Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
Article
Determination of Grayanotoxins from Rhododendron brachycarpum in Dietary Supplements and Homemade Wine by LC-Q-TOF-MS and LC-MS/MS Taeik Hwang, Eunyoung Noh, Ji Hye Jeong, Sung-Kwan Park, Dongwoo Shin, and Hoil Kang J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.7b05054 • Publication Date (Web): 13 Feb 2018 Downloaded from http://pubs.acs.org on February 14, 2018
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 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 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 22
Journal of Agricultural and Food Chemistry
Determination of Grayanotoxins from Rhododendron brachycarpum in Dietary Supplements and Homemade Wine by LC-QTOF-MS and LC-MS/MS
Taeik Hwang, Eunyoung Noh, Ji Hye Jeong, Sung-Kwan Park, Dongwoo Shin, Hoil Kang*
Division of Advanced Analysis, National Institute of Food and Drug Safety Evaluation, Ministry of Food and Drug Safety, Cheongju-si, Chungcheongbuk-do, 28159, Republic of Korea.
Corresponding Author (Tel: +82-43-719-5301; Fax: +82-43-719-5300; E-mail:
[email protected])
1
1 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 2 of 22
2
Abstract
3
A sensitive and specific high-performance liquid chromatography–quadrupole–time-of-flight mass spec-
4
trometry (LC-QTOF-MS) method combined with liquid chromatography–tandem mass spectrometry
5
(LC-MS/MS) was developed for the determination of grayanotoxins I and III in dietary supplements and
6
homemade wine. Grayanotoxins I and III were successfully extracted using solid-phase extraction car-
7
tridges, characterized by LC-QTOF-MS, and quantitated by LC-MS/MS. The LC-MS/MS calibration
8
curves were linear over concentrations of 10–100 ng/mL (grayanotoxin I) and 20–400 ng/mL (grayano-
9
toxin III). Grayanotoxins I and III were found in 51 foodstuffs, with quantitative determinations reveal-
10
ing total toxin concentrations of 18.4–101000 ng/mL (grayanotoxin I) and 15.3–56000 ng/mL (grayano-
11
toxin III). The potential of the validated method was demonstrated by successful quantitative analysis of
12
grayanotoxins I and III in dietary supplements and homemade wine; the method appears suitable for the
13
routine detection of grayanotoxins I and III from Rhododendron brachycarpum.
14
Keywords: grayanotoxin, LC-MS/MS, homemade wine, Rhododendron brachycarpum, LC-QTOF-MS
15
2 ACS Paragon Plus Environment
Page 3 of 22
Journal of Agricultural and Food Chemistry
16
Introduction
17
Grayanotoxins are a group of chemical compounds found naturally in the plant family Ericaceous,
18
which includes the Rhododendron species. Over 60 grayanotoxin-related compounds exist,1 of which
19
grayanotoxin I, 1, and grayanotoxin III, 2 (Figure 1) are the most toxic.2 The consumption of grayano-
20
toxins contained in mad honey, leaves, nectar, and flowers causes grayanotoxin poisoning, because of
21
the widespread belief in the medicinal properties of these plant-based products. Grayanotoxin poisoning
22
is often caused by misidentification of the plant species. More specifically, the ingestion of flowers,
23
flower juice, wine, or boiled extracts of R. schlippenbachii and R. brachycarpum has led to grayanotox-
24
in poisoning in Korea,3–6 as R. schlippenbachii (which contains grayanotoxins) seemingly resembles R.
25
mucronulatum. Indeed, R. mucronulatum flowers have been usually applied for preparing vegetable
26
pancakes, tea, and wine, in addition to being used as a herbal medicine for the treatment of the symp-
27
toms of the common cold, and as a diuretic.7 Grayanotoxins cause intoxication by increasing the mem-
28
brane permeability to sodium ions.8 Grayanotoxin-containing Rhododendron species grow naturally in
29
various countries, including the United States, Brazil, Spain, Portugal, Turkey, Nepal, and Japan.9
30
Interestingly, the toxic Rhododendron species, R. ponticum, which is native to the Black Sea region of
31
Turkey, is commonly used as a folk medicine to treat various health problems, including colds, edemas,
32
and general aches and pains.10 In addition, mad honey, which is produced from the pollen and nectar of
33
these flowers, is also used as a folk medicine to treat gastritis, ulcers, hyperglycemia, hypertension, ar-
34
thritis, and sexual impotence.11 However, as indicated above, the consumption of such products can
35
cause intoxication in humans, and the severity of symptoms may vary from mild to life-threatening, de-
36
pending on the quantity consumed. Indeed, grayanotoxin poisoning remains one of the most frequently
37
encountered causes of food intoxication in Turkey, although cases have also been reported in Austria,
38
Korea, and Nepal.12-14
39
Recently, concerns regarding poisoning from Rhododendron species were raised by the German Federal
40
Institute for Risk Assessment, which called for a greater systematic control of honey from specific re3 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 4 of 22
41
gions. Thus, due to the unseen danger to public health from grayanotoxins, the development of rapid
42
and reliable analytical methods for the detection of grayanotoxins in Rhododendron-containing dietary
43
supplements and homemade wines is of particular importance. However, to date, only a few analytical
44
methods have been reported for the determination of grayanotoxin in honey samples and plant materials,
45
such as thin-layer chromatography,15–17 direct spectrometric methods,18 gas chromatography (GC) fol-
46
lowing trimethylsilyl derivatization,19 and high-performance liquid chromatography–tandem mass spec-
47
trometry.20-23 However, no analytical methods have been described for the determination of grayanotox-
48
ins in dietary supplements and homemade wine.
49
Herein, we present the development and optimization of the first LC-QTOF-MS and triple quadrupole
50
LC–MS/MS method for the simultaneous determination of grayanotoxin I and grayanotoxin III in die-
51
tary supplements and homemade wine. The proposed method employs a solid-phase extraction (SPE)
52
step for sample cleanup, and therefore, it is expected to exhibit excellent sensitivity. We propose that
53
these chromatographic methods will allow accurate and reliable determination of such compounds, and
54
method was fully validated for grayanotoxin I and grayanotoxin III analysis in dietary supplements and
55
homemade wine for routine laboratory use.
56 57
Materials and Methods
58
Chemicals and reagents
59
Grayanotoxin I was kindly provided by the Kangwon National University (Kangwondo, Korea).
60
Grayanotoxin III was purchased from the National Development Institute of Korean Medicine
61
(NIKOM). The structures of grayanotoxins I and III are shown in Figure 1. Formic acid (analytical
62
grade) and acetic acid (analytical grade) were purchased from Sigma-Aldrich (St Louis, MO). Methanol
63
(HPLC grade) and acetonitrile (HPLC grade) were obtained from Merck (HPLC grade, Darmsradt,
64
Germany).
65 4 ACS Paragon Plus Environment
Page 5 of 22
Journal of Agricultural and Food Chemistry
66
Preparation of the standard solutions
67
Standard stock solutions were prepared by dissolving grayanotoxin I or grayanotoxin III (10 mg) in
68
methanol (10 mL). Working standard solutions were prepared daily by diluting the appropriate volumes
69
of the grayanotoxin I (10, 20, 40, 60, 80, and 100 ng/mL) and grayanotoxin III (20, 40, 80, 100, 200,
70
and 400 ng/mL) stock solutions in methanol.
71 72
Sample preparation
73
51 dietary supplement and homemade wine samples made from R. brachycarpum or R. mucronulatum
74
flowers were purchased from a local market and from four different provinces of Korea. Samples con-
75
sisted of both solids (two of the powder type, four of the capsule type, seven of the pill type, nine of the
76
tablet type, four of the tea type), and liquid (25 types) products. The contents from capsules, tablets,
77
pills, powder, and liquid products were vortex mixed briefly to obtain a uniform sample. For example,
78
for capsules, the ingredients present inside were used, and tablets and pills were broken into pieces and
79
ground to obtain a powder, so that all the samples were homogenized. To determine the grayanotoxin
80
contents present in the dietary supplements, the following sample preparation method was employed to
81
obtain analytical samples for injection. A portion of the liquid or solid sample (1 g) was dissolved in
82
methanol (10 or 50 mL, respectively). After mixing for 30 min under ultrasonication, the resulting solu-
83
tion was employed for the SPE cleanup step. Specifically, Oasis HLB LP cartridge, 6 cc, 500 mg (Wa-
84
ters, Milford, MA) were conditioned with methanol (5 mL) and distilled water (5 mL) prior to sample
85
loading. After loading a portion (2 mL) of the desired sample onto the SPE cartridge, elution of
86
grayanotoxins I and III was carried out using methanol (2 × 5 mL), and the eluted fractions were made
87
up to 20 mL in a volumetric flask using methanol. Finally, the sample extracts were filtered using a
88
0.2 µm membrane filter (VWR, Darmstadt, Germany) prior to analysis.
89 90
UPLC-QTOF-MS analysis 5 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 6 of 22
91
An ACQUITY ultra performance liquid chromatography (UPLC) system (Waters, Milford, MA) was
92
coupled with a QTOF Premier hybrid quadrupole orthogonal acceleration time-of-flight mass spectrom-
93
eter (Waters Micromass, Manchester, UK) equipped with an orthogonal Z-spray electrospray ionization
94
(ESI) interface operated in both positive and negative ion modes. LC separation was achieved using a
95
100 mm × 2.1 mm i.d., 1.7 µm, Waters ACQUITY BEH C18 column maintained at 40 °C. The mobile
96
phase was composed of both distilled water (solvent A) and methanol (solvent B) containing 0.1% for-
97
mic acid. The gradient elution profile employed as follows: 0–3 min 15% B; 3.0–3.1 min 15–90% B;
98
3.1–8.0 min 90% B; 8.0–8.1 min 90–15% B; and 8.1–10 min 15% B;. A mobile phase flow rate of
99
0.3 mL min−1 was employed along with a sample injection volume of 4 µL. Other acquisition conditions
100
were as follows: capillary voltage, 3.0 kV; desolvation temperature, 600 °C; desolvation gas flow, 800 L
101
h−1; cone gas flow, 30 L h−1. Lock-mass and mass calibration of the Q-TOF-MS were achieved using
102
leucine-enkephalin (m/z 556.2771 [M+H]+, m/z 554.2615 [M−H]−) and a sodium formate calibration
103
solution (0.1 M sodium hydroxide, 10% formic acid, and 80% acetonitrile). The mass resolution was in
104
the range of 10,000–20,000 at m/z 50–1000. Nitrogen gas was used as the collision gas in a hexapole
105
collision cell. Two types of acquisition, namely MS/MS and MSE, were performed. For the MS/MS ex-
106
periments, the cone voltage and collision energy ramp were optimized individually for each compound.
107
For the MSE experiments, two acquisition functions with different collision energies were created,
108
namely the low energy (LE) and high energy (HE) functions. The compound-dependent optimized cone
109
voltage and collision energy ramp determined for the MS/MS experiments were also employed.
110 111
LC-MS/MS analysis
112
Analytes were separated using an Agilent 1200 Series HPLC instrument (Agilent Technologies, Palo
113
Alto, CA) consisting of a quaternary pump, a vacuum degasser, and a thermostated autosampler. Sample
114
extracts (2 µL) were injected onto a 2.0 mm × 100 mm i.d., 3 µm, Capcell Pak C18 MGII column
115
(Shiseido, Tokyo, Japan). The mobile phase was composed of 0.1% formic acid in distilled water (sol6 ACS Paragon Plus Environment
Page 7 of 22
Journal of Agricultural and Food Chemistry
116
vent A) and 0.1% formic acid in methanol (solvent B). A flow rate of 0.25 mL min−1 was employed, and
117
the gradient elution profile was as follows: 0–4 min 5–40% B; 4–6.5 min 40–70% B; 6.5–7 min 70% B;
118
7–7.5 min 70–95% B; 7.5–8 min 95% B; 8–8.1 min 95–5% B; and 8.1–10 min 5% B;. The analytes
119
were detected using a QTrap5500 linear ion trap quadrupole mass spectrometer (AB SCIEX, Concord,
120
ON, Canada). The TurboIonSpray source was employed in the positive ion mode with a source tempera-
121
ture of 450 °C, an ion voltage of 5500 V, and a curtain gas pressure of 30 psi. To identify the target
122
compounds, the precursors and fragmentation patterns for each analyte were optimized through a series
123
of continuous infusion experiments. Initially, >2 multiple reaction monitoring (MRM) transitions were
124
monitored for each analyte, and two transitions were selected in the final method, one for quantitation
125
and the other for confirmation. To improve the sensitivity and selectivity of the system, the target com-
126
pounds were measured using the time-scheduled MRM mode. Analytical instrument control, data acqui-
127
sition, and data treatment were performed using the Analyst 1.5.1 software.
128 129
Results and Discussion
130
LC-QTOF-MS/MS analysis and proposed fragmentation patterns
131
To determine the sensitivity and selectivity of targets, standard solutions were infused directly into the
132
ESI source that is operated in the positive ion mode by varying the different MS parameters. MS condi-
133
tions were optimized and investigated for the quantitation of both grayanotoxin I and III, for which the
134
MS/MS product ion spectra are shown in Figure 2. In the case of grayanotoxin I, a prominent sodium
135
adduct of the molecular ion [M+Na] + was observed at m/z 435.22 in the Q1 full-scan mass spectrum,
136
although the most abundant signal was observed at m/z 375.21, which corresponded to
137
[M+Na−CH3COOH−H2O]+. More specifically, grayanotoxin I fragmented readily by loss of an acetyl
138
group and formed a more stable structure, which resembled grayanotoxin III (Figure 3A). We therefore
139
selected the m/z 435.22→375.21 ion transition for grayanotoxin I, although it should be noted that an
7 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 8 of 22
140
additional product ion at m/z 357.21 was also formed by the loss of another water molecule to afford
141
[M+Na−CH3COOH−2H2O]+. Similarly, the sodium adduct of the molecular ion [M+Na]+ was observed
142
at m/z 393.22 for grayanotoxin III in the Q1 full-scan mass spectrum. Successive loss of water produced
143
signals at m/z 375.21, 357.20, and 317.21, which corresponded to [M+H−H2O]+, [M+H−2H2O]+, and
144
[M+H−3H2O−rearrangement(C13-C16)]+, respectively (Figure 3B). It was assumed that the loss of H2O
145
renders the base ion more stable, with the number of H2O molecules lost being dependent on the number
146
of hydroxyl substituents present in the structure. Additional product ions at m/z 299.20, 289.17, and
147
246.16 corresponded to the stepwise loss of H2O, C2H4, and C2H3O from the m/z 357.20 ion.
148 149
LC-MS/MS Analysis
150
The initial stage of method development for LC-MS/MS analysis involved optimization of the mass
151
spectrometer conditions for the detection of the grayanotoxin I and grayanotoxin III. For this purpose,
152
the MS/MS parameters were optimized by individually infusing standard solutions of grayanotoxin.
153
Both positive (ESI+) and negative (ESI−) ionization modes were tested, although only the positive mode
154
gave satisfactory results. For each analyte (i.e., grayanotoxin I and grayanotoxin III), the two most in-
155
tense transitions were selected for quantitative determination of the components, and the remaining ions
156
were used for the confirmation of grayanotoxin I and grayanotoxin III. As the grayanotoxin III structure
157
fragmented readily to lose two molecules of H2O, this precursor ion was selected for detection by LC-
158
TQ/MS. Therefore, the precursor ion of m/z 335.0 was selected for the quantitation of grayanotoxin III,
159
as the optimized MS/MS patterns of the LC-MS/MS and LC-QTOF-MS techniques produced different
160
precursor and product ions. Under the optimized solvent gradient conditions, good separation of
161
grayanotoxin I and grayanotoxin III was achieved within 4 min.
162 163
Method Validation
164
Specificity 8 ACS Paragon Plus Environment
Page 9 of 22
Journal of Agricultural and Food Chemistry
165
Method selectivity was then assessed by duplicate analysis of blank samples from different dietary sup-
166
plements, and no peaks corresponding to interfering compounds were observed in the blank samples
167
close to the retention times of the target grayanotoxin I and grayanotoxin III. Typical chromatograms
168
obtained from the analysis of the blank and spiked blank samples are given in Figure 4.
169
The specificity was then determined using independent blank samples from the different analytes. Typi-
170
cal chromatograms of the blank samples spiked with 100 ng mL−1 of the grayanotoxin I and grayanotox-
171
in III are shown in Figure 4. More specifically, as shown in Figures 4B and D, no significant interfer-
172
ence was observed at the retention times of the analytes (i.e., 4.71 and 4.78 min for grayanotoxins I and
173
III, respectively). These results indicated that this method is both specific and selective, and therefore, it
174
is suitable for application in the analysis of dietary supplement samples.
175
Linearity and the limit of quantitation (LOQ)
176
Calibration curves for grayanotoxin I and III were constructed by plotting the peak area ratio (y) of each
177
compound against the nominal concentration (x) using weighted (1 x−1) least-squares linear regression
178
analysis. Linearity was determined over the concentration ranges of 10–100 ng mL−1 (grayanotoxin I)
179
and 20–400 ng mL−1 (grayanotoxin III), and the mean regression equations of the calibration curves
180
were derived as: y = 1040x − 368.68, r2 = 0.999 for grayanotoxin I, and y = 6663x − 27352, r2 = 0.999
181
for grayanotoxin III. In addition, the limits of quantitation (LOQs) for grayanotoxin I and grayanotoxin
182
III were 7.5 and 15 ng mL−1, respectively, which are sufficiently low to render this method suitable for
183
their determination in dietary supplements.
184
Precision and accuracy
185
The intra- and inter-day precision and accuracy values for grayanotoxin I and grayanotoxin III were de-
186
termined. These values were determined by analyzing quality control (QC) samples at the LOQ, and at
187
low, medium, and high concentrations in triplicate over one day (intra-day) and over three separate days
188
(inter-day). Thus, the intra- and inter-day precisions of these analytes were
