Anda di halaman 1dari 8

G Model

CHROMB-19774; No. of Pages 8 ARTICLE IN PRESS


Journal of Chromatography B, xxx (2015) xxxxxx

Contents lists available at ScienceDirect

Journal of Chromatography B
journal homepage: www.elsevier.com/locate/chromb

Simultaneous determination of capsaicin and dihydrocapsaicin for


vegetable oil adulteration by immunoafnity chromatography
cleanup coupled with LCMS/MS
Fei Ma a,b,c,d,e,,1 , Qingqing Yang a,b,c,d,1 , Bertrand Matthus f , Peiwu Li a,b,c,d,e, ,
Qi Zhang a,b,c,d, , Liangxiao Zhang a,b,c,e
a
Oil Crops Research Institute, Chinese Academy of Agricultural Sciences, Wuhan 430062, China
b
Key Laboratory of Biology and Genetic Improvement of Oil Crops, Ministry of Agriculture, Wuhan 430062, China
c
Key Laboratory of Detection for Mycotoxins, Ministry of Agriculture, Wuhan 430062, China
d
Laboratory of Quality & Safety Risk Assessment for Oilseeds Products (Wuhan), Ministry of Agriculture, Wuhan 430062, China
e
Quality Inspection & Test Center for Oilseeds Products, Ministry of Agriculture, Wuhan 430062, China
f
Federal Research Institute of Nutrition and Food, Max Rubner-Institut, Detmold 32756, Germany

a r t i c l e i n f o a b s t r a c t

Article history: Capsaicin and dihydrocapsaicin were selected as adulteration markers to authenticate vegetable
Received 30 July 2015 oils. In this study, a method of immunoafnity chromatography (IAC) combined with liquid
Received in revised form chromatographytandem mass spectrometry was established for the determination of capsaicin and
18 November 2015
dihydrocapsaicin in vegetable oils. In this method, immunosorbents were obtained by covalently coupling
Accepted 10 December 2015
highly specic capsaicinoid polyclonal antibodieswith CNBr-activated Sepharose 4B, and then packed into
Available online xxx
a polyethylene column. In this paper, the major parameters affecting IAC extraction efciency, including
loading, washing and eluting conditions, were also investigated. The IAC column displayed high selec-
Chemical compounds studied in this article:
Capsaicin (Pubchem CID: 1548943)
tivity for capsaicin and dihydrocapsaicin with the maximum capacity of 240 ng. The limit of detection
Dihydrocapsaicin (Pubchem CID: 107982) (LOD) and limit of quantication (LOQ) for capsaicin were calculated as 0.02 and 0.08 g kg1 , and for
Capsaicin-d3 (Pubchem CID: 57369257 dihydrocapsaicin were 0.03 and 0.10 g kg1 . The recoveries of capsaicin and dihydrocapsaicin in oil
Dihydrocapsaicin-d3 (Pubchem CID: samples were in the range of 87.395.2% with the relative standard deviation (RSD) of less than 6.1%. The
71316141) results indicated that capsaicinoid compounds could not be found in edible vegetable oils. Therefore, the
proposed method is simple, reliable and adequate for routine monitoring of capsaicinoid compounds in
Keywords:
vegetable oils and has an excellent potential for detection of adulteration with inedible waste oil.
Capsaicinoids
Immunoafnity chromatography 2015 Elsevier B.V. All rights reserved.
Isotope dilution internal standard
LCMS/MS
Inedible waste oil

1. Introduction inedible oils are generally contaminated by hazardous compounds


such as condiments, heavy metal ions, bio-toxins, polycyclic
Waste oil, consisting of cooking oil collected from food wastes, aromatic hydrocarbons (PAH), 3-monochloropropane-1,2-diol (3-
waste cooking oils recycled from drains or grease traps, residual MCPD) and its esters, phthalates, dioxins and polychlorinated
oils extracted from animal fats, deep frying oils, and other inedible biphenyls (PCBs) [29] or degradation products of oxidation pro-
oils, have caused serious food safety scandals in developing coun- cess. In the illicit rening industry, the complete removal of
tries when they are illegally sold and used as edible oils [1]. These contaminants or additives is difcult. Parts of them still remain
in recycled used oils though they are treated by degumminrg,
neutralizing, washing, drying, bleaching, ltering and deodorizing
processes [10]. As a result, those oils can adversely impact human
Corresponding authors at: Oil Crops Research Institute, Chinese Academy of
health due to oxidation products and other hazardous materials
Agricultural Sciences, Wuhan 430062, China. Fax: +86 27 86812862.
E-mail addresses: mafeicpu@163.com (F. Ma), peiwuli@oilcrops.cn (P. Li), from the draining and reprocessing processes [11].
Zhangqi01@caas.cn (Q. Zhang).
1
These authors contributed equally to this study.

http://dx.doi.org/10.1016/j.jchromb.2015.12.017
1570-0232/ 2015 Elsevier B.V. All rights reserved.

Please cite this article in press as: F. Ma, et al., Simultaneous determination of capsaicin and dihydrocapsaicin for
vegetable oil adulteration by immunoafnity chromatography cleanup coupled with LCMS/MS, J. Chromatogr. B (2015),
http://dx.doi.org/10.1016/j.jchromb.2015.12.017
G Model
CHROMB-19774; No. of Pages 8 ARTICLE IN PRESS
2 F. Ma et al. / J. Chromatogr. B xxx (2015) xxxxxx

Generally, sensory evaluation was used to detect adulteration (Shanghai, China). Sodium acetate (NaAc), sodium chloride (NaCl),
of inedible waste oils. However, sensory analysis depends on the sodium dihydrogen phosphate, disodium hydrogen phosphate,
experience of analysts, and the subjective judgment may cause false glacial acetic acid, tris(hydroxymethyl) aminomethane and sodium
negative results. In this case, many analytical methods were devel- azide (NaN3 ) were from Sinopharm Chemical Reagent (Shanghai,
oped to detect and quantify the adulterants. As rapid analytical China). Hydrochloric acid (HCl), caprylic acid, hexane, ammonium
methods, NMR, NIR and Raman could characterize oil samples in the sulfate, sodium bicarbonate and potassium dihydrogen phos-
whole spectrum [1214]. Recently, the metabolite proles of fatty phate were obtained from Kermel Chemicals (Tianjin, China). All
acids, triacylglycerols and volatile components were used to ana- other chemicals and organic solvents were of analytical reagent
lyze inedible oils and detect potential adulteration [1517]. During grade. Ultra-pure water (18 m) was obtained from a Milli-Q
data processing, the optimized model needs many standards and water purication system (Milford, MA, USA). Nonspecic rabbit
possible adulterated samples and might not be effective for samples immunoglobulins were produced in our laboratory.
out of the training set. Compared with the preceding methods, liq- The stock solutions of capsaicin, dihydrocapsaicin, capsaicin-d3
uid chromatographytandem mass spectrometry (LCMS/MS) is a and dihydrocapsaicin-d3 were separately prepared by accurately
promising approach to assess adulteration and quality of vegetable weighing 1 mg of the reference materials and dissolving them in
oils since it satises routine analysis requirements for rapid deter- 20 mL methanol. A series of standard solutions were prepared by
mination and identication of contaminants or additives, which can diluting the stock solutions with methanol. All standard solutions
be used as marker compounds [18,19]. were sealed with paralm, covered with aluminum foil, and stored
With the sensory attributes of pungency, aroma and color, in the dark at 4 C until use.
hot peppers are popular food additives and have been consumed
in large quantities throughout China and the world. The most 2.2. Instrumentation
abundant capsaicinoids are capsaicin and dihydrocapsaicin (the
molecular structures of these chemicals are present in Fig. 1), which A Thermo lyophilizer (Savant, England) was used to freeze
constitute nearly 90% of the total pungency of pepper fruits [20]. dry antibodies. A shaker (Chinese Academy of Sciences Scientic
Previous researches indicated that capsaicin and dihydrocapsaicin Instrument, Wuhan) was used to prepare the immunosorbents.
were lipophilic and stable during the rening process of inedible Centrifugation of the vegetable samples was performed on a
waste oils with high boiling points. Therefore these compounds centrifuge (Hitachi CF 16RX), and a homogenizer (IKA Labora-
could be selected as specic biomarkers to detect and authenticate tory Equipment, Germany) was used for sample preparation.
edible vegetable oils [21]. Sample analysis was performed in selected reaction monitoring
Considering the complexity of lipid matrix and low concentra- (SRM) mode on the Accela HPLC system coupled to TSQ Quan-
tion of capsaicinoids in inedible oils, the analytical methods usually tum Ultra EMR (Thermo Fisher Scientic, USA). Separation was
use thin layer chromatography (TLC) or solid phase extraction performed at 35 C on a Thermo Scientic C18 column (Hyper-
(SPE) to extract and enrich the analytes [2224]. The procedures sil Gold, 100 mm 2.1 mm, 3.0 m). Mobile phase A consisted of
generally consist of liquidliquid extraction, saponication and 0.1% aqueous formic acid and mobile phase B acetonitrile. A lin-
neutralization, which provide non-specic retention and are time- ear gradient elution program was applied as follows: initial B was
consuming, laborious, low sensitivity and consuming large volumes linearly increased from 33.5% to 83.5% in 10 min, held for 1 min,
of organic solvents. Immunoafnity chromatography (IAC) is a sep- and returned to 33.5% B in 3 min, which was held for 4 additional
aration method which takes advantage of the specic interaction minutes for re-equilibration, giving a total run time of 18 min. The
between antibodies and antigens. It is a simple and selective tech- mobile phase ow rate was 250 L min1 , and an aliquot of 10 L
nique to purify analytes without tedious pretreatment and can save sample was injected into the HPLC system. The mass spectra were
organic solvents during pretreatment [25,26]. Many reports have obtained by Thermo TSQ Quantum Ultra EMR triple quadrupole
been found for the determination of toxins, pesticide residues, vet- (Thermo Scientic, USA) coupled with electrospray interface (ESI).
erinary drugs and illegal additives using IAC cleanup [2729]. To the The MS/MS conditions were set as follows: spray voltage, 3.5 kV
best of our knowledge, IAC cleanup has not been used to determine in positive mode; capillary temperature, 350 C; sheath gas pres-
capsaicin and dihydrocapsaicin in combination with LCMS/MS. sure (N2 ), 30 units; auxiliary gas pressure (N2 ), 5 units; collision
The aim of this study was to establish IAC using poly- gas (Ar), 1.5 mTorr; scan time, 0.1s. As shown in Table 1, two
clonal antibodies (pAb) covalently immobilized on CNBr-activated parent-to-product transitions for each analyte were simultane-
Sepharose-4B for simultaneous determination of capsaicin and ously monitored in SRM mode.
dihydrocapsaicin in vegetable oils, which were further quantied All the experiments were conducted in triplicate, and the aver-
by LCMS/MS. The extraction conditions of the IAC column for age values standard deviations (SD) were reported. LCMS/MS
capsaicinoid compounds were optimized and the IAC column was data were processed using Xcalibur 2.0.7 SP1. Statistical analyses
characterized in terms of binding capacity, extraction recovery and were performed with the @Risk 5.5.1 software package (Palisade,
reproducibility. Under the optimized conditions, the analytes in Australia).
lipid matrix were successfully quantied and used to detect oils
adulterated with inedible waste oil. 2.3. Production and purication of polyclonal antibodies

Polyclonal antibodies (pAb) against capsaicinoids were pro-


2. Experimental duced according to the procedure described in our previous study
[30]. As illustrated in Fig. 2, the hydroxyl group of capsaicin
2.1. Reagents and materials molecules was changed to the ester group under bromination reac-
tion, and then the esters were alkaline hydrolyzed to the carboxyl
CNBr-activated Sepharose 4B was purchased from GE Health- group which was covalently coupled to the carrier protein (bovine
care (Uppsala, Sweden). Capsaicin and dihydrocapsaicin standards serum albumin, BSA) using the modied active ester method.
were obtained from SigmaAldrich (St. Louis, MO, USA). Capsaicin- The capsaicin-BSA conjugate was dialyzed and then used as an
d3 and dihydrocapsaicin-d3 standards were obtained from Toronto immunogen to immunize New Zealand white rabbits for polyclonal
Research Chemicals (North York, Canada). Formic acid, methanol antibody production. The obtained antiserum was further puried
(MeOH) and acetonitrile of HPLC grade were from SigmaAldrich according to the modied saturated ammonium sulfate caprylic

Please cite this article in press as: F. Ma, et al., Simultaneous determination of capsaicin and dihydrocapsaicin for
vegetable oil adulteration by immunoafnity chromatography cleanup coupled with LCMS/MS, J. Chromatogr. B (2015),
http://dx.doi.org/10.1016/j.jchromb.2015.12.017
G Model
CHROMB-19774; No. of Pages 8 ARTICLE IN PRESS
F. Ma et al. / J. Chromatogr. B xxx (2015) xxxxxx 3

Fig. 1. Molecular structures of capsaicin and dihydrocapsaicin and their stable isotope labeled internal standards selected as the biomarkers for oil detection and authenti-
cation.

Table 1
LCMS/MS parameters for the determination of capsaicinoid compounds.

Compound Parent ion (m/z) Product ion (m/z) Collision energy (eV) Tube lens voltage (V)

Capsaicin 306.1 136.9a 18 126


306.1 122.0 18 126

Dihydrocapsaicin 308.2 136.9a 18 136


308.2 122.0 36 136

Capsaicin-D3 309.1 140.2a 25 126


309.1 182.3 15 125

Dihydrocapsaicin-D3 311.1 140.2a 15 136


311.1 184.3 15 136
a
The most abundant product ion selected for the quantitative analysis.

Fig. 2. Synthesis scheme of the capsaicinoid-protein conjugate.

acidammonium sulfate method [25]. The antiserum was diluted dried and stored at 20 C until use. The cross-reactivity (CR) was
with an equal volume of saline, and then the saturated ammonium calculated as the following equation,
sulfate solution (pH 7.4) was gently added under magnetic stirring
at 4 C for 2 h. After the turbid solution centrifuged at 12,000 rpm
IC50 of capsaicin
for 30 min, the precipitate was dissolved in 1/8 antiserum volume CR% = 100% (1)
IC50 of analytes
of phosphate buffer solution (PBS) (0.01 M, pH 7.4) and dialyzed
against PBS for 48 h. The puried antibody solution was then freeze-
where IC50 is the concentration at which 50% of the antibody is
bound to the analytes and CR were calculated for capsaicin and
dihydrocapsaicin, respectively.

Please cite this article in press as: F. Ma, et al., Simultaneous determination of capsaicin and dihydrocapsaicin for
vegetable oil adulteration by immunoafnity chromatography cleanup coupled with LCMS/MS, J. Chromatogr. B (2015),
http://dx.doi.org/10.1016/j.jchromb.2015.12.017
G Model
CHROMB-19774; No. of Pages 8 ARTICLE IN PRESS
4 F. Ma et al. / J. Chromatogr. B xxx (2015) xxxxxx

2.3.1. Preparation of the IAC column for establishing a single IAC column which could simultaneously
Immunosorbents were prepared following the manufacturers capture capsaicin and dihydrocapsaicin.
instructions and related literature [25]. Briey, 0.3 g CNBr-activated
Sepharose 4B was swelled and washed with 40 mL 1 mM HCl for
15 min, and then mixed with 4 mL coupling buffer (0.1 M NaHCO3 , 3.2. Preparation of the IAC column
0.5 M NaCl, pH 8.3) containing 10.2 mg puried antibody. The cou-
pling solution reacted for 3 h in a shaker at room temperature, and The absorbents supporting the IAC column have a signicant
then it was transferred to the column and the unbound antibody role in the covalent reaction with the antibody and purica-
was washed with the coupling buffer of at least 5 times the medium tion of the sample matrix. The basic properties of the supporting
volume. The active gel groups were capped with the blocking buffer absorbents are biochemical inertness, mechanical and chemical
(0.1 M TrisHCl, pH 8.0) at 4 C for 2 h without shaking. To remove stability, porosity and uniformity in particle size, and covalent
excess uncoupled immunosorbents, the gel was washed with high reaction with the antibody under mild conditions. Compared
and low pH buffer solutions (0.1 M HAc-NaAc buffer, pH 4.0, con- with the supporting absorbents for immobilization procedures,
taining 0.5 M NaCl, and 0.1 M TrisHCl buffer, pH 8.0, containing agarose-based supports are the most common absorbents used
0.5 M NaCl) of at least 5 times the medium volume for three cycles. for preparation of the IAC column, especially for off-line purica-
Finally, the gel was equilibrated with 0.01 M PBS and stored in tion [25,31]. CNBr-activated Sepharose 4B, which could be easily
0.01 M PBS containing 0.01% NaN3 at 4 C before use. coupled to the antibodies with the activated group of the parti-
cle surface, is chemically and mechanically inert, mild derivative,
water stable and hydrophilic and retains the bio-specic activity of
2.4. Sample preparation the antibody. In the present study, it was selected as the supporting
absorbents for IAC extraction. Sepharose 4B of (0.3 g) was dissolved
For the development and validation of the method, at least 1.0 kg in 4 mL gel, and then 10.2 mg of pAb was covalently reacted with
vegetable oil samples, including maize oil (5), sesame oil (5), sun- the gel. From the data of UVvis spectrometry, 8.8 mg of pAb was
ower oil (5), soybean oil (5), olive oil (7), rapeseed oil (7) and immobilized on the gel. According to the ratio of the antibody cou-
blend oil (5) were collected from supermarkets. Waste oil sam- pled with the solid support gel to the initial antibody, the coupling
ples (9) were collected from local restaurant grease traps. The efciency of the capsaicinoid antibody on the immunosorbents was
detailed information of samples including type, place of produc- 86.3%. Finally, 0.2 mL of the immunosorbents were transferred into
tion, rening step were added in Supplementary material Table 1S. a 1 mL polyethylene column, and stored in 0.01 M PBS containing
All oil samples were stored at 25 C before use. Each oil sample 0.01% NaN3 at 4 C.
was homogenized by thoroughly stirring with a mechanical mor-
tar and then stored in the dark at 4 C until analysis. The oil samples,
in which no detectable amount of capsaicinoids was conrmed by 3.3. Optimization of IAC conditions
LCMS/MS, were used as the blank sample (soybean oil). Then, the
blank samples spiked with standard solutions of capsaicin, dihydro- The afnity interaction between antigens and antibodies is
capsaicin, capsaicin-d3 and dihydrocapsaicin-d3 prior to extraction generally thought to be a comprehensive combination of various
were used as inedible waste oils. non-covalent bonding mechanisms including electrostatic attrac-
5.0 g oil sample spiked with 200 L of internal standards tion force, van der Waals attraction force, hydrogen bonding, and
(capsaicin-d3 and dihydrocapsaicin-d3 ) was placed in a 25 mL hydrophobicity. The characteristic properties of the loading, wash-
polypropylene-capped centrifuge tube. Subsequently, it was mixed ing, and eluting solutions, such as polarity, pH value and ionic
with 5 mL methanol and vortexed for 2 min before being cen- strength, have important effects on the afnity interaction of IAC.
trifuged at 4500 rpm for 5 min at 4 C. The upper layer was To obtain high extraction efciency and avoid irreversibly change
transferred and ltered through an organic membrane (0.45 m). of antibody afnities, the properties of the loading, washing and
elution conditions were investigated and optimized according to
the previous study [25]. All fractions in the loading, washing and
2.5. IAC extraction eluting steps were collected respectively, and the analytes were
determined by LCMS/MS.
For IAC cleanup, 0.5 mL of the organic extracts was diluted to
5 mL with PBS (pH 7.4) and then passed through the IAC column
with a ow rate of 1 mL min1 to remove the interference resulted 3.3.1. Loading condition
from the lipid matrix. The column was washed with water (10 mL) To achieve satisfactory loading procedure, different percent-
and eluted with 1 mL of methanol. The eluting solvent was evapo- ages of MeOH-PBS (pH 7.4) mixture (0:100, 5:95, 10:90, 15:85,
rated to dryness under a gentle stream of nitrogen at 40 C. Finally, 20:80, 30:70, v/v) was investigated and compared. As shown in
the residue was reconstituted with 200 L of acetonitrile/H2 O Fig. 3a, when the concentration of MeOH was less than 5%, the
(50:50, v/v) and ltered through a PTEE lter (0.22 m) before recovery increased with the increasing MeOH concentration. When
LCMS/MS. the MeOH concentration was further increased from 5%, the recov-
ery decreased to less than 78.7%. The recoveries were affected by
the solubility of capsaicinoids in organic solvents. The results indi-
3. Results and discussion cated that the percentage of organic solvents was important for
the equilibrium extraction process of analyte dissolution and anti-
3.1. Antibody characterization body desaturation. Therefore, pH 7.4 PBS containing 5% MeOH was
selected as the loading solution.
With the purpose of IAC cleanup considered, the characteriza- The volumes of the loading medium (1, 2, 5, 8, 10, and 15 mL)
tion of the antibody plays an important role in the preparation inuencing the recoveries were also studied (Fig. 3b). The results
of the IAC column for a single analyte or for an analyte and its showed that with the increase of the loading solvent volume from
derivatives. The antibody against capsaicin showed a high titer 1 mL to 5 mL, the extraction recoveries of capsaicin and dihydro-
(1/128,000), showing a cross-reactivity of 95% and 88% for capsaicin capsaicin gradually increased from 63.2% to 94.4% and from 66.8%
and dihydrocapsaicin, respectively. Thus, the antibody was suitable to 96.2%, respectively. Further increase of the loading solvent did

Please cite this article in press as: F. Ma, et al., Simultaneous determination of capsaicin and dihydrocapsaicin for
vegetable oil adulteration by immunoafnity chromatography cleanup coupled with LCMS/MS, J. Chromatogr. B (2015),
http://dx.doi.org/10.1016/j.jchromb.2015.12.017
G Model
CHROMB-19774; No. of Pages 8 ARTICLE IN PRESS
F. Ma et al. / J. Chromatogr. B xxx (2015) xxxxxx 5

Fig. 3. The inuence of several parameters on IAC extraction efciency: (a) loading solvent, (b) loading volume, (c) washing solvent, and (d) eluting solvent. The oil sample
solutions with capsaicin and dihydrocapsaicin spiked at 2.0 g kg1 .

Fig. 4. The amount of capsaicin and dihydrocapsaicin retained in the IAC column
with continuous sample loading.

not signicantly improve the recoveries, thus 5 mL was chosen as


the loading volume in the following experiments.

3.3.2. Washing condition


To avoid matrix effects and poor chromatographic separation,
the oil- and fat-based interfering substances which were retained at
the sol-gel column should be removed as much as possible. The fol-
Fig. 5. LCMS/MS chromatograms of the blank vegetable oils with capsaicin and
lowing washing media were tested: PBST (Tween 20/PBS, 0.5:99.5,
dihydrocapsaicin spiked at 0.5 g kg1 , respectively.
v/v), MeOHwater (10:90, v/v), PBS, and water. In Fig. 3c, the exper-
imental results indicate no signicant differences between these
washing mediums except PBST. When nonionic detergent-Tween 3.3.3. Eluting condition
20 was added, the nonspecic hydrophobic interactions between To dissociate the analytes from the antibodyantigen complex,
the detergent and the lipophilic capsaicinoid compounds in the elution solutions were selected as organic aqueous solvents (MeOH
aqueous system resulted in low recoveries. Consequently, 10 mL was widely applied), which were successfully applied in most
water was adopted to minimize nonspecic binding and improve off-line IAC procedures [2529]. In this study, different percent-
sensitivity of LCMS/MS analysis. ages of waterMeOH (20:80, 15:85, 10:90, 5:95, 0:100, v/v) were

Please cite this article in press as: F. Ma, et al., Simultaneous determination of capsaicin and dihydrocapsaicin for
vegetable oil adulteration by immunoafnity chromatography cleanup coupled with LCMS/MS, J. Chromatogr. B (2015),
http://dx.doi.org/10.1016/j.jchromb.2015.12.017
G Model
CHROMB-19774; No. of Pages 8 ARTICLE IN PRESS
6 F. Ma et al. / J. Chromatogr. B xxx (2015) xxxxxx

Table 2
Recoveries and precisions for the determination of capsaicinoid compounds in oil samples.

Analyte Accuracy, mean recoverya (%), n = 3 Intra-day (RSD%, n = 6) Inter-day (RSD%, n = 3)


1
1 g kg1 5 g kg1 10 g kg1 1 g kg1 5 g kg1 10 g kg1 1 g kg1 5 g kg1 10 g kg

Capsaicin 87.3 94.1 95.2 3.5 4.1 3.8 3.0 2.7 4.2
Dihydrocapsaicin 90.4 94.2 94.3 2.6 4.3 4.2 6.1 4.6 4.1
a
Recoveries are calculated as mean value of triplicate analysis.

Table 3 3.4. Method validation


Results for the determination of capsaicin and dihydrocapsaicin in oil samples.

Sample Number of samples Capsaicin a Dihydrocapsaicin a 3.4.1. Linearity


(g kg1 ) (g kg1 ) Calibration curves were constructed using the ratio of the
Waste oil 1 9 8.4 0.5 3.4 0.3 chromatographic peaks of the analyte to its internal stan-
Waste oil 2 4.4 0.3 11.4 1.1 dard, at different levels in the range of 0.2050.0 g kg1 . To
Waste oil 3 9.7 0.7 26.7 1.9 calibrate the effect of ion suppression or enhancement, capsaicin-
Waste oil 4 2.8 0.1 9.5 0.7
d3 and dihydrocapsaicin-d3 were used as internal standards
Waste oil 5 19.5 0.7 10.8 0.9
Waste oil 6 26.7 1.9 9.8 0.7 for capsaicin and dihydrocapsaicin, respectively. The linearity
Waste oil 7 17.5 4.9 8.6 0.2 of the afore-mentioned ranges was established by analyzing
Waste oil 8 41.2 6.4 19.9 2.8 seven calibration levels (0.20 g kg1 , 0.50 g kg1 , 1.00 g kg1 ,
Waste oil 9 45.2 2.1 16.8 0.9
5.00 g kg1 , 10.00 g kg1 , 20.00 g kg1 , and 50.00 g kg1 ) in
Maize oil 5 N.D.b N.D.
Sesame oil 5 N.D. N.D. three replicates. Good linearity was obtained for each analyte
Sunower oil 5 N.D. N.D. throughout the concentration range, and the calibration equations
Soybean oil 5 N.D. N.D. were calculated as: y = 0.28435 0.00113 with the correlation
Olive oil 7 N.D. N.D. coefcient (r2 ) of 0.9998 for capsaicin and y = 0.21286 0.01895
Rapeseed oil 7 N.D. N.D.
with the correlation coefcient (r2 ) of 0.9992 for dihydrocapsaicin,
Blend oil 5 N.D. N.D.
respectively.
a
Values represent the mean of the triplicate analyses the standard deviation.
b
N.D., not detected.
3.4.2. Limit of detection (LOD) and limit of quantication (LOQ)
Blank samples were selected and spiked with mixed-standard
solutions, and then prepared and analyzed under the optimized
conditions. The LOD was dened as the lowest detectable concen-
evaluated. As shown in Fig. 3d, the best recovery was obtained tration with a signal-to-noise ratio of at least 3, whereas the LOQ
when using 1 mL MeOH. Elution media with lower percentage of was dened as the lowest quantiable concentration with a signal-
methanol may cause less damage to the antibodies and can increase to-noise ratio of at least 10. The LODs and LOQs were calculated to
the reusability of the IAC column. During the evaporation step, be 0.02 and 0.08 g kg1 for capsaicin and 0.03 and 0.10 g kg1
water contained in the eluate makes the cleanup procedure more for dihydrocapsaicin. The results indicated that this method was
time-consuming. sensitive to the capsaicinoid compound in vegetable oils. In Fig. 5,
the LCMS/MS chromatograms were present for the vegetable oil
spiked with 0.5 g kg1 of capsaicin and 0.5 g kg1 of dihydrocap-
3.3.4. Reconstitution medium saicin.
A tailing peak was obtained when MeOH was used as the
reconstitution medium because the analytes have an ionic char- 3.4.3. Accuracy and precision
acter. The reconstitution medium including different percentages The accuracy of the proposed method was studied in terms
of MeOH/H2 O, acetonitrile/H2 O, acetonitrile/PBS (50:50, 40:60, of trueness (systematic error) and expressed as the mean recov-
30:70, 20:80, 10:90, 5:95, v/v) and ACN were evaluated. The results ery. Firstly, blank oil samples were spiked with capsaicin and
showed that acetonitrile/PBS (50:50, v/v) and acetonitrile/H2 O dihydrocapsaicin at different concentrations ranging from 1.0 to
(50:50, v/v) gave larger MS/MS signal than other solutions including 10.0 g kg1 , and the analyte concentrations in crude oil sam-
aqueous MeOH. The ionization suppression of the analytes could ples were calculated based on the calibration curves. Then, the
been caused by the ions from PBS during the electrospray process, oil samples were spiked at different concentrations, and the con-
which would worsen the sensitivity of the MS/MS and contaminate centrations were also calculated in the method mentioned above.
ion source. Thus acetonitrile/H2 O (50:50, v/v) was optimized as the Finally, the recoveries were obtained by comparing the calculated
reconstitution medium in the following experiments. spiked concentration with the corresponding spiked values. In
Table 2, the results indicated that the recoveries of the samples
spiked with different concentrations of capsaicin and dihydro-
capsaicin ranged from 87.3% to 95.2% and from 90.4% to 94.3%,
3.3.5. Binding capacity respectively.
Under the optimal IAC conditions for capsaicinoids described The reproducibility of the method was determined by the intra-
above, the column capacity was evaluated by loading 30 mL and inter-day precision detection of blank vegetable oils spiked
standard solution mixed with capsaicin and dihydrocapsaicin with capsaicin and dihydrocapsaicin at three different concentra-
(10 ng mL1 ). As shown in Fig. 4, the results indicated that the max- tions. The intra-day precisions were determined by six parallel
imum binding capacity of IAC was 240 ng for the total amount of extractions of the sample solution at different concentration levels
capsaicin and dihydrocapsaicin. After the loading volume exceeded within one day, and the inter-day precisions were determined by
the maximum capacity, the amount of analytes remaining on the three parallel extractions of the sample solution at different con-
IAC column attened out, which illustrated the saturation of cap- centration levels for three consecutive days. The results showed
saicinoid binding sites. that the intra- and inter-day RSDs were less than 4.3% and 6.1%,

Please cite this article in press as: F. Ma, et al., Simultaneous determination of capsaicin and dihydrocapsaicin for
vegetable oil adulteration by immunoafnity chromatography cleanup coupled with LCMS/MS, J. Chromatogr. B (2015),
http://dx.doi.org/10.1016/j.jchromb.2015.12.017
G Model
CHROMB-19774; No. of Pages 8 ARTICLE IN PRESS
F. Ma et al. / J. Chromatogr. B xxx (2015) xxxxxx 7

respectively (Table 2), indicating that satisfactory reproducibility Appendix A. Supplementary data
was achieved by the method.
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.jchromb.2015.
3.4.4. Matrix effect
12.017.
The matrix of the complex samples, as well as other variables
such as sample preparation, chromatography and mass spectrom-
etry, could diminish or enhance the ion intensity of the analyte References
and affect reproducibility and accuracy of the analytic method. In
[1] F. Lu, X. Wu, China food safety hits the gutter, Food Control 41 (2014)
addition to optimizing sample preparation and purication, the co- 134138.
eluting stable isotope labeled internal standard (SIL-IS) plays an [2] M. Abalos, J. Parera, E. Abad, J. Rivera, PCDD/Fs and DL-PCBs in feeding fats
important role in solving this problem [32]. When added to the obtained as co-products or by-products derived from the food chain,
Chemosphere 71 (2008) 11151126.
sample at the beginning of the extraction, the SIL-IS experienced all [3] P. Felizardo, M.J. Neiva Correia, I. Raposo, J.F. Mendes, R. Berkemeier, J.M.
changes experienced by the analyte, including the ion suppression Bordado, Production of biodiesel from waste frying oils, Waste Manage. 26
effects affecting the ESI response. It was found that when SIL-ISs (2006) 487494.
[4] G. Hageman, R. Kikken, F. Ten Hoor, J. Kleinjans, Assessment of mutagenic
were used for analysis, the precision of the slopes of the calibration activity of repeatedly used deep-frying fats, Mutat. Res. 204 (1988) 593604.
standards prepared in ve different lots of vegetable oils was within [5] C.J. Jacob, C. Lok, K. Morley, D.A. Powell, Government management of two
the range 1.33.5%, irrespective of the HPLC-MS interface utilized. media-facilitated crises involving dioxin contamination of food, Public
Underst. Sci. 20 (2011) 261269.
The results indicated that the selectivity of the immunoafnity [6] Y. Guo, Z. Zhang, L. Liu, Y. Li, N. Ren, K. Kannan, Occurrence and proles of
chromatography method was acceptable for routine analysis of the phthalates in foodstuffs from China and their implications for human
capsaicinoid compounds in vegetable oils. exposure, J. Agric. Food Chem. 60 (2012) 69136919.
[7] R. Weihaar, Fatty acid esters of 3-MCPD: overview of occurrence and
exposure estimates, Eur. J. Lipid Sci. Tech. 113 (2011) 304308.
3.5. Application immunoafnity chromatography cleanup [8] X.F. Leong, A. Aishah, U. Nor Aini, S. Das, K. Jaarin, Heated palm oil causes rise
in blood pressure and cardiac changes in heart muscle in experimental rats,
coupled with LCMS/MS for oil adulteration Arch. Med. Res. 39 (2008) 567572.
[9] M. Grootveld, C.J.L. Silwood, P. Addis, A. Claxson, B.B. Serra, M. Viana, Health
Under the optimized conditions, the proposed method was effects of oxidized heated oil, Foodservice Res. Int. 13 (2001) 4155.
[10] W.M. Cao, X.H. Sun, F.X. Chen, B. Xue, W. Chen, Q. Jin, X. Wang, Prospect on
applied to determination of the capsaicinoid compounds in oil sam- discerning technology of swill-cooked dirty oil, China Oils Fats 37 (2012) 15.
ples including inedible waste oil, maize oil, sesame oil, sunower [11] H. Esterbauer, Cytotoxicity and genotoxicity of lipid-oxidation products, Am.
oil, soybean oil, olive oil, rapeseed oil, and blend oil. As present in J. Clini. Nutr. 57 (1993) 779S785S.
[12] M.T. Osorio, S.A. Haughey, C.T. Elliott, A. Koidis, Evaluation of methodologies
Table 3, capsaicinoids compounds were found in waste oils, and to determine vegetable oil species present in oil mixtures: proposition of an
the overall ranges of analytes were 2.845.2 g kg1 for capsaicin approach to meet the EU legislation demands for correct vegetable oils
and 3.426.7 g kg1 for dihydrocapsaicin. The results also showed labelling, Food Res. Int. 60 (2014) 6675.
[13] E. Cubero-Leon, R. Penalver, A. Maquet, Review on metabolomics for food
that no capsaicinoid compounds were detected in the vegetable oils authentication, Food Res. Int. 60 (2014) 95107.
from the market. As expected, capsaicin and dihydrocapsaicin were [14] C.A. Nunes, Vibrational spectroscopy and chemometrics to assess
only found in waste oil samples, which could act as the putative authenticity, adulteration and intrinsic quality parameters of edible oils and
fats, Food Res. Int. 60 (2014) 255261.
biomarkers to detect and authenticate edible vegetable oils.
[15] L. Zhang, P. Li, X. Sun, W. Hu, X. Wang, Q. Zhang, X. Ding, Untargeted fatty acid
proles based on the selected ion monitoring mode, Anal. Chim. Acta 839
(2014) 4450.
4. Conclusions [16] R. Salghi, W. Armbruster, W. Schwack, Detection of argan oil adulteration
with vegetable oils by high-performance liquid chromatography-evaporative
To the best of our knowledge, this is the rst time the anti- light scattering detection, Food Chem. 153 (2014) 387392.

[17] S. Mildner-Szkudlarz, H.H. Jelen, R. Zawirska-Wojtasiak, E. Wasowicz,
body was generated against capsaicinoid compounds and applied Application of headspacesolid phase microextraction and multivariate
to IAC cleanup. In this study, a simple, sensitive and accurate IAC- analysis for plant oils differentiation, Food Chem. 83 (2003) 515522.
LCMS/MS method was established for determination of capsaicin [18] D.S. Wishart, Metabolomics: applications to food science and nutrition
research, Trends Food Sci. Technol. 19 (2008) 482493.
and dihydrocapsaicin in vegetable oil samples. By covalently cou- [19] F. Puiggros, R. Sola, C. Blade, M.-J. Salvado, L. Arola, Nutritional biomarkers
pling the highly specic capsaicinoid pAb with CNBr-activated and foodomic methodologies for qualitative and quantitative analysis of
Sepharose 4B and packing the immunosorbents into a polyethylene bioactive ingredients in dietary intervention studies, J. Chromatogr. A 1218
(2011) 73997414.
column, the IAC column was utilized for extraction and purica- [20] S.F. Kosuge, M. Furata, Studies on the pungent principle of capsicum. Part XIV:
tion. Satisfactory results were obtained with regard to sensitivity, chemical constitution of the pungent principle, Agric. Biol. Chem. 34 (1970)
accuracy and precision, and the method was effective in remov- 248256.
[21] L. Ang, J. Jin, S. Wang, X. Wang, Y. Tian, J. Chen, A novel method for the
ing potential interference from vegetable oils since no interference
identication of illegal cooking oil (1): detection of three capsaicinoids with
was observed for quantication of capsaicinoid compounds. The liquid chromatographymass spectrometry, Se Pu 30 (2012) 10941099.
results also showed that no capsaicinoid compounds were detected [22] G.F. Barbero, A. Liazid, M. Palma, C.G. Barroso, Fast determination of
capsaicinoids from peppers by high-performance liquid chromatography
in the vegetable oils from market, and capsaicin and dihydro-
using a reversed phase monolithic column, Food Chem. 107 (2008)
capsaicin were only found in inedible waste oils. Therefore, the 12761282.
IAC-LCMS/MS method was adequate for routine monitoring of [23] A. Pena-Alvarez, E. Ramirez-Maya, L.A. Alvarado-Suarez, Analysis of capsaicin
capsaicinoid compounds in vegetable oil samples, and provided a and dihydrocapsaicin in peppers and pepper sauces by solid phase
microextractiongas chromatographymass spectrometry, J. Chromatogr. A
practical tool for detection of adulteration with inedible waste oil. 1216 (2009) 28432847.
[24] G.F. Barbero, A. Liazid, M. Palma, C.G. Barroso, Ultrasound-assisted extraction
of capsaicinoids from peppers, Talanta 75 (2008) 13321337.
Acknowledgements [25] Y.R. Wang, Q. Zhang, P.W. Li, W. Zhang, Y. Li, X.X. Ding, Selective sample
cleanup by immunoafnity chromatography for determination of fenvalerate,
J. Chromatogr. B 879 (2011) 35313537.
This work was supported by the Project of National Science & [26] F. Ma, R. Chen, P. Li, Q. Zhang, W. Zhang, X. Hu, Preparation of an
Technology Pillar Plan (2012BAK08B03), the National Natural Sci- immunoafnity column with amino-silica gel microparticles and its
ence Foundation of China (31201447), Deutscher Akademischer application in sample cleanup for aatoxin detection in agri-products,
Molecules 18 (2013) 22222235.
Austausch Dienst (DAAD) and the National Major Project for Agro-
[27] F.A. Esteve-Turrillas, J.V. Mercader, C. Agullo, A. Abad-Somovilla, A.
product Quality & Safety Risk Assessment (No. GJFP2015006). Abad-Fuentes, Development of immunoafnity columns for pyraclostrobin

Please cite this article in press as: F. Ma, et al., Simultaneous determination of capsaicin and dihydrocapsaicin for
vegetable oil adulteration by immunoafnity chromatography cleanup coupled with LCMS/MS, J. Chromatogr. B (2015),
http://dx.doi.org/10.1016/j.jchromb.2015.12.017
G Model
CHROMB-19774; No. of Pages 8 ARTICLE IN PRESS
8 F. Ma et al. / J. Chromatogr. B xxx (2015) xxxxxx

extraction from fruit juices and analysis by liquid chromatography with UV [30] P.W. Li, F. Ma, Q. Zhang, Q.Q. Yang, L.X. Zhang, X.X. Ding, W. Zhang, The design
detection, J. Chromatogr. A 1218 (2011) 49024909. and synthesis of capsaincin hapten and its application in polyclonal
[28] H.Z. Senyuva, J. Gilbert, Immunoafnity column clean-up techniques in food antibodies, China Patent Int. Cl., CN103951577A, 2014.07.30.
analysis: a review, J. Chromatogr. B 878 (2010) 115132. [31] J. Xie, T. Peng, J.L. He, Y. Shao, C.L. Fan, Y. Chen, W.X. Jiang, M. Chen, Q. Wang,
[29] M. Gasparini, M. Curatolo, W. Assini, E. Bozzoni, N. Tognoli, G. Dusi, X.Y. Pei, S.Y. Ding, H.Y. Jiang, Preparation and characterization of an
Conrmatory method for the determination of nandrolone and trenbolone in immunoafnity column for the selective extraction of aatoxin B1 in 13 kinds
urine samples using immunoafnity cleanup and liquid of foodstuffs, J. Chromatogr. B 998 (2015) 5056.
chromatographytandem mass spectrometry, J. Chromatogr. A 1216 (2009) [32] A.K. Hewavitharana, Matrix matching in liquid chromatographymass
80598066. spectrometry with stable isotope labelled internal standardsis it necessary?
J. Chromatogr. A 1218 (2011) 359361.

Please cite this article in press as: F. Ma, et al., Simultaneous determination of capsaicin and dihydrocapsaicin for
vegetable oil adulteration by immunoafnity chromatography cleanup coupled with LCMS/MS, J. Chromatogr. B (2015),
http://dx.doi.org/10.1016/j.jchromb.2015.12.017