suggesting that high derivatization efficiencies were achieved. As the plant samples are much more complex than standard solutions, the amounts of DMED should be reexamined for the derivatization of GAs in O. sativa samples. For this purpose, plant samples were spiked with 4 GAs (1 ng/g for each analytes) and pretreated according to procedures described in Section 2.4. Our results showed that the peak areas of GAs derivatives reached a plateau when the concentrations of DMED standard solutions were increased to 20 mmol/mL (Fig. S4). The stability of the derivatives was evaluated by monitoring the change at different time intervals. The results showed that the derivatives were stable for at least 24 h in ACN/H2O (5/95, v/v) at 25 °C (Fig. S5). 3.3. Fragmentation of the DMED labeled GAs The fragmentation behavior of DMED labeled GA3 was first investigated. Chemical structure and possible fragmentation pattern of DMED labeled GA3 (CE, 40 eV) were shown in Fig. 5. It shows that five main fragments were produced, and the signal of product ions with m/z 310.2 and m/z 372.2 were obvious higher than that of the other product ions. Fragments at m/z 372.2 was produced from the neutral loss (NL) of NH(CH3)2, while m/z 310.2 was resulted from the dissociation of GA3 skeleton. Product ions relative with dissociation of skeleton of analytes other than labeling reagent may help to improve the detection selectivity by tandem MS analysis. This assumption was proved by monitoring the MRM transitions of both 417.24 372.2 and 417.24310.2 during the analysis of plant samples. As shown in Fig. 6, the LC–MS/ MS chromatogram corresponding to MRM transition of 417.2 4310.2 showed much less interference and background noise compared with that of 417.24 372.2, which improved the signal to noise (S/N) ratio and decreased the limit of detection (LOD) values significantly. Weinstock et al. pointed out that it is important to consider the background noise for samples with
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
345
Fig. 4. Optimization of derivatization conditions for GAs standards: (A) optimization of TEA concentration; (B) optimization of reaction temperature; (C) optimization of DMED molar ratios over a total amount of GAs; (D) optimization of reaction time.
Table 1 Derivatization efficiencies under the optimum reaction condition. Analytes Derivatization efficiency (%)
GA8 99.1
GA29 99.3
GA3 99.0
GA1 99.6
GA6 99.7
GA5 99.7
GA34 99.4
GA51 99.7
GA7 99.2
GA4 99.9
GA9 99.8
Fig. 5. Chemical structure and possible fragmentation pattern of DMED-GA3 at CE ¼ 40 eV.
complicated matrix when comparing overall sensitivity improvements, as improvements in signal sensitivity can be completely neutralized by a detrimental rise in chemical noise [21]. As for other GAs, HPLC–MS/MS chromatogram derived from MRM transition of the structure specific fragments also owned more selectivity than that of the fragments resulting from the neutral loss of NH(CH3)2 (Fig. S6). Therefore, the structure specific fragments were chosen for quantitative analysis of GAs to enhance reliability of the method, even when they were not the fragments of the
highest abundance. 3.4. Signal enhancement of the derivatives GAs are a class of acid phytohormones that contain carboxyl groups in structures. An important issue facing us in the determination of GAs by ESI-MS is the poor sensitivity resulting from the low ionization efficiency in negative ion mode. Herein, we compared the limits of quantitation (LOQs) of GAs before and after
346
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
Fig. 6. The MRM chromatograms of DMED-GA3 in Oryza sativa leaves under different transitions. (A) 310.2, a structure specific fragment resulting from the dissociation of GA3 skeleton was chosen as the product ion for MRM; (B) 372.2, fragment after the neutral loss of NH(CH3)2 was chosen as the product ion for MRM. Table 2 LODs, LOQs of the 11 GAs with and without derivatization. Analytes LOQ/pg (S/N ¼ 10)
GA8 GA29 GA3 GA1 GA6 GA5 GA34 GA51 GA7 GA4 GA9
LOD/pg (S/N ¼ 3)
Nonlabeled
Labeled Nonlabeled
Labeled
18.18 13.79 0.77 3.64 0.75 3.64 2.03 6.80 0.64 2.94 4.29
2.31 2.45 0.08 0.19 0.47 0.25 0.89 0.21 0.07 0.17 0.19
0.69 0.74 0.03 0.06 0.14 0.08 0.27 0.06 0.02 0.05 0.06
5.45 4.14 0.23 1.09 0.23 1.09 0.61 2.04 0.19 0.88 1.29
Approximate sensitivity enhancement (fold)
8 6 9 19 2 15 2 32 9 17 23
derivatization under their optimal conditions. The signal enhancement factor was determined by comparing the LOQs of 11 GAs. As shown in Table 2, the LOQs of 11 GAs were decreased by 2– 32 folds upon derivatization.
3.5. Method validation The LC–MS/MS chromatogram of 11 GAs-DMED was shown in Fig. 7. The calibration curves were constructed by plotting the mean peak area ratios of DMED/d4-DMED derivatives against nominal concentration ratios based on data obtained from duplicate measurements. As shown in Table 3, good linearities were obtained with correlation coefficients (R2) of 0.9976–0.9999. The LODs and the LOQs were calculated as the concentration of the analytes at a S/N ratio of 3 and 10, respectively. The results showed that LODs and LOQs for the 11 derived GAs ranged from 0.02 to 0.74 pg and 0.07 to 2.45 pg, respectively (Table 2). In addition, the reliability and accuracy of the method were evaluated by spiking GAs standards into the rice matrices. The recoveries of 11 GAs in O. sativa leave samples at three concentration levels were found to be 72–128%. The relative standard deviations (RSDs) were 1.9–13.9% (Table 4). These results suggested that the accuracy and precision of the proposed method were satisfactory for the determination of GAs in plant samples, and the stable isotope labeling strategy worked well for absolute quantification of GAs.
Fig. 7. The MRM chromatograms of 11 DMED-derived GA (5 pg) standards analyzed by HPLC–ESI-MS/MS.
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
347
Table 3 Calibration curves of the 11 DMED-labeled GAs. Analytes Concentration range of GAs-DMED (ng/mL)
GA8 GA29 GA3 GA1 GA6 GA5 GA34 GA51 GA7 GA4
0.05–20 0.05–20 0.002–20 0.005–20 0.01–20 0.005–20 0.02–20 0.005–20 0.002–20 0.005–20
Fixed concentration of GAs-d4-DMED (ng/mL)
1 1 0.5 0.5 1 1 1 0.5 0.5 0.5
Nominal concentration ratios of DMED/d4-DMED labeled GAs
0.05–20 0.05–20 0.004–40 0.01–40 0.01–20 0.005–20 0.02–20 0.01–40 0.004–40 0.01–40
Table 4 Recoveries of 11 GAs in 50 mg Oryza sativa leaves (n¼ 3).
Calibration curve Slope
Intercept R2
1.0015 1.0524 1.0522 1.0646 0.9761 1.0423 1.0762 0.9777 0.9955 1.0104
0.0080 0.0976 0.0381 0.0824 0.0054 0.0179 0.0399 0.1090 0.0556 0.0014
0.9999 0.9976 0.9999 0.9998 0.9998 0.9999 0.9998 0.9994 0.9998 0.9996
Table 5 Determination of endogenous GAs in 50 mg (FW) Oryza sativa leaves (n¼3).
Analytes
Added (ng/g) low, medium, high
Recoveries (%)
RSD (%)
GA1 GA3 GA7 GA9 GA5 GA8 GA29 GA4 GA6 GA34 GA51
0.2, 1, 10 0.2, 1, 10 0.2, 1, 10 0.2, 1, 10 1, 5, 20 1, 5, 20 1, 5, 20 1, 5, 20 1, 5, 20 1, 5, 20 0.2, 1, 10
93, 104, 83 83, 87, 72 98, 99, 90 128, 106, 109 95, 89, 86 95, 77, 81 90, 88, 82 83, 87, 85 87, 80, 79 81, 83, 82 108, 102, 95
3.4, 5.1, 2.6 4.6, 3.7, 1.0 7.4, 7.5, 2.0 4.4, 1.4, 2.7 10, 1.5, 1.7 13.9, 6.4, 3.9 2.0, 8.0, 2.3 4.6, 7.3, 1.4 1.2, 8.0, 1.6 11.0, 4.9, 4.4 7.7, 8.3, 4.3
Analytes
Found (ng/g)
GA1 GA3 GA7 GA9 GA5 GA8 GA29 GA4 GA6 GA34 GA51
0.157 0.01 0.217 0.01 N.D. 0.58 7 0.04 N.D. N.D. N.D. 8.487 0.47 0.27 70.02 N.D. N.D.
N.D., not found.
suggesting that the high sensitive and reliable stable isotope labeling strategy can be successfully applied for identification and quantification of endogenous GAs in complex plant matrix
4. Conclusions
Fig. 8. The MRM chromatograms of detected endogenous GAs in 50 mg Oryza sativa leaves.
In this study, we developed an isotope mass probe labeling method for the sensitive and reliable analysis of GAs in plant samples by HPLC–ESI-MS/MS. The proposed method overcomes the limitations of using isotopically labeled internal standards in the traditional isotope dilution approach. Sensitivity was enhanced by incorporating ionized moieties into analytes to improve ionization efficiency. In addition, the derivatives could produce specific ions, which impart high selectivity for MRM analysis, thus enhanced the reliability of the method. The developed method was employed for identification and quantification of endogenous GAs in 50 mg O. sativa leaves. Five endogenous GAs including the bioactive GA1, GA3, GA4 were detected and quantified. In all, the high sensitive and reliable stable isotope labeling strategy has the potential to facilitate plant hormone research.
3.6. Determination of endogenous GAs in plant samples
Acknowledgments
GAs play important roles in plant growth and development. A satisfactory method should meet the need of detecting GAs of trace amount from complicated plant matrix. Fig. 8 showed the MRM chromatograms of endogenous GAs detected by the current method in 50 mg O. sativa leaves. It can be seen that five GAs including the bioactive GA1, GA3, GA4 could be successfully observed at the content ranging from 0.15 to 8.5 ng/g FW (Table 5),
The authors thank the National Natural Science Foundation of China for financial support (21475098, 91217309) and the Natural Science Foundation of Hubei Province, China (2014CFA002). Appendix A. Supplementary material Supplementary data associated with this article can be found in
348
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
the online version at http://dx.doi.org/10.1016/j.talanta.2015.06. 056.
References [1] M. Hamano, Y. Yamato, H. Yamazaki, H. Miura, J. Hortic. Sci. Biotech. 77 (2002) 220–225. [2] S. Yamaguchi, M. Ogawa, Y. Yamauchi, Y. Kamiya, Plant Cell Physiol. 43 (2002), S20–S20. [3] N. Arnaud, T. Girin, K. Sorefan, S. Fuentes, T.A. Wood, T. Lawrenson, R. Sablowski, L. Ostergaard, Genes Dev. 24 (2010) 2127–2132. [4] T. Urbanova, D. Tarkowska, O. Novak, P. Hedden, M. Strnad, Talanta 112 (2013) 85–94. [5] J.F. Liu, J. Ding, B.F. Yuan, Y.Q. Feng, Analyst 139 (2014) 5605–5613. [6] S.C. Liu, W.Q. Chen, L. Qu, Y. Gai, X.N. Jiang, Anal. Bioanal. Chem. 405 (2013) 1257–1266. [7] Z. Ma, L. Ge, A.S. Lee, J.W. Yong, S.N. Tan, E.S. Ong, Anal. Chim. Acta 610 (2008) 274–281. [8] M. Muller, S. Munne-Bosch, Plant Methods 7 (2011) 37. [9] X. Pan, X. Wang, J. Chromatogr. B 877 (2009) 2806–2813.
[10] X. Pan, R. Welti, X. Wang, Nat. Protoc. 5 (2010) 986–992. [11] Y.Q. Huang, J.Q. Liu, H.Y. Gong, J. Yang, Y.S. Li, Y.Q. Feng, Analyst 136 (2011) 1515–1522. [12] Y.Q. Huang, G.D. Ruan, J.Q. Liu, Q. Gao, Y.Q. Feng, Anal. Biochem. 416 (2011) 159–166. [13] S. Ogawa, S. Ooki, K. Shinoda, T. Higashi, Anal. Bioanal. Chem. 406 (2014) 6647–6654. [14] X. Sun, Y. Ouyang, J. Chu, J. Yan, Y. Yu, X. Li, J. Yang, C. Yan, J. Chromatogr. A 1338 (2014) 67–76. [15] T. Toyo’oka, J. Pharm. Biomed. Anal. 69 (2012) 174–184. [16] S. Zhang, J. You, S. Ning, C. Song, Y.R. Suo, J. Chromatogr. A 1280 (2013) 84–91. [17] M.L. Chen, Y.Q. Huang, J.Q. Liu, B.F. Yuan, Y.Q. Feng, J. Chromatogr. B 879 (2011) 938–944. [18] W.J. Leavens, S.J. Lane, R.M. Carr, A.M. Lockie, I. Waterhouse, Rapid Commun. Mass Spectrom. 16 (2002) 433–441. [19] Y.Q. Huang, Q.Y. Wang, J.Q. Liu, Y.H. Hao, B.F. Yuan, Y.Q. Feng, Analyst 139 (2014) 3446–3454. [20] Y. Tang, J. Xiong, H.P. Jiang, S.J. Zheng, Y.Q. Feng, B.F. Yuan, Anal. Chem. 86 (2014) 7764–7772. [21] M. Star-Weinstock, B.L. Williamson, S. Dey, S. Pillai, S. Purkayastha, Anal. Chem. 84 (2012) 9310–9317.
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
Stable isotope labeling assisted liquid chromatography–electrospray tandem mass spectrometry for quantitative analysis of endogenous gibberellins Yan-Hong Hao, Zheng Zhang, Lu Wang, Chao Liu, Ai-Wen Lei, Bi-Feng Yuan, Yu-Qi Feng n Key Laboratory of Analytical Chemistry for Biology and Medicine (Ministry of Education), Department of Chemistry, Wuhan University, Wuhan 430072, PR China
art ic l e i nf o
a b s t r a c t
Article history: Received 4 April 2015 Received in revised form 17 June 2015 Accepted 20 June 2015 Available online 24 June 2015
In the current study, we developed a stable isotope labeling strategy for the absolute quantification of gibberellins (GAs) by high performance liquid chromatography–electrospray tandem mass spectrometry (HPLC–ESI-MS/MS). N,N-dimethyl ethylenediamine (DMED) and its deuterated counterpart d4-DMED were used to derivatize GAs extracted from plant tissue samples and GA standards respectively. The both derivatives of GAs were mixed and then subjected to HPLC–ESI-MS/MS analysis. The absolute quantification of GAs in plant tissues could be achieved by calculating the peak area ratios of DMED labeled GAs/d4-DMED labeled GAs. In the proposed strategy, the derivatization reaction of the labeling reagents with GAs could be completed rapidly (within 5 min) with high efficiency ( 499%) under mild conditions. The resulting derivatives could produce specific fragments in collision induced dissociation (CID), leading to high selectivity in multiple-reaction monitoring (MRM) mode, thus enhanced the reliability of the LC– MS/MS method. Furthermore, the limits of quantitation (LOQs) of GAs were considerably decreased (2– 32 folds) due to incorporating easily ionized moieties into GAs, and the quantification of GAs in plant tissue could be achieved without isotopically labeled GA standards. Good linearity was obtained with correlation coefficients R2 values of 40.99. The limits of detection (LODs) and quantitation (LOQs) ranged from 0.02 to 0.74 pg and 0.07 to 2.45 pg, respectively. Eleven GAs could be successfully determined in spiked sample with 72–128% recoveries and the relative standard deviations (RSDs) were between 1.0% and 13.9%. Finally, the developed method was successfully applied for the detection of GAs in 50 mg (fresh weight) Oryza sativa leaves. & 2015 Elsevier B.V. All rights reserved.
Keywords: Gibberellins HPLC–ESI-MS/MS Stable isotope labeling
1. Introduction Gibberellins (GAs) are a class of acid phytohormones that regulate many aspects of plant growth and development, mainly including stem elongation, germination, flowering and fruit development [1–3]. Elucidation of GAs functions and the molecular mechanism of how GAs control the developmental processes of plants are important for agriculture production. Since the physiological actions of GAs partly depend on its endogenous concentrations, it is important to build a method for accurate quantification of GAs in plant samples. Liquid chromatography–electrospray ionization-tandem mass spectrometry (LC–ESI-MS/MS) has been used for GAs analysis due to its high selectivity and sensitivity [4–10]. However, the ionization efficiency of GAs in negative ion mode is poor, leading to the n
Corresponding author. Fax: þ 86 27 68755595. E-mail address: [email protected] (Y.-Q. Feng).
http://dx.doi.org/10.1016/j.talanta.2015.06.056 0039-9140/& 2015 Elsevier B.V. All rights reserved.
low sensitivity of GAs by LC–ESI-MS/MS. Most of the endogenous GAs in plant tissue exists in an extremely low level. In this case, it is difficult to get a full profile of the endogenous GAs by LC–ESIMS/MS in negative ion mode. Furthermore, the complicated matrix of plant extracts leads to severe ionization suppression of GAs, and the ionization suppression varies widely depending on the coelution. For this reason, isotopically labeled internal standards are essential to correct matrix effects during ionization. However, isotopically labeled internal standards of GAs are often expensive or difficult to obtain. Besides, the commercial available isotopically labeled internal standards of GAs only contain two deuterium atoms in structure, thus the cross interference derived from isotope forms of the native GAs is hard to avoid. Stable isotope labeling has been proved to be an effective strategy for relative quantification of various kinds of analytes including plant hormones in samples with complex matrix [11– 16]. Instead of synthesizing the stable isotope labeled internal standards of the analytes of interest, an alternative strategy is to introduce stable isotopes into analytes in comparative samples by
342
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
Fig. 1. Chemical structures of 11 GAs.
chemical derivatization. The two different labeled samples are mixed together and then subjected to LC–MS analysis. As the isotope labeled derivatives are expected to co-elute during the LC separation procedure, the matrix effects and run-to-run ionization differences are identical for them. Therefore, the peak intensity ratio of the different isotope labeled analyte pair provides relative quantification of analyte in two comparative samples. When one of the samples is a standard solution with a known concentration, absolute quantification can be achieved without the use of individual isotopically labeled internal standards. For detection of analytes of low concentrations, an ideal labeling reagent should contain ionizable functional groups in structure. Thus sensitivity can be improved by incorporating easily ionized moieties into analytes. Up to now, stable isotope labeling method has not been found in application of absolute quantitative analysis of plant hormones including GAs. In this study, N,N-dimethyl ethylenediamine (DMED) and its isotope-labeled counterpart d4-DMED were developed as a pair of isotope mass probes for absolute quantitative analysis of 11 GAs (Fig. 1) in complex plant samples by LC–MS/MS. The derivatization reaction could be completed rapidly with high efficiency under mild conditions, and the resulting GAs derivatives were found to produce specific fragments which impart high selectivity for MRM analysis. The developed stable isotope labeling assisted LC–MS/MS method exhibited high sensitivity and accuracy in the quantification of GAs without using individual isotopically labeled GAs. Finally, the developed method was successfully applied for determination of GAs in plant tissue samples, demonstrating its excellent application potential in plant hormone research.
2. Materials and methods 2.1. Chemicals and reagents GAS standards: GA8, GA29, GA3, GA1, GA6, GA5, GA34, GA51, GA7, GA4 and GA9 were purchased from Olchemim Ltd. (Olomouc, Czech Republic). HPLC-grade acetonitrile (ACN) and methanol (MeOH) were purchased from Merck (Darmstadt, Germany). MilliQ water (Millipore, Bradford, USA) was used in all experiments. Formic acid (FA) and triethylamine (TEA), ethyl ether, 1,2-diaminoethane, ammonium formate, dichloromethane (DCM), methanol (MeOH) and acetic acid were bought from Sinopharm Chemical Reagent (Shanghai, China). Benzylorthochloroformiate, 10% palladium on activated charcoal, NaBH3CN and 2-chloro-1-methylpyridinium iodide (CMPI) were purchased from Aladdin Reagent Co. (Shanghai, China). N,N-dimethyl ethylenediamine (DMED, 98%) and deuterium formaldehyde (d4-HCHO, 20% in water) were purchased from Sigma (St. Louis, MO, USA). C18 SPE cartridges (1 mL, 50 mg) were obtained from Weltech Co. (Wuhan, China). The stock solutions of CMPI (100 mmol/mL), DMED (100 mmol/mL), d4-DMED (100 mmol/mL) and TEA (100 mmol/mL) were prepared by dissolving appropriate amount of CMPI, DMED and TEA in ACN, respectively. For GA8, GA29, GA3, GA1, GA6, GA5, GA34, GA51, GA7, GA4 and GA9, the stock solutions were prepared at the concentration of 10 mg/mL in ACN. All stock solutions were stored at 18 °C. The stock solutions were diluted with ACN to working solutions before analysis. 2.2. Synthesis of d4-DMED The pathway for synthesis of d4-DMED is shown in Fig. 2. 2.2.1. Synthesis of (2-amino-ethyl)-carbamic acid benzyl ester To a solution of 1,2-diaminoethane (30.0 g, 0.50 mmol) in DCM (500 mL) benzylorthochloroformiate (8.53 g, 50 mmol) in DCM
Fig. 2. Synthesis procedure of the d4-DMED reagents.
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
(250 mL) was added drop wise at 0 °C over 8 h. After stirring at ambient temperature over night, the mixture was washed five times with brine (100 mL) and then with water (100 mL). The combined organic phases were dried over MgSO4 and concentrated in vacuo to afford colorless oil, which solidifies in the refrigerator as a white solid (7.58 g, 39 mmol, 78%). 2.2.2. Synthesis of d4-(2-dimethylamino-ethyl)-carbamic acid benzyl ester To a vigorously stirred suspension of (2-amino-ethyl)-carbamic acid benzyl ester (2.22 g, 11.4 mmol) in methanol (30 mL), deuterium formaldehyde (20% in water, 6.3 mL) and NaBH3CN (1.03 g, 16 mmol) were added. After 15 min, acetic acid was added until the pH was neutral and stirring was continued for 24 h. After evaporation of solvents, the mixture was dissolved in DCM (30 mL) and washed by three times with brine (10 mL). After evaporation of the solvent under reduced pressure at 20 °C, the mixture was then purified by silica gel column chromatography (eluent: DCM/ MeOH/TEA, 94/1/5) to afford the expected compound (2.11 g, 9.5 mmol, 83% yield) as an orange oil. 2.2.3. Synthesis of d4-N1,N1-dimethylethane-1, 2-diamine (d4-DMED) To a solution of (2-dimethylamino-ethyl)-carbamic acid benzyl ester (449.1 mg, 2.0 mmol) in methanol (15 mL), 10% palladium on activated charcoal (30 mg) was added. The reaction was performed at 50 °C over 6 h in the present of hydrogen gas (20 atm). The suspension was then filtered to remove Pd–C. 1H-NMR purity: 498% (Fig. S1). In addition, DMED and d4-DMED were mixed equally and subjected to LC–MS analysis. In addition, it could be seen that the peak area ratio of the two compounds was approximate to 1, indicating the high purity of the synthesized reagent (Fig. S2). 2.3. Plant materials Oryza sativa was grown in an artificial environmental chamber greenhouse at 30 °C under 16 h light/8 h dark photoperiods. 10 days old O. sativa leaves were harvested, weighted, immediately
343
frozen in liquid nitrogen, and stored at 80 °C until needed. 2.4. Plant sample pretreatment and stable isotope labeled derivatization The overall procedure for sample pretreatment and stable isotope labeled derivatization is summarized in Fig. 3. The pretreatment of plant samples was performed according to our previously published work [17]. Briefly, Plant materials (50 mg fresh weight) were frozen in liquid nitrogen, ground into powder, and extracted with 1 mL methanol containing 20% water at 4 °C for 12 h. The extract was centrifuged at 12,000 g under 4 °C for 20 min. The supernatant was collected and then passed through a C18 SPE-cartridge (1 mL, 50 mg) which was pre-conditioned with 2 mL methanol and 2 mL methanol containing 20% water. The elution was pooled and dried under nitrogen gas stream, reconstituted in 100 mL 1% FA H2O (v/v), and then extracted with ethyl ether (4 1 mL). The ether phases were combined and then evaporated to dryness under nitrogen gas stream. The standard GAs and the pretreated plant samples were separately dissolved in 100 mL ACN, and the stable isotope labeled derivatization was started by adding 10 mL TEA (10 mmol/mL), 10 mL CMPI (20 mmol/mL). After mixing, 20 mL DMED (20 mmol/mL) was added to the plant extracts and 20 mL d4-DMED (20 mmol/mL) was added to the standards solution for derivatization. The reaction solutions were incubated at 40 °C for 5 min with shaking at 1500 rpm and then evaporated under a stream of nitrogen gas until dryness. At last, the “light” labeled plant samples and “heavy” labeled standard GAs were respectively dissolved in 50 mL ACN/H2O 5/95 (v/v) and mixed together. Then 30 mL of the combined solutions was injected for analysis. 2.5. HPLC–ESI-MS/MS analysis Analysis of GAs and its derivatives were performed on a HPLC– ESI-MS/MS system consisting of a Shimadzu MS-8050 mass spectrometer (Tokyo, Japan) with an electrospray ionization source (Turbo Ionspray), and a Shimadzu LC-20AD HPLC system (Tokyo, Japan), which was equipped with two LC-20AD pumps, a
Fig. 3. Procedure for sample pretreatment and stable isotope labeled derivatization.
344
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
SIL-20A autosampler, a CTO-20AC column thermostat and a DGU20A3 degasser. Data acquisition and processing were performed using LabSolutions software (version 5.53 sp2, Shimadzu, Tokyo, Japan). To illustrate the sensitivity enhancement by chemical labeling, both GAs and its derivatives were analyzed in multiple reaction monitoring (MRM) mode. Optimized parameters for quantification of DMED labeled GAs and the native GA molecules by MRM were listed in Table S1. The HPLC separation was performed on a Shimpack VP-ODS (150 mm 2.0 mm i.d., 5 mm, Shimadzu, Tokyo, Japan) column at 40 °C with a flow rate of 0.2 mL/min. GAs were analyzed in negative ion mode, FA in H2O (0.01%, v/v, solvent A) and ACN (solvent B) were used as mobile phase. Gradient elution was achieved by the following program: 2 min, 10% B, 20 min, 10– 70% B, 1 min, 70–10% B, 10 min, 10% B. DMED/d4-DMED labeled GAs were analyzed in positive ion mode. 5 mmol/L ammonium formate aqueous solution (solvent A) and ACN (solvent B) were chosen for the separation procedure. The gradient elution program was set as follows: 2 min 5% B, 20 min 5–45% B, 20–21 min 45–5% B, 30 min 5% B. The optimal ESI source conditions for GAs and its derivatives were as follows: DL temperature, 250 °C; heat block temperature, 400 °C; nebulizing gas, 2 L/min; drying gas, 10 L/min and heating gas, 10 L/min.
3. Results and discussion 3.1. Chemical derivatization CMPI was found to activate carboxyl groups to form a reactive ester, and then amines attacked the reactive ester rapidly to form an amide bond under slightly alkaline conditions [18–20]. In a preliminary experiment, DMED was added to the mixtures of GA1, GA3, GA4, GA7 and GA9 for derivatization at a molar ratio of 10/1 with CMPI as an activator. As CMPI also reacts with amines, CMPI was added to the GAs containing solution to active the analytes first and then DMED was put in for derivatization. The molar ratio of the derivatization reagent to the activator was fixed at 2:1 to avoid complete DMED consumption by CMPI. The reaction was performed at 40 °C for 1 h, and then the resulting solution was subjected to MS analysis by direct infusion to examine the derivatization products. As expected, in positive ion mode, GA9-DMED (m/z 387.2), GA7-DMED (m/z 401.2), GA4-DMED (m/z 403.2), GA3-DMED (m/z 417.2) and GA1-DMED (m/z 419.2) were distinctly observed on the MS spectrum, while m/z 102.2 and m/z 180.1 represent the parent ions of TEA and CMPI–DMED, respectively. And no apparent peaks of impurities were observed on the MS spectrum. In negative ion mode, m/z 127.1, which resulted from I of CMPI, was the main peak on the MS spectrum. And no apparent peaks of unlabeled GA9 (m/z 315.2), GA7 (m/z 401.2), GA4 (m/z 329.2), GA3 (m/z 331.2) and GA1 (m/z 347.2) were observed (Fig. S3). These experimental results indicate high conversion rates of the target analytes (GAs) to the products (GAs-DMED). 3.2. Optimization of derivatization conditions In order to achieve high derivatization efficiency, derivatization conditions including TEA concentration, reaction temperature, DMED amount and reaction time were optimized. Four bioactive GAs including GA1, GA3, GA4, GA7 (1 ng for each GA) were used for the optimization procedures. The reaction was stopped by immediate freezing at 20 °C for 10 min and then evaporated to dryness under a flow of nitrogen. d4-DMED labeled GAs (1 ng for each GA) were spiked before LC–MS/MS analysis to calibrate matrix effects and signal fluctuations during ionization. TEA concentration was firstly investigated. GAs were dissolved
in 100 μL ACN, followed by addition of 10 μL TEA of different concentrations ranging from 1 to 20 μmol/mL. Then 10 μL CMPI (20 μmol/mL) was added. After mixing, 20 μL DMED (20 μmol/mL) was put in for derivatization. The reaction was incubated at 40 °C for 60 min with shaking at 1500 rpm. Our results showed that the signal of derivatives kept stable when the concentration of TEA increased from 1 mmol/mL to 20 mmol/mL (Fig. 4A). Reaction temperature was then examined. Derivatization was performed under different temperatures from 10 to 60 °C for 60 min with excess amount of DMED as described above. The results showed that temperature had little influence on labeling efficiency (Fig. 4B). To optimize the amount of DMED, the molar ratios of DMED to the total amount of GAs were changed from 20 to 5000. The reaction was kept at 40 °C for 60 min. The results show that 200-fold molar excess of DMED was enough for the efficient labeling of GAs (Fig. 4C). At last, derivatization time was optimized ranging from 5 to 60 min with the TEA concentration, reaction temperature and DMED amount optimized in the previous procedures. The result showed that derivatization of GAs with DMED was very fast and 5 min was enough for efficient reaction (Fig. 4D). Taken together, the optimized derivatization conditions for GAs with DMED were 40 °C for 5 min with a 200/1 molar ratio of DMED/GAs. 10 mL TEA (10 mmol/mL) was added to create slightly alkaline conditions. Under the optimized derivatization conditions, derivatization efficiency was determined by dividing the peak areas of underivatized GAs after derivatization with equal amount of GA standards without derivatization, the ratio was then subtracted from one. The results showed that more than 99% of GAs were consumed (Table 1), suggesting that high derivatization efficiencies were achieved. As the plant samples are much more complex than standard solutions, the amounts of DMED should be reexamined for the derivatization of GAs in O. sativa samples. For this purpose, plant samples were spiked with 4 GAs (1 ng/g for each analytes) and pretreated according to procedures described in Section 2.4. Our results showed that the peak areas of GAs derivatives reached a plateau when the concentrations of DMED standard solutions were increased to 20 mmol/mL (Fig. S4). The stability of the derivatives was evaluated by monitoring the change at different time intervals. The results showed that the derivatives were stable for at least 24 h in ACN/H2O (5/95, v/v) at 25 °C (Fig. S5). 3.3. Fragmentation of the DMED labeled GAs The fragmentation behavior of DMED labeled GA3 was first investigated. Chemical structure and possible fragmentation pattern of DMED labeled GA3 (CE, 40 eV) were shown in Fig. 5. It shows that five main fragments were produced, and the signal of product ions with m/z 310.2 and m/z 372.2 were obvious higher than that of the other product ions. Fragments at m/z 372.2 was produced from the neutral loss (NL) of NH(CH3)2, while m/z 310.2 was resulted from the dissociation of GA3 skeleton. Product ions relative with dissociation of skeleton of analytes other than labeling reagent may help to improve the detection selectivity by tandem MS analysis. This assumption was proved by monitoring the MRM transitions of both 417.24 372.2 and 417.24310.2 during the analysis of plant samples. As shown in Fig. 6, the LC–MS/ MS chromatogram corresponding to MRM transition of 417.2 4310.2 showed much less interference and background noise compared with that of 417.24 372.2, which improved the signal to noise (S/N) ratio and decreased the limit of detection (LOD) values significantly. Weinstock et al. pointed out that it is important to consider the background noise for samples with
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
345
Fig. 4. Optimization of derivatization conditions for GAs standards: (A) optimization of TEA concentration; (B) optimization of reaction temperature; (C) optimization of DMED molar ratios over a total amount of GAs; (D) optimization of reaction time.
Table 1 Derivatization efficiencies under the optimum reaction condition. Analytes Derivatization efficiency (%)
GA8 99.1
GA29 99.3
GA3 99.0
GA1 99.6
GA6 99.7
GA5 99.7
GA34 99.4
GA51 99.7
GA7 99.2
GA4 99.9
GA9 99.8
Fig. 5. Chemical structure and possible fragmentation pattern of DMED-GA3 at CE ¼ 40 eV.
complicated matrix when comparing overall sensitivity improvements, as improvements in signal sensitivity can be completely neutralized by a detrimental rise in chemical noise [21]. As for other GAs, HPLC–MS/MS chromatogram derived from MRM transition of the structure specific fragments also owned more selectivity than that of the fragments resulting from the neutral loss of NH(CH3)2 (Fig. S6). Therefore, the structure specific fragments were chosen for quantitative analysis of GAs to enhance reliability of the method, even when they were not the fragments of the
highest abundance. 3.4. Signal enhancement of the derivatives GAs are a class of acid phytohormones that contain carboxyl groups in structures. An important issue facing us in the determination of GAs by ESI-MS is the poor sensitivity resulting from the low ionization efficiency in negative ion mode. Herein, we compared the limits of quantitation (LOQs) of GAs before and after
346
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
Fig. 6. The MRM chromatograms of DMED-GA3 in Oryza sativa leaves under different transitions. (A) 310.2, a structure specific fragment resulting from the dissociation of GA3 skeleton was chosen as the product ion for MRM; (B) 372.2, fragment after the neutral loss of NH(CH3)2 was chosen as the product ion for MRM. Table 2 LODs, LOQs of the 11 GAs with and without derivatization. Analytes LOQ/pg (S/N ¼ 10)
GA8 GA29 GA3 GA1 GA6 GA5 GA34 GA51 GA7 GA4 GA9
LOD/pg (S/N ¼ 3)
Nonlabeled
Labeled Nonlabeled
Labeled
18.18 13.79 0.77 3.64 0.75 3.64 2.03 6.80 0.64 2.94 4.29
2.31 2.45 0.08 0.19 0.47 0.25 0.89 0.21 0.07 0.17 0.19
0.69 0.74 0.03 0.06 0.14 0.08 0.27 0.06 0.02 0.05 0.06
5.45 4.14 0.23 1.09 0.23 1.09 0.61 2.04 0.19 0.88 1.29
Approximate sensitivity enhancement (fold)
8 6 9 19 2 15 2 32 9 17 23
derivatization under their optimal conditions. The signal enhancement factor was determined by comparing the LOQs of 11 GAs. As shown in Table 2, the LOQs of 11 GAs were decreased by 2– 32 folds upon derivatization.
3.5. Method validation The LC–MS/MS chromatogram of 11 GAs-DMED was shown in Fig. 7. The calibration curves were constructed by plotting the mean peak area ratios of DMED/d4-DMED derivatives against nominal concentration ratios based on data obtained from duplicate measurements. As shown in Table 3, good linearities were obtained with correlation coefficients (R2) of 0.9976–0.9999. The LODs and the LOQs were calculated as the concentration of the analytes at a S/N ratio of 3 and 10, respectively. The results showed that LODs and LOQs for the 11 derived GAs ranged from 0.02 to 0.74 pg and 0.07 to 2.45 pg, respectively (Table 2). In addition, the reliability and accuracy of the method were evaluated by spiking GAs standards into the rice matrices. The recoveries of 11 GAs in O. sativa leave samples at three concentration levels were found to be 72–128%. The relative standard deviations (RSDs) were 1.9–13.9% (Table 4). These results suggested that the accuracy and precision of the proposed method were satisfactory for the determination of GAs in plant samples, and the stable isotope labeling strategy worked well for absolute quantification of GAs.
Fig. 7. The MRM chromatograms of 11 DMED-derived GA (5 pg) standards analyzed by HPLC–ESI-MS/MS.
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
347
Table 3 Calibration curves of the 11 DMED-labeled GAs. Analytes Concentration range of GAs-DMED (ng/mL)
GA8 GA29 GA3 GA1 GA6 GA5 GA34 GA51 GA7 GA4
0.05–20 0.05–20 0.002–20 0.005–20 0.01–20 0.005–20 0.02–20 0.005–20 0.002–20 0.005–20
Fixed concentration of GAs-d4-DMED (ng/mL)
1 1 0.5 0.5 1 1 1 0.5 0.5 0.5
Nominal concentration ratios of DMED/d4-DMED labeled GAs
0.05–20 0.05–20 0.004–40 0.01–40 0.01–20 0.005–20 0.02–20 0.01–40 0.004–40 0.01–40
Table 4 Recoveries of 11 GAs in 50 mg Oryza sativa leaves (n¼ 3).
Calibration curve Slope
Intercept R2
1.0015 1.0524 1.0522 1.0646 0.9761 1.0423 1.0762 0.9777 0.9955 1.0104
0.0080 0.0976 0.0381 0.0824 0.0054 0.0179 0.0399 0.1090 0.0556 0.0014
0.9999 0.9976 0.9999 0.9998 0.9998 0.9999 0.9998 0.9994 0.9998 0.9996
Table 5 Determination of endogenous GAs in 50 mg (FW) Oryza sativa leaves (n¼3).
Analytes
Added (ng/g) low, medium, high
Recoveries (%)
RSD (%)
GA1 GA3 GA7 GA9 GA5 GA8 GA29 GA4 GA6 GA34 GA51
0.2, 1, 10 0.2, 1, 10 0.2, 1, 10 0.2, 1, 10 1, 5, 20 1, 5, 20 1, 5, 20 1, 5, 20 1, 5, 20 1, 5, 20 0.2, 1, 10
93, 104, 83 83, 87, 72 98, 99, 90 128, 106, 109 95, 89, 86 95, 77, 81 90, 88, 82 83, 87, 85 87, 80, 79 81, 83, 82 108, 102, 95
3.4, 5.1, 2.6 4.6, 3.7, 1.0 7.4, 7.5, 2.0 4.4, 1.4, 2.7 10, 1.5, 1.7 13.9, 6.4, 3.9 2.0, 8.0, 2.3 4.6, 7.3, 1.4 1.2, 8.0, 1.6 11.0, 4.9, 4.4 7.7, 8.3, 4.3
Analytes
Found (ng/g)
GA1 GA3 GA7 GA9 GA5 GA8 GA29 GA4 GA6 GA34 GA51
0.157 0.01 0.217 0.01 N.D. 0.58 7 0.04 N.D. N.D. N.D. 8.487 0.47 0.27 70.02 N.D. N.D.
N.D., not found.
suggesting that the high sensitive and reliable stable isotope labeling strategy can be successfully applied for identification and quantification of endogenous GAs in complex plant matrix
4. Conclusions
Fig. 8. The MRM chromatograms of detected endogenous GAs in 50 mg Oryza sativa leaves.
In this study, we developed an isotope mass probe labeling method for the sensitive and reliable analysis of GAs in plant samples by HPLC–ESI-MS/MS. The proposed method overcomes the limitations of using isotopically labeled internal standards in the traditional isotope dilution approach. Sensitivity was enhanced by incorporating ionized moieties into analytes to improve ionization efficiency. In addition, the derivatives could produce specific ions, which impart high selectivity for MRM analysis, thus enhanced the reliability of the method. The developed method was employed for identification and quantification of endogenous GAs in 50 mg O. sativa leaves. Five endogenous GAs including the bioactive GA1, GA3, GA4 were detected and quantified. In all, the high sensitive and reliable stable isotope labeling strategy has the potential to facilitate plant hormone research.
3.6. Determination of endogenous GAs in plant samples
Acknowledgments
GAs play important roles in plant growth and development. A satisfactory method should meet the need of detecting GAs of trace amount from complicated plant matrix. Fig. 8 showed the MRM chromatograms of endogenous GAs detected by the current method in 50 mg O. sativa leaves. It can be seen that five GAs including the bioactive GA1, GA3, GA4 could be successfully observed at the content ranging from 0.15 to 8.5 ng/g FW (Table 5),
The authors thank the National Natural Science Foundation of China for financial support (21475098, 91217309) and the Natural Science Foundation of Hubei Province, China (2014CFA002). Appendix A. Supplementary material Supplementary data associated with this article can be found in
348
Y.-H. Hao et al. / Talanta 144 (2015) 341–348
the online version at http://dx.doi.org/10.1016/j.talanta.2015.06. 056.
References [1] M. Hamano, Y. Yamato, H. Yamazaki, H. Miura, J. Hortic. Sci. Biotech. 77 (2002) 220–225. [2] S. Yamaguchi, M. Ogawa, Y. Yamauchi, Y. Kamiya, Plant Cell Physiol. 43 (2002), S20–S20. [3] N. Arnaud, T. Girin, K. Sorefan, S. Fuentes, T.A. Wood, T. Lawrenson, R. Sablowski, L. Ostergaard, Genes Dev. 24 (2010) 2127–2132. [4] T. Urbanova, D. Tarkowska, O. Novak, P. Hedden, M. Strnad, Talanta 112 (2013) 85–94. [5] J.F. Liu, J. Ding, B.F. Yuan, Y.Q. Feng, Analyst 139 (2014) 5605–5613. [6] S.C. Liu, W.Q. Chen, L. Qu, Y. Gai, X.N. Jiang, Anal. Bioanal. Chem. 405 (2013) 1257–1266. [7] Z. Ma, L. Ge, A.S. Lee, J.W. Yong, S.N. Tan, E.S. Ong, Anal. Chim. Acta 610 (2008) 274–281. [8] M. Muller, S. Munne-Bosch, Plant Methods 7 (2011) 37. [9] X. Pan, X. Wang, J. Chromatogr. B 877 (2009) 2806–2813.
[10] X. Pan, R. Welti, X. Wang, Nat. Protoc. 5 (2010) 986–992. [11] Y.Q. Huang, J.Q. Liu, H.Y. Gong, J. Yang, Y.S. Li, Y.Q. Feng, Analyst 136 (2011) 1515–1522. [12] Y.Q. Huang, G.D. Ruan, J.Q. Liu, Q. Gao, Y.Q. Feng, Anal. Biochem. 416 (2011) 159–166. [13] S. Ogawa, S. Ooki, K. Shinoda, T. Higashi, Anal. Bioanal. Chem. 406 (2014) 6647–6654. [14] X. Sun, Y. Ouyang, J. Chu, J. Yan, Y. Yu, X. Li, J. Yang, C. Yan, J. Chromatogr. A 1338 (2014) 67–76. [15] T. Toyo’oka, J. Pharm. Biomed. Anal. 69 (2012) 174–184. [16] S. Zhang, J. You, S. Ning, C. Song, Y.R. Suo, J. Chromatogr. A 1280 (2013) 84–91. [17] M.L. Chen, Y.Q. Huang, J.Q. Liu, B.F. Yuan, Y.Q. Feng, J. Chromatogr. B 879 (2011) 938–944. [18] W.J. Leavens, S.J. Lane, R.M. Carr, A.M. Lockie, I. Waterhouse, Rapid Commun. Mass Spectrom. 16 (2002) 433–441. [19] Y.Q. Huang, Q.Y. Wang, J.Q. Liu, Y.H. Hao, B.F. Yuan, Y.Q. Feng, Analyst 139 (2014) 3446–3454. [20] Y. Tang, J. Xiong, H.P. Jiang, S.J. Zheng, Y.Q. Feng, B.F. Yuan, Anal. Chem. 86 (2014) 7764–7772. [21] M. Star-Weinstock, B.L. Williamson, S. Dey, S. Pillai, S. Purkayastha, Anal. Chem. 84 (2012) 9310–9317.