Direct Determination of Six Cytokinin Nucleotide Monophosphates in

Subscriber access provided by UNIV OF MISSOURI ST LOUIS

Article

Direct Determination of Six Cytokinin Nucleotide Monophosphates in Coconut Flesh by Reversed-Phase Liquid Chromatography-Tandem Mass Spectrometry Zhao-yun Cao, Youning Ma, Li-Hua Sun, Ren-Xiang Mou, Zhi-Wei Zhu, and Mingxue Chen J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.7b03798 • Publication Date (Web): 23 Oct 2017 Downloaded from http://pubs.acs.org on October 27, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Agricultural and Food Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29

Journal of Agricultural and Food Chemistry

1

Direct Determination of Six Cytokinin Nucleotide

2

Monophosphates in Coconut Flesh by Reversed-Phase

3

Liquid Chromatography-Tandem Mass Spectrometry

4

†,§

†,§

‡

†

†

5

Zhao-Yun Cao , You-Ning Ma , Li-Hua Sun , Ren-Xiang Mou , Zhi-Wei Zhu ,

6

Ming-Xue Chen*,

†,iD

7 8 9 10 11

†

Rice Product Quality Inspection and Supervision Center of Ministry of Agriculture,

China National Rice Research Institute, Hangzhou 310006, China ‡

Institute of Health Food, Zhejiang Academy of Medical Sciences, Hangzhou 310013,

China

1

ACS Paragon Plus Environment


Page 2 of 29

12

ABSTRACT: Coconut contains many uncharacterized cytokinins that have important

13

physiological effects in plants and humans. In this work, a method based on liquid

14

chromatography-tandem mass spectrometry was developed for identification and

15

quantification of six cytokinin nucleotide monophosphates in coconut flesh. Excellent

16

separation was achieved using a low-coverage C18 bonded-phase column with an acidic

17

mobile phase, which greatly improved the retention of target compounds. To enable

18

high-throughput analysis, a single-step solid-phase extraction using mixed-mode

19

anion-exchange cartridges was employed for sample preparation. This proved to be an

20

effective method to minimize matrix effects and ensure high selectivity. The limits of

21

detection varied from 0.06 to 0.3 ng/mL and the limits of quantification ranged from 0.2 to

22

1.0 ng/mL. The linearity was statistically verified over two orders of magnitude, giving

23

coefficients of determination (R2) greater than 0.9981. The mean recoveries were from

24

81% to 108%, the intra-day precision (n = 6) was less than 11%, and the inter-day

25

precision (n = 11) was within 14%. The developed method was applied to the

26

determination of cytokinin nucleotide monophosphates in coconut flesh samples, and four

27

of them were successfully identified and quantified. The results showed that trans-zeatin

28

riboside-5'-monophosphate was the dominant cytokinin with concentrations of 2.7–34.2

29

ng/g, followed by N6-isopentenyladenosine-5'-monophosphate (≤ 12.9 ng/g), while the

30

concentrations

31

riboside-5'-monophosphate were less than 2.2 and 4.9 ng/g, respectively.

32

KEYWORDS:

33

solid-phase extraction, coconut flesh

of

cis-zeatin

cytokinin

riboside-5'-monophosphate

nucleotide

monophosphates,

and

LC-MS/MS,

dihydrozeatin

mixed-mode

34 2


Page 3 of 29


35

INTRODUCTION

36

The coconut (Cocos nucifera L.) is an important fruit tree in tropical and sub-tropical

37

regions. Its fruit has been widely consumed as food and beverages all over the world, not

38

only for its food value, but also for its medical benefits (e.g., activity against myocardial

39

infarction, anti-inflammatory and wound healing properties).1 An increasing number of

40

studies have suggested that certain medical effects are closely correlated with the presence

41

of uncharacterized cytokinins in coconut, especially in coconut flesh,2–4 but these

42

compounds have remained unidentified until now.

43

Cytokinins, an important group of plant hormones, play crucial roles in the regulation of

44

cell division and differentiation, and control various processes during plant growth and

45

development.5–7 The naturally occurring cytokinins are adenine derivatives carrying either

46

an isoprene-derived (including isopentenyladenine-type and zeatin-type) or aromatic (i.e.,

47

aromatic cytokinins) side chain at the N6 terminus.8 Nucleosides, nucleotides and other

48

sugar-conjugates have also been identified from various plant species.7 Among them, the

49

nucleotide forms (see Table 1 for chemical structures) have attracted increasing attention

50

in recent years worldwide, not only because they can act as translocation forms of active

51

cytokinins in plants,9 but also because they exhibit interesting therapeutic effects such as

52

triggering apoptosis of cancer cells.10 Unfortunately, the physiological effects of these

53

nucleotides have not yet been fully elucidated.9 This could be partially due to the lack of

54

direct analytical methods for detection of various cytokinin nucleotides.11

55

Typically, cytokinin species are present in extremely low concentrations (less than 30

56

pmol/g of fresh weight) in plants.7,12 This, together with the matrix complexity and their 3



Page 4 of 29

57

structural similarity requires not only a sensitive detector, but also highly efficient

58

separation and adequate cleanup of sample extracts. Although mass spectrometry

59

(MS)-based quantification of cytokinin nucleotides has been widely accepted due to its

60

high sensitivity and selectivity,12,13 separation remains a challenge because of the high

61

polarity of the compounds. Therefore, an extra step involving either enzymatic

62

dephosphorylation14 or esterification15 has often been required prior to routine

63

reversed-phase (RP) separation and MS detection. The biggest drawback of such a

64

procedure is that it fails to discriminate between individual nucleotide species. In addition,

65

the results obtained using this approach are greatly affected by phosphatase activity or

66

derivatization efficiency. To obtain information about intact cytokinin nucleotides,

67

ion-pairing RP chromatography has been employed for simultaneous separation of

68

multiple nucleotides and other intracellular metabolites,16 but it is often incompatible with

69

MS detection because of the high salt concentrations in the mobile phases, resulting in

70

poor reproducibility. More recently, separation by means of capillary electrophoresis (CE)

71

has been reported,17–19 which enabled identification and quantification of individual intact

72

cytokinin nucleotides. Although this technique is very attractive because it can provide

73

high resolving power and separation speed, the major limitation is its relatively poor

74

concentration sensitivity compared to liquid chromatography.20–22

75

To date, various tools for extraction and pre-concentration of cytokinin nucleotides from

76

plant extracts have been reported, such as liquid-liquid extraction,18 solid-phase extraction

77

(SPE, e.g., C18 SPE, ion exchange SPE, and mixed-mode SPE),17,23 and immunoaffinity

78

chromatography (IAC).24 Although IAC offers higher selectivity than conventional

4


Page 5 of 29


79

liquid–liquid extraction, throughput is relatively low because of complicated and

80

multi-step procedures. For example, cytokinin nucleotides were typically washed from

81

IAC columns, cleaned up with ion-exchange SPE, subjected to hydrolysis with alkaline

82

phosphatase, and then further cleaned up using another IAC column.24 More recently,

83

SPE-based methodology has attracted increasing attention as an alternative protocol for

84

separation of cytokinin nucleotides from complex matrices,12 since the distinct advantages

85

of SPE over liquid–liquid extraction or IAC are that it consumes less organic solvent and a

86

wider range of extraction mechanisms can be utilized. Unfortunately, most current SPE

87

methods still require at least a dual-step SPE procedure, such as a combination of RP SPE

88

and mixed-mode cation exchanger SPE,12,25 making the process labor-intensive and costly.

89

Thus, to achieve high-throughput sample preparation, a single-step method is desirable.

90

In past decades, although several methods for measuring these substances in coconut

91

have been reported, few components have been identified because of the poor sensitivity

92

of the methods.3,17 Herein, we describe the development and validation of an analytical

93

method for the analysis of six cytokinin nucleotide monophosphates in coconut flesh

94

based on liquid chromatography-tandem mass spectrometry (LC-MS/MS). The separation

95

was conducted on a low-coverage C18 bonded-phase column without using ion-pair

96

reagents or a time-consuming enzymatic dephosphorylation step and tedious derivatization.

97

Furthermore, a single-step SPE method using mixed-mode anion-exchange (MAX)

98

cartridges was also developed for cleanup of sample extracts for high-throughput analysis.

99

The method was then evaluated and used to screen for the presence of endogenous

100

cytokinin nucleotide monophosphates in coconut flesh samples.

5



Page 6 of 29

101

EXPERIMENTAL

102

Chemicals and Solvents. HPLC-grade acetonitrile and methanol were obtained

103

from Merck (Shanghai, China). Formic acid (for mass spectrometry, ca. 98%), phosphoric

104

acid (ACS reagent, ≥ 85 wt.% in H2O), and ammonium hydroxide solution (ACS reagent,

105

28.0%−30.0% NH3 basis) were purchased from Sigma–Aldrich (Shanghai, China).

106

Analytical standards of trans-zeatin riboside-5'-monophosphate (ZMP), dihydrozeatin

107

riboside-5'-monophosphate (DZMP), cis-zeatin riboside-5'-monophosphate (cZMP),

108

kinetin riboside-5'-monophosphate (KMP), N6-isopentenyladenosine-5'-monophosphate

109

(iPMP) and N6-benzyladenosine-5'-monophosphate (BAMP) were obtained from

110

OlChemIm (Olomouc, Czech Republic). The purity of all standards was above 95%. Oasis

111

MAX cartridges (3 mL, 60 mg) were purchased from Waters Corporation (Shanghai,

112

China). All reagent concentrations are expressed as volume fractions in this study, unless

113

otherwise stated. Ultra-pure water used throughout the study was obtained using a Milli-Q

114

system (Milford, MA, USA).

115

Sample Preparation. Coconut fruits were obtained from local supermarkets. On

116

arrival at the laboratory, they were removed from the shell with a spoon and chopped. The

117

chopped samples were frozen with liquid nitrogen and then ground with a ball-mill at 25

118

Hz for 30 s (Model MM400, Retsch GmbH, Haan, Germany). The homogenized samples

119

were stored in the dark at −80°C prior to processing.

120

An accurately weighed 1 g portion was added to 5 mL of pre-cooled extraction solvent

121

(4°C, 80% methanol in water) in a 15 mL polypropylene centrifuge tube. The mixture was

122

extracted for 10 min with the aid of ultrasonication (KQ-800TDE, Kunshan, China), and

6


Page 7 of 29


123

then centrifuged at 6000 g for 10 min at 4°C. The supernatant was collected in a 10 mL

124

volumetric flask. The residual material was extracted again with 3 mL of pre-cooled

125

extraction solvent in the same way. The combined supernatants were made up to the full

126

volume with 80% methanol in water, and then subjected to SPE cleanup.

127

MAX cartridges (3 mL, 60 mg) were conditioned by sequential washing with 2 mL

128

methanol and 2 mL ammonium hydroxide solution (5%). A 2 mL aliquot of extract was

129

then loaded onto the equilibrated cartridge. After sequential washing with 1 mL water and

130

1 mL methanol, the retained analytes were eluted with 2 mL formic acid in methanol (5%).

131

The eluate was concentrated almost to dryness at 45°C under a stream of nitrogen. The

132

residue was dissolved in 1.0 mL water. Aliquots of supernatant were transferred to

133

autosampler vials and submitted to LC-MS/MS analysis.

134

Preparation of Standards. Individual stock solutions of ZMP, DZMP, cZMP,

135

KMP, iPMP and BAMP, were prepared at concentrations of 100 µg/mL by accurately

136

weighing appropriate amounts of the standard compounds and dissolving them in 10 mM

137

phosphoric acid. Mixed stock solutions (5 µg/mL) were prepared by transferring aliquots

138

(0.5 mL) of the individual stock solutions to a 10 mL volumetric flask and diluting to 10

139

mL with water. The solutions were stored at −80°C in the dark. For all analytes, five

140

calibration concentrations were used to define the standard curve (0.6, 4, 20, 100 and 500

141

ng/mL for ZMP and KMP, 1, 5, 25, 100 and 500 ng/mL for DZMP, and 0.2, 1, 5, 25 and

142

100 ng/mL for the remaining compounds), which were prepared daily by serial dilution of

143

the mixed standard stock solutions with water. The solutions were stored at 4°C in the

144

dark.

7



Page 8 of 29

145

LC-MS/MS Analysis. The analysis of cytokinin nucleotide monophosphates was

146

performed using a TSQ Quantum Access Max mass spectrometer (Thermo Fisher

147

Scientific, Bremen, Germany), equipped with an electrospray ionization interface (ESI).

148

Selected reaction monitoring (SRM) data were acquired and processed in positive ESI

149

mode. The optimized ESI operating conditions were as follows: needle spray voltage, 3.0

150

kV; sheath gas pressure (N2), 35 arbitrary units; auxiliary gas pressure (N2), 5 arbitrary

151

units; ion transfer capillary temperature, 300°C; collision gas pressure (Ar), 1.5 mTorr.

152

Other parameters, such as transition ions, collision energies, and tube lens offsets for all

153

analytes from the optimization procedure are listed in Table 2.

154

The LC analyses were performed using a Thermo Finnigan Surveyor LC system

155

equipped with a Surveyor MS pump, a solvent degasser, and a Surveyor autosampler. An

156

HSS T3 column (150 × 2.1 mm, i.d. 3.5 µm, Waters Corporation) was chosen to separate

157

target compounds at a column temperature of 30°C. The mobile phases consisted of water

158

containing 0.1% formic acid as eluent A and acetonitrile as eluent B at a flow rate of 0.3

159

mL min–1. The linear gradient was as follows: 0–14 min, 5%–25% B; 14.1–20 min, 90% B;

160

20.1–28 min, 5% B. The injection volume was 5 µL.

161

Method Validation. The method was validated according to the National Standards

162

of the People’s Republic of China and relevant literature reports26–28 in terms of limit of

163

detection (LOD), limit of quantification (LOQ), linearity, selectivity, specificity, accuracy

164

and precision, including inter- and intra-day precision.

165

The LOD and LOQ were calculated by analyzing blank spiked samples as

166

concentrations corresponding to a signal-to-noise ratio of 3 or above and 10 or above,

8


Page 9 of 29


167

respectively. If a blank sample contained certain targets, their LOD and LOQ were

168

estimated by injection of standard solutions. Calibration curves were constructed by linear

169

regression of the analyte peak area versus concentration at six levels with equal weighting

170

(three replicates for each level). To evaluate the (non) linearity of the regression model

171

over the working range concentration, Mandel's fitting test was used to evaluate whether

172

an alternative regression model was appropriate.29 Furthermore, the significance of the

173

intercept at the 95% confidence level was also assessed by running a t-test. The selectivity

174

and specificity were evaluated by comparing the responses of the spiked and blank

175

samples obtained from eight batches of coconut fruits.

176

The selectivity of the developed method was assessed by the absence of interfering

177

peaks near the elution times of the analytes for blank chromatograms of different coconut

178

fruits.28 The specificity was tested using spiked ZMP, KMP, iPMP, BAMP, cZMP and

179

DZMP at 5 ng/g in blank samples. For identification purposes, the retention times of six

180

nucleotides in standards and samples were compared at a tolerance of ± 5% required by

181

the 2002/657/EC Decision.30 The relative ion intensities of analytes studied in the standard

182

solution and the spiked samples at the concentration levels used for the calibration curve

183

were compared in accordance with the criteria.30

184

Since there were no reference materials available, accuracy was calculated as recovery

185

assessed by measuring cytokinin nucleotide concentrations in coconut flesh samples

186

spiked at three different levels (six replicates per level). Recovery was determined by

187

comparing the amount of cytokinin nucleotide added with the amount detected. The intra-

188

and inter-day precisions were assessed by replicate analyses of the same spiked sample on

9



189

Page 10 of 29

the same day with six replicates and on 11 sequential days with 11 replicates, respectively.

190

RESULTS AND DISCUSSION

191

Optimization of Liquid Chromatography Conditions. Since cytokinin

192

nucleotides have very similar structures, especially ZMP and cZMP that have identical

193

precursor and product ions (see Table 2), sufficient chromatographic separation is required

194

prior to MS detection. Thus, the analytical column and eluent conditions were both

195

optimized.

196

It has been suggested that cytokinin nucleotides are weakly retained on traditional

197

reversed-phase columns due to their polar and hydrophilic nature, and that hydrophilic

198

interaction liquid chromatography (HILIC) could be a promising candidate to separate

199

these analogous compounds.31 Thus, several typical HILIC columns were initially

200

investigated to achieve adequate separation and retention of these six compounds,

201

including bare silica and silica or polymeric materials modified with amide, amino (–NH2),

202

or zwitterionic functional groups as the stationary phases. Unfortunately, poor retention

203

was observed with these analytical columns for all analytes (data not shown), and there

204

was no obvious improvement when different mobile phases (including water, buffer salt

205

solutions at various pH values, and acetonitrile) were used. The reason for the failure to

206

separate these compounds on several HILIC columns was probably because the analytes

207

tested in previous reports22,31 had different chemical properties compared to those in our

208

study. Additionally, the mobile phase (usually non-volatile) used in the previous studies

209

cannot be tested in our experiment because of incompatibility with mass spectrometry

210

interfaces. 10


Page 11 of 29


211

In view of these results, we then proposed a liquid chromatography method using an

212

HSS T3 column, a low-coverage C18 bonded-phase column designed to retain polar

213

compounds in reversed-phase mode.31 The HSS T3 column gave better retention and

214

separation for most of the cytokinin nucleotides than in HILIC mode. Nevertheless,

215

separation of ZMP and cZMP isomers was not achieved, and severely distorted and tailing

216

peaks were also observed for iPMP and BAMP in this case (Figure 1A). Further

217

optimization of the mobile phase composition was therefore performed in order to gain

218

better resolution and peak shapes. As shown in Figure 1B, an improved peak shape and a

219

significant increase in retention (especially for ZMP, cZMP and DZMP) were observed for

220

all analytes when formic acid was added to the aqueous mobile phase (containing 0.05%

221

formic acid). Subsequently, we found that excellent peak shapes were obtained for all

222

nucleotides at formic acid concentrations up to 0.1%, and baseline separation of two of the

223

isomers was achieved (Figure 1C). However, the retention of most analytes was not

224

significantly affected by a further increase of formic acid concentration. This can be

225

explained by the fact that the pH value of the mobile phase (i.e., a mixture of 0.1% formic

226

acid and acetonitrile, an approximate pH of 2.7) may be close to the isoelectric point

227

(approximately 2.5) of these compounds,17 where the retention maxima are found for

228

amphoteric substances on RP materials. Further addition of formic acid does not

229

significantly affect the pH value because of its weak ionic nature, and accordingly

230

retention is almost unchanged.

231

Optimization of SPE Procedure. In the present method, a single MAX SPE process

232

was chosen to pre-concentrate and clean up the extract of coconut flesh. Potential

11



Page 12 of 29

233

interfering compounds were removed from the co-extracts by successive washes with

234

water and methanol after introduction of the crude extract, dissolved in 80% methanol,

235

into the MAX cartridge. Subsequently, the analytes were eluted from the column by

236

methanol acidified with formic acid, in which the content of formic acid was optimized

237

between 1% and 10% in order to achieve maximum recoveries. The results showed that

238

5% formic acid in methanol as the elution solvent was sufficient for the purpose, giving

239

recoveries of 96%–107% for all analytes. On the basis of these results, it was clear that the

240

interaction between solute molecules and MAX SPE sorbents was attributable mainly to

241

an ion exchange mechanism. Hence, to avoid any loss of analytes during sample loading

242

and washing steps, a basic environment was provided by conditioning with 5%

243

ammonium hydroxide solution prior to sample loading.

244

In general, apart from pre-concentration, the aim of the SPE procedure was primarily to

245

eliminate or reduce the matrix effects (MEs) and improve accuracy of the analytical

246

method. To evaluate the ME, several calibration curves were obtained by injecting

247

matrix-matched and solvent (water) samples. The ME was quantified using the following

248

formula:32 ME (%) = [slope (matrix-matched calibration)/slope (calibration curves in

249

solvent) − 1] × 100. Each calibration curve was composed of the same five concentrations

250

(1–100 ng/mL) in triplicate. ME values below −15% or above 15% were regarded as

251

matrix suppression or enhancement. For values within that range, it can be assumed that

252

there was no matrix effect.33

253

As shown in Figure 2, the signals for all nucleotides were greatly suppressed in coconut

254

flesh extract, with matrix effects in the range of −46% to −17%, when the sample of crude

12


Page 13 of 29


255

extract without any cleanup was used. Satisfyingly, the matrix effect for each nucleotide

256

was significantly reduced, or even eliminated at whatever concentration it was added,

257

when the crude extract of the sample was cleaned up using MAX SPE cartridges, with

258

matrix effects ranging from −7% to 4%. In other words, the developed MAX SPE method

259

was sufficient to eliminate matrix interference from coconut flesh extracts, making it

260

possible to achieve accurate quantification of nucleotides through solvent calibration.

261

Method Validation. As listed in Table 3, the LOD values for six cytokinin nucleotides

262

were from 0.06 to 0.3 ng/mL, corresponding to 0.14–0.69 nM, while the LOQs ranged

263

from 0.2 to 1.0 ng/mL. Compared to other methods, the LODs were far lower than those

264

with CE-MS/MS (60–840 nM),17,18 and more than 10-fold lower than with LC-MS/MS

265

after derivatization.15 Furthermore, the LOD for each analyte obtained from these methods

266

was estimated by injection of standards instead of sample matrices. Thus, the proposed

267

method provided greater sensitivity for detection of low concentrations of cytokinin

268

nucleotides.

269

Five point standard calibration curves were obtained with concentrations ranging from

270

0.6–500 ng/mL for ZMP and KMP, 1–500 ng/mL for DZMP, and 0.2–100 ng/mL for the

271

remaining compounds. Linearity of these calibration curves was verified using Mandel’s

272

fitting test. According to Mandel’s fitting test, when the obtained F-value is below the

273

tabulated one (F0.99 = 9.33), the alternative regression model is not significantly better than

274

the linear regression model. In our cases, the obtained F-values were between 0.011 and

275

5.362 for all compounds tested (Table S1 in the Supporting Information). As a result, a

276

linear regression model was preferred, and all compounds showed good linearity with

13



Page 14 of 29

277

coefficients of determination (R2) ≥ 0.9981 over the investigated range. Furthermore, a

278

t-test performed on the intercept provided p values at the 95% confidence level ranging

279

from 0.074 to 0.941, demonstrating that all calibration equations were in the y = bx form

280

and free from systematic errors.

281

The results of the selectivity experiment showed that no interference peaks for the blank

282

samples occurred near the retention times of the analytes. The typical SRM traces of blank

283

and spiked samples are illustrated in Figure S1 of the Supporting Information. Since ZMP,

284

DZMP, cZMP and iPMP were found in the blank sample, the selectivity was assessed by

285

comparing the MS/MS spectra of these four analytes present in the matrix with the

286

MS/MS spectra of their standards.29 Almost no additional peaks were observed in the

287

MS/MS spectrum of each analyte in the matrix compared to the MS/MS spectrum of its

288

standard, which indicated that the peaks at 6.55, 6.68, 7.22 and 10.47 min were indeed

289

ZMP, DZMP, cZMP and iPMP, respectively, rather than interference (see Figure S2 of the

290

Supporting Information). Therefore, the described method was considered to be selective

291

for all compounds. In regard to specificity, the retention time in the proposed method was

292

quite stable for each compound tested. The maximum shift of retention time in the batch

293

(including 25 injections of sample) compared with standards was between −0.9% and

294

1.6% for all analytes. Moreover, firm identification of the analytes in the actual samples

295

was assured by performing the analysis in SRM mode. Thus, the method was specific for

296

target quantification according to the criteria.30

297

The accuracy of the proposed method was evaluated by spiking coconut flesh samples

298

at three different concentrations (1, 5, and 25 ng/g for cZMP, iPMP, and BAMP; 5, 25, and

14


Page 15 of 29


299

100 ng/g for ZMP, DZMP, and KMP). The experiments were repeated six times. As shown

300

in Table 3, the recoveries of all analytes were in the range of 81%–108% at all spiked

301

levels. In regard to overall precision, good results were obtained both in terms of intra-

302

and inter-day precision, with relative standard deviation values below 11% and 14% (see

303

Table 3), respectively. These validation experiments demonstrate the reliability and

304

usefulness of the method for routine determination of six cytokinin nucleotides in coconut

305

flesh materials.

306

Application. In total, 11 samples of fresh coconut fruits obtained from the local market

307

were analyzed for nucleotides by the validated method described here. Identification

308

criteria for the target compounds were the retention time and the intensity ratio of the two

309

monitored transitions for each compound. In addition, samples spiked at a level of 10 ng/g

310

were used for quality control and were analyzed with each batch of samples.

311

As listed in Table 4, ZMP, DZMP, cZMP and iPMP were successfully identified in

312

coconut flesh samples. KMP and BAMP, however, were not detected in the samples tested

313

in this study, indicating that they were present at levels below the LOD or even absent

314

from coconut flesh. Overall, from the quantitative results, it was clear that ZMP was the

315

dominant species in test samples, with concentrations of 2.7 to 34.2 ng/g, followed by

316

iPMP (≤ 12.9 ng/g), while the concentrations of cZMP and DZMP were less than 2.2 and

317

4.9 ng/g, respectively. The total ion chromatograms of sample 2 are shown in Figure S3 of

318

the Supporting Information. Moreover, the recoveries from samples used for quality

319

control were between 83% and 104%, demonstrating reliable quantification over the

320

whole batch of samples.

15



Page 16 of 29

321

Previously, the same compounds were also measured in coconut water by CE-MS/MS.17

322

Only ZMP was identified and estimated quantitatively at a concentration of 10 nM

323

(corresponding to 4.3 ng/mL) after 200-fold concentration of the extract, which indicated

324

that the concentrations of these compounds (except ZMP) were quite different from those

325

in coconut flesh. To the best of our knowledge, this is the first report of quantitative

326

analysis of naturally occurring cytokinin nucleotides in coconut flesh, which will

327

hopefully provide useful techniques to further understanding of cytokinins and their

328

physiological effects.

329

SUPPORTING INFORMATION

330

Typical selected-reaction-monitoring traces of the (A) blank and (B) spiked coconut

331

flesh samples for all compounds at 5 ng/g (Figure S1); the MS/MS spectra of (A–D)

332

ZMP, DZMP, cZMP and iPMP present in the matrix and (E–H) their standards (Figure

333

S2); the total ion chromatograms of sample 2 (Figure S3); calibration curves and

334

F-values obtained by Mandel’s Fitting test (99% confidence level) (Table S1).

335

AUTHOR INFORMATION

336

Corresponding Author

337

*Telephone: +86-571-63370275. E-mail: [email protected]

338

ORCID

339

Mingxue Chen: 0002-9400-4656

340

Author Contributions

341

§

Z.-Y. Cao and Y.-N. Ma contributed equally to this work.

16


Page 17 of 29


342

M.-X. Chen, Z.-Y. Cao and Y.-N. Ma designed the study. Z.-Y. Cao and Y.-N. Ma

343

carried out the laboratory work and analyzed data. L.-H. Sun and R.-X. Mou interpreted

344

the data. Z.-Y. Cao wrote the paper. Z.-W. Zhu gave technical support and conceptual

345

advice.

346

Funding

347

This work was supported by China National Quality and Safety Risk Assessment

348

Project for Grain and oil Crops (grant No. GJFP2014006).

349

Notes

350

The authors declare no competing financial interest.

351

REFERENCES

352

(1) DebMandal, M.; Mandal, S., Coconut (Cocos nucifera L.: Arecaceae): In health

353

promotion and disease prevention. Asian Pac. J. Trop. Med. 2011, 4, 241–247.

354

(2) Vermeulen, K.; Strnad, M.; Krystof, V.; Havlicek, L.; Van der Aa, A.; Lenjou, M.;

355

Nijs, G.; Rodrigus, I.; Stockman, B.; van Onckelen, H.; Van Bockstaele, D. R.; Berneman,

356

Z. N., Antiproliferative effect of plant cytokinin analogues with an inhibitory activity on

357

cyclin-dependent kinases. Leukemia 2002, 16, 299–305.

358

(3) Ge, L.; Yong, J. W. H.; Tan, S. N.; Yang, X. H.; Ong, E. S., Analysis of some

359

cytokinins in coconut (Cocos nucifera L.) water by micellar electrokinetic capillary

360

chromatography after solid-phase extraction. J. Chromatogr. A 2004, 1048, 119–126.

361 362 363

(4) Shaw, M.; Srivastava, B. I. S., Purine-like substances from coconut endosperm and their effect on senescence in excised cereal leaves. Plant Physiol. 1964, 39, 528–532. (5) Mok, D. W. S.; Mok, M. C., Cytokinin metabolism and action. Annu. Rev. Plant. 17



364 365 366 367 368

Page 18 of 29

Biol. 2001, 52, 89–118. (6) Amasino, R., 1955: Kinetin Arrives. The 50th anniversary of a new plant hormone. Plant Physiol. 2005, 138, 1177–1184. (7) Sakakibara, H., Cytokinins: activity, biosynthesis, and translocation. Annu. Rev. Plant Biol. 2006, 57, 431– 449.

369

(8) Matsumoto-Kitano, M.; Kusumoto, T.; Tarkowski, P.; Kinoshita-Tsujimura, K.;

370

Vaclavikova, K.; Miyawaki, K.; Kakimoto, T., Cytokinins are central regulators of

371

cambial activity. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 20027–20057.

372 373

(9) Davies, P. J., The Plant Hormones: Their Nature, Occurrence, and Functions, In Plant Hormones, Davies, P. J., Eds.; Springer: Dordrecht, Netherlands, 2010, pp. 1–15.

374

(10) Voller, J.; Zatloukal, M.; Lenobel, R.; Doležal, K.; Béreš, T.; Kryštof, V.; Spíchal,

375

L.; Niemann, P.; Džubák, P.; Hajdúch, M.; Strnad, M., Anticancer activity of natural

376

cytokinins: A structure–activity relationship study. Phytochemistry 2010, 71, 1350–1359.

377

(11) Mlejnek, P.; Doležel, P., Apoptosis induced by N6-substituted derivatives of

378

adenosine is related to intracellular accumulation of corresponding mononucleotides in

379

HL-60 cells. Toxicol. in Vitro 2005, 19, 985–990.

380 381 382 383

(12) Tarkowski, P.; Ge, L.; Yong, J. W. H.; Tan, S. N., Analytical methods for cytokinins. TrAC-Trends Anal. Chem. 2009, 28, 323–335. (13) Bai, Y.; Du, F.; Liu, H., Determination strategies of phytohormones: recent advances. Anal. Methods 2010, 2, 1867–1873.

384

(14) Kakimoto, T., Identification of plant cytokinin biosynthetic enzymes as

385

dimethylallyl diphosphate:ATP/ADP isopentenyltransferases. Plant Cell Physiol. 2001, 42,

18


Page 19 of 29


386 387

677–685. (15) Nordström, A.; Tarkowski, P.; Tarkowska, D.; Dolezal, K.; Åstot, C.; Sandberg, G.;

388

Moritz, T., Derivatization for LC-Electrospray Ionization-MS: A tool for improving

389

reversed-phase separation and esi responses of bases, ribosides, and intact nucleotides.

390

Anal. Chem. 2004, 76, 2869–2877.

391

(16) Luo, B.; Groenke, K.; Takors, R.; Wandrey, C.; Oldiges, M., Simultaneous

392

determination of multiple intracellular metabolites in glycolysis, pentose phosphate

393

pathway and tricarboxylic acid cycle by liquid chromatography–mass spectrometry. J.

394

Chromatogr. A 2007, 1147, 153–164.

395

(17) Ge, L.; Yong, J. W. H.; Tan, S. N.; Yang, X. H.; Ong, E. S., Analysis of cytokinin

396

nucleotides

in

coconut

(Cocos

nucifera

L.)

water

using

capillary

zone

397

electrophoresis-tandem mass spectrometry after solid-phase extraction. J. Chromatogr. A

398

2006, 1133, 322–331.

399

(18) Béres, T.; Gemrotová, M.; Tarkowski, P.; Ganzera, M.; Maier, V.; Friedecký, D.;

400

Dessoy, M. A.; Wessjohann, L. A.; Spíchal, L.; Strnad, M.; Doležal, K., Analysis of

401

cytokinin nucleotides by capillary zone electrophoresis with diode array and mass

402

spectrometric detection in a recombinant enzyme in vitro reaction. Anal. Chim. Acta 2012,

403

751, 176–181.

404 405

(19) Chen, X.-J.; Yang, F.-Q.; Wang, Y.-T.; Li, S.-P., CE and CEC of nucleosides and nucleotides in food materials. Electrophoresis 2010, 31, 2092–2105.

406

(20) Kok, M. G. M.; Somsen, G. W.; de Jong, G. J., Comparison of capillary

407

electrophoresis–mass spectrometry and hydrophilic interaction chromatography–mass

19



408 409 410

Page 20 of 29

spectrometry for anionic metabolic profiling of urine. Talanta 2015, 132, 1–7. (21) Tamizi, E.; Jouyban, A., The potential of the capillary electrophoresis techniques for quality control of biopharmaceuticals-A review. Electrophoresis 2015, 36, 831–858.

411

(22) Aral, H.; Haşimi, D.; Aral, T.; Levent, A.; Ziyadanoğullari, B., Separation,

412

optimization, and quantification of cytokinins by a recently developed amide-embedded

413

stationary phase. J. Liq. Chromatogr. Relat. Technol. 2017, 40 , 549–555.

414

(23) Ge, L.; Yong, J. W. H.; Tan, S. N.; Yang, X. H.; Ong, E. S., Analysis of positional

415

isomers of hydroxylated aromatic cytokinins by micellar electrokinetic chromatography.

416

Electrophoresis 2005, 26, 1768-1777.

417

(24) Novák, O.; Tarkowski, P.; Tarkowská, D.; Doležal, K.; Lenobel, R.; Strnad, M.,

418

Quantitative analysis of cytokinins in plants by liquid chromatography–single-quadrupole

419

mass spectrometry. Anal. Chim. Acta 2003, 480, 207–218.

420

(25) Béres, T.; Zatloukal, M.; Voller, J.; Niemann, P.; Gahsche, M. C.; Tarkowski, P.;

421

Novák, O.; Hanuš, J.; Strnad, M.; Doležal, K., Tandem mass spectrometry identification

422

and LC-MS quantification of intact cytokinin nucleotides in K-562 human leukemia cells.

423

Anal. Bioanal. Chem. 2010, 398, 2071–2080.

424

(26) Standardization Administration of China and General Administration of Quality

425

Supervision Inspection and Quarantine of the People’s Republic of China, Criterion on

426

quality control of laboratories-Chemical testing of food, GB/T 27404–2008.

427

(27) Bosco, R.; Daeseleire, E.; Van Pamel, E.; Scariot, V.; Leus, L., Development of an

428

ultrahigh-performance liquid chromatography–electrospray ionization–tandem mass

429

spectrometry method for the simultaneous determination of salicylic acid, jasmonic acid,

20


Page 21 of 29


430

and abscisic acid in rose leaves. J. Agr. Food Chem. 2014, 62, 6278–6284.

431

(28) Almeida Trapp, M.; De Souza, G. D.; Rodrigues-Filho, E.; Boland, W.; Mithofer,

432

A., Validated method for phytohormone quantification in plants. Front. Plant Sci. 2014, 5,

433

417–427.

434

(29) Fraselle, S.; Derop, V.; Degroodt, J. M.; Van Loco, J., Validation of a method for

435

the detection and confirmation of nitroimidazoles and the corresponding hydroxy

436

metabolites in pig plasma by high performance liquid chromatography-tandem mass

437

spectrometry. Anal. Chim. Acta 2007, 586, 383–393.

438

(30) European Commission Decision (2002/657/EC) of 12 August 2002 implementing

439

council directive 96/23/EC concerning the performance of analytical methods and

440

interpretation of results, Off. J. Eur. Communities (2002) 8–36.

441

(31) Aral, H.; Aral, T.; Ziyadanoğulları, B.; Ziyadanoğulları, R., Development of a

442

novel amide-silica stationary phase for the reversed-phase HPLC separation of different

443

classes of phytohormones. Talanta 2013, 116, 155–163.

444

(32) Besil, N.; Cesio, V.; Heinzen, H.; Fernandez-Alba, A. R., Matrix effects and

445

interferences of different citrus fruit coextractives in pesticide residue analysis using

446

ultrahigh-performance liquid chromatography–high-resolution mass spectrometry. J. Agr.

447

Food Chem. 2017, 65 (23), 4819–4829.

448

(33) Periat, A.; Kohler, I.; Thomas, A.; Nicoli, R.; Boccard, J.; Veuthey, J.-L.;

449

Schappler, J.; Guillarme, D., Systematic evaluation of matrix effects in hydrophilic

450

interaction chromatography versus reversed phase liquid chromatography coupled to mass

451

spectrometry. J. Chromatogr. A 2016, 1439, 42–53.

452 21



Page 22 of 29

453

Figure captions

454

Figure 1. Effect of the mobile phases on chromatographic resolution of six cytokinin

455

nucleotide monophosphates as analyzed on the HSS T3 column (150 × 2.1 mm, i.d. 3.5

456

µm). (A) The mobile phases consisted of acetonitrile and water; (B) The mobile phases

457

consisted of acetonitrile and water containing 0.05% formic acid; (C) The mobile phases

458

consisted of acetonitrile and water containing 0.1% formic acid. In these cases, the same

459

gradient procedure was used: 0–14 min, 5%–25% acetonitrile; 14.1–20 min, 90%

460

acetonitrile; 20.1–28 min, 5% acetonitrile. 1, ZMP; 2, DZMP; 3, cZMP; 4, KMP; 5, iPMP;

461

6, BAMP.

462 463

Figure 2. Investigation of matrix effects in analysis of cytokinin nucleotide

464

monophosphates in coconut flesh. Error bars represent means ± SD (n = 3)

22


Page 23 of 29


Table 1 Chemical Structures of the Six Cytokinin Nucleotide Monophosphates in this Work. HN N

R

N

O HO

P

N

O

N

O

OH

H

H

OH

H OH

H

R

compound

abbreviation

molecular weight (g/mol)

trans-Zeatin riboside-5'-monophosphate

ZMP

431.3

Dihydrozeatin riboside-5'-monophosphate

DZMP

433.4

cis-Zeatin riboside-5'-monophosphate

cZMP

431.3

Kinetin riboside-5'-monophosphate

KMP

427.3

N6-Isopentenyladenosine-5'-monophosphate

iPMP

415.3

N6-Benzyladenosine-5'-monophosphate

BAMP

437.3

CH2OH CH2

CH3 CH2OH

CH2

CH3 CH3

CH2

CH2OH

H 2C O

CH3 CH2

H2 C

CH3

23



Page 24 of 29

Table 2 Parameters for Selected Reaction Monitoring of Cytokinin Nucleotide Monophosphates.

compound

retention time (min)

tube lens (V)

ZMP

6.53

DZMP

transition 1

transition 2

relative intensitya (% of quantitative ion)

quantitative ion (m/z)

collision energy (V)

qualitative ion (m/z)

collision energy (V)

110

432.1/220.1

20

432.1/136.2

35

33

6.72

105

434.1/222.1

19

434.1/136.2

33

28

cZMP

7.24

110

432.1/220.1

20

432.1/136.2

35

44

KMP

9.15

100

428.1/216.1

17

428.1/148.1

31

16

iPMP

10.45

100

416.1/204.1

19

416.1/136.1

30

46

BAMP

11.98

114

438.1/226.1

20

438.1/91.2

50

29

a

the intensity ratio of the two monitored ions.

24


Page 25 of 29


Table 3 Summarized Validation Parameters for Six Cytokinin Nucleotide Monophosphates in Coconut Flesh. analyte

precisions (RSD, %) inter-day precision (n = 11)

intra-day precision (n = 6)

11

11

94

12

8

82

100

8

4

108

5

14

9

89

9

9

100

100

12

5

85

1

14

10

86

13

8

87

25

6

4

104

5

13

7

81

7

6

93

100

6

2

96

1

9

6

87

10

4

101

25

5

3

94

1

13

7

86

8

6

102

7

4

95

spiked level (ng/g)

limit of detection (ng/mL) (nM)

limit of quantitation (ng/mL)

5 ZMP

DZMP

cZMP

KMP

iPMP

BAMP

25

25

5

25

5

5 25

0.2

0.3

0.06

0.2

0.06

0.06

0.46

0.69

0.14

0.47

0.14

0.14

0.6

1.0

0.2

0.6

0.2

0.2

recovery (%)

25



Page 26 of 29

Table 4 Concentrationsa of Cytokinin Nucleotide Monophosphates in Coconut Flesh Samples. concentration (ng g–1)

sample No.

ZMP

DZMP

cZMP

iPMP

1

3.1 ± 0.5

2.1 ± 0.5

1.6 ± 0.2

n.d.b

2

6.6 ± 0.6

1.7 ± 0.4

1.7 ± 0.2

2.1 ± 0.3

3

34.2 ± 1.3

2.7 ± 0.4

2.2 ± 0.4

4.3 ± 0.4

4

2.7 ± 0.5

1.5 ± 0.3

n.d.

n.d.

5

9.7 ± 1.1

1.9 ± 0.3

1.6 ± 0.3

6.2 ± 0.4

6

12.0 ± 1.5

1.6 ± 0.4

n.d.

4.4 ± 0.7

7

19.0 ± 1.1

n.d.

n.d.

6.0 ± 0.5

8

18.1 ± 1.3

1.8 ± 0.3

n.d.

3.5 ± 0.6

9

14.3 ± 0.8

3.1 ± 0.2

n.d.

7.0 ± 0.9

10

3.7 ± 0.6

4.9 ± 0.3

n.d.

6.9 ± 0.9

11

5.9 ± 0.8

3.8 ± 0.5

1.7 ± 0.3

12.9 ± 1.0

a

Each reported value is the mean ± SD (n = 3).

b

Not detected using our method.

26


Page 27 of 29


Figure 1

27



Page 28 of 29

Figure 2

28


Page 29 of 29


Topic of Content

29


Direct Determination of Six Cytokinin Nucleotide Monophosphates in

Recommend Documents