Sei sulla pagina 1di 10

Research Article

Received: 22 February 2007 Revised: 29 February 2008 Accepted: 12 August 2008 Published online in Wiley Interscience: 7 October 2008

(www.interscience.com) DOI 10.1002/jsfa.3411

Residues of antibiotics and sulfonamides


in honeys from Basque Country (NE Spain)
J Serra Bonvehı́a∗ and A Lacalle Gutiérrezb

Abstract
BACKGROUND: The aim of this work was to ensure that Label Basque market honey is free of veterinary residues.

RESULTS: A total of 567 Basque honey samples were previously analyzed with the respective Charm II system – 68 samples were
presumptive positive for sulfonamides (SA-s), 24 samples for tetracyclines (TC-s), and no positive samples for chloramphenicol
(CAP) (<0.3 µg kg−1 ) residues. The residues were mostly confirmed by liquid chromatography fluorescence detection (LC-FD)
and tandem mass spectrometry (LC-MS/MS), according to the latest European Union criteria for the analyses of veterinary drug
residues (2002/657/EC). These techniques confirmed that 19 of the 68 samples, presumptive contaminated with SA-s, contained
sulfathiazole (STZ) residues at levels from 20 to 210 µg kg−1 , and the 24 samples presumptive contaminated with TC-s, were also
confirmed, showing tetracycline (TC) levels from 15 to 920 µg kg−1 . Linearity range, decision limit (CCα), detection capability
(CCβ), precision and reproducibility were also determined.

CONCLUSION: Residues of veterinary drugs were confirmed in a very limited number of honey samples: sulfathiazole (3.40%)
and tetracycline (4.22%). This work reports the advantages of the Charm II assay, but also its limitations, detecting SA-s in
most (87.7%) of the heather (Erica vagans) honey samples. The false positives detected in this honey were assumed to be of an
unknown compound that has not been confirmed as a drug residue. Until now, no studies have been performed to find out if
other heather honeys of different geographical origins give similar false positives for SA-s.
c 2008 Society of Chemical Industry

Keywords: sulfonamides; tetracyclines; raw honey; Charm II test; liquid chromatography/tandem mass spectrometry

INTRODUCTION Required Performance Limit (MRPL) of the analytical method of


Antibiotics are used in the prevention and treatment of European detection was established for chloramphenicol (0.3 µg kg−1 ).6 The
foulbrood (EFB) (Melissococcus pluton) and American foulbrood consumer does not expect any contamination with antibiotics in
(AFB) (Paenibacillus larvae) in honeybees.1 Since the mid-1990s, a natural product such as honey, because honey itself has a bac-
there has been an increase number of AFB cases in Europe.2 Drugs teriostatic action.7 Switzerland has established a general MRLs
known to be effective against these diseases are tetracyclines of 50 µg kg−1 for sulfonamides, 20 µg kg−1 for tetracyclines, and
(TCs) [mainly oxytetracycline hydrochloride (OTC), and tetracycline 20 µg kg−1 for streptomycin in honey.8 A procedure to decontam-
(TC)], streptomycin and sulfathiazole (STZ). Powdered Terramycin inate AFB-infected equipment and to control AFB is the utilization
(oxytetracycline hydrochloride) is used in Spain (200 mg per hive). of gamma-radiation.9 This practice showed good results in Aus-
The drug is administered in three applications of sugar syrup or tralia, the only country where gamma-radiation is allowed on a
three dustings at 4- to 5-day intervals. Terramycin should be fed commercial basis. Several screening kits (Tetrasensor, Elisa, Scin-
early in the spring or fall and consumed by the honeybees before tillation detection)10 – 14 are available to detect the presence of
the main honey flow begins, to avoid contamination of honey antibiotic residues in honey at low levels, but they show sev-
production. The treatment has to be removed at least 6 weeks eral handicaps: (1) they do not identify analyte; (2) they display
before main honey flow. Apicicline was also used in veterinary interferences and false positive results; (3) presence of epimers
practice in Spain; it contains 0.4% oxytetracycline and 4% sul- reduces the sensitivity; and (4) matrix effects are observed. Charm
fathiazole as active compounds. However, the drugs do not kill II assay has been used (since about 2000) in the detection of
the P. larvae because of the presence of resistant bacteria.1,2 The antibiotics in animal food and other matrices, including honey.
safety of sulfonamides (SA-s) to consumers has been questioned
because of their apparent toxicity.3 OTC and TC are also used in
∗ Correspondence to: J Serra Bonvehı́, Research and Development of Nederland,
treatment of human bacterial infections and, thus, this method of
Co., P.O. Box 34, 08890 Viladecans (Barcelona), Spain.
control is not advisable. As a result, honey could become contam- E-mail: serrjosep@gmail.com
inated with these medical products. In Germany and several other
countries (Japan, Switzerland, Scandinavia, etc.), the use of such a Research and Development of Nederland, Co., P.O. Box 34, 08890 Viladecans
antibiotics is not approved for the treatment of honeybees.4 In the (Barcelona), Spain
Council Regulation (EEC) 2377/90,5 no maximum residues (MRLs) b Neiker, Nekazal Ikerketa eta Garapenerako Euskal Erakundea, Berreaga Kalea
63

are fixed for antibiotics in honey; however in 2003, a Minimum 1, 48160 Derio (Bizkaia), Spain

J Sci Food Agric 2009; 89: 63–72 www.soci.org 


c 2008 Society of Chemical Industry
www.soci.org J Serra Bonvehı́, A Lacalle Gutiérrez

This screening procedure is based on a radio immuno assays de-


Table 1. Detection and confirmation of presumptive positive
tection system for chemical families of drug residues with different residues in unifloral and multifloral honeys
limits of detection.15 It uses class-specific receptors or antibodies
in imune/binding assay formats. Results are given in numerical Charm II Test LC-FD
counts, exhibiting high sensitivity and good repeatibility for the Botanical origin
(no. samples) SAs TCs CAP STZ TC
antibiotics tested. For the detection of any antimicrobial drug, two
reagents are required: a [14 C]- or [3 H]-labeled antimicrobial drug Erica spp. (65) 57 15 0 8 15
and a binding reagent (microbial cell containing receptors for Eucalyptus sp. (37) 0 1 0 0 1
the antimicrobial drugs). A variety of chromatographic methods Castanea sativa (12) 0 0 0 0 0
has been used to confirm and quantificate presumptive antibi- Multifloral (453) 11 8 0 11 8
otic residues in honey. In this paper, two analytical methods are
proposed for the quantitative determination of SA-s and TC-s SA-s: sulfonamides; TC-s: tetracyclines; CAP: chloramphenicol
in honey by LC-FD, and the positive samples were also con-
firmed by liquid chromatography coupled to mass spectrometry
(LC-MS/MS). Several liquid chromatographic methods have been honey. For each sample, data about its handling by the producer,
reported for determination of SA-s and TC-s using ultraviolet (UV) extraction date, type of bee-hive and route of conveyance was
detection; however, these methods have insufficient sensitivity, collected. Maturity and deterioration criteria indicated that the
and interferences from the honey matrix do not allow quantifi- quality of the honeys was good, and none of the samples showed
cation by UV at the level 10 µg kg−1 .16 – 19 A rapid reverse-phase signs of fermentation. After this laboratory study, the samples
high-performance liquid chromatography method with fluorimet- were reclassified as unifloral or multifloral honey, according
ric detection is mainly applied, due to the low sensitivity of UV requirements defined in the Basque Label Quality.26 The unifloral
detection and the high cost of the mass spectrometry instru- honeys (provided by Basque beekeepers and produced in the years
ment. SA-s have bean determined by fluorescence detection with 2003 and 2004) were defined as sweet chestnut (Castanea sativa),
precolumn derivatization with fluorescamine, using previous liq- heather (Erica vagans plus E. cinerea), and eucalyptus (Eucalyptus
uid–liquid extraction.20,21 In addition, in the experiment reported sp.) (Table 1).
here, extract ion and cleanup procedure was applied to suc-
cessfully isolate TC-s, which combined postcolumn magnesium
Chemicals
acetate with fluorescence detection to quantify TC-s at this low-
HPLC-grade dichloromethane, methanol, and acetonitrile were
level.22 Fortunately, the cleanup for honey extracts through Oasis
from Scharlab (Barcelona, Spain). Analytical grade citric acid, suc-
HLB cartridges considerably reduces the interferences and their
cinic acid, boric acid, oxalic acid, magnesium acetate, sodium
negative influences on the selectivity of the analytical signal and
hydroxide, trichloroacetic acid, disodium phosphate, phosphoric
maintenance of instruments, especially in LC-MS/MS analysis.8,23
acid, hydrochloric acid (37%), formic acid (98%), and ammo-
To improve the precision and accuracy of the analytical procedure
nia (25%) were supplied by Acros (Geel, Belgium), and flu-
utilized, the validation was established taking into account the
orescamine was obtained from Merck (Darmstadt, Germany).
European Community Decision 2002/657/EC.24 For the analysis
Sodium 1-decanesulfonate (for ion-pair chromatography) was
of banned compounds (such as sulfonamides and tetracyclines
obtained from Sigma Chemical Co. (St Louis, MO, USA). Ultra-
for which no MRLs have been set) by LC-MS/MS, a minimum of
pure water (Milli-Q, Millipore Corp., Bedford, MA) was prepared
four identification points (IP-s) are required. Pang et al.23 present
for chromatographic use. Sulfonamides – sulfamethizole (SML),
LC-MS/MS procedure with satisfactory results without the hy-
sulfamethoxypyridazine (SMP), sulfadoxine (SDX), sulfathiazole
drolysis step required by other chromatographic methods with
(STZ), sulfamerazine (SMR), sulfamethazine (SMZ), sulfachloropy-
different detection systems, including LC-MS analysis. However,
ridazine (SCP), sulfadimethoxine (SDM), sulfamethoxazole (SOZ),
in the present study a hydrolysis step prior to extraction was
sulfisoxazole (SXZ), sulfaguanidine (SGD), sulfadiazine (SDZ), sul-
adopted to ensure that sugar-bound sulfonamides in honey were
fanilamide (SNL), sulfacetamide (SAA), and sulfapyridine (SPY),
included in the final result.8,20,25 In addition, the above-mentioned
and the hydrochlorides of tetracycline (TC), oxytetracycline (OTC),
data confirm that honey interactions between sugars and SA-s
and chlortetracycline (CTC) – were from Sigma, sulfachlorpyrazine
decreases biological activity, making the product more available
(SPZ) (Ciba-Geigy Corporation, Ardsley, NY), 4-epitetracycline (4-
under acidic conditions. In consequence, it could be transformed
epiTC), 4-epioxytetracycline (4-epiOTC), and 4-epichlortetracycline
into free active compound in the human digestive system.
(4-epiCTC) were obtained from Acros (Fisher Scientific, Schwerte,
The aim of the present work was to obtain an overview to ensure
Germany).
that Label Basque marketed honey is free from drug residues.26
Reagents, standards and buffers [tablet reagents (radioactive
Moreover, all honey and syrup stored during medication must be
material), MSU multi-antimicrobial concentrate standard, zero
removed to avoid human consumption. In fact, honey purchasers
control standard, MSU extraction buffer, and scintillation fluid
requests suppliers to certify that the offered honey is pure and free
(Optifluor)] are included in the Charm II drug test for honey
of veterinary drugs.
(Charm Sciences, Inc., Lawrence, MA).

MATERIALS AND METHODS Instruments


Materials Charm II system
The study was carried out on 567 selected samples from different The Charm II 7600 analyzer (Charm Sciences, Inc.) assay is a rapid
floral origins. All samples were unheated, and stored in dark antibody assay that uses [3 H]- and [14 C]-tagged drug tracers with
and dry places at room temperature. The samples were taken broadly specific binding agents in a multipurpose single sample,
liquid scintillation counter (LSC) and luminometer.14,27 When the
64

directly from the containers that beekeepers use for the storage of

www.interscience.wiley.com/jsfa 
c 2008 Society of Chemical Industry J Sci Food Agric 2009; 89: 63–72
Antibiotic and sulfonamide residues in Basque Country honey www.soci.org

binding reagent is added to a sample with antimicrobial drugs, the standard solution was added (tablet reagent dissolved in 300 µL
contaminating antimicrobial drug binds to receptors in the cell. of distilled water), and incubated at 85 ◦ C for 3 min. The samples
This prevents the [14 C] or [3 H] antimicrobial drug from binding to were then homogenized and centrifuged (2000 × g, 10 min). The
these sites. Therefore, the more [14 C]- or [3 H]-labeled antimicrobial supernatant was discarded and the residue redissolved in 300 µL of
drug bound, the less antimicrobial drug there is in the sample. The distilled water and 3 mL scintillation fluid (Opti-fluor). Immediately,
detection reaction is stopped with a centrifugation step, where the solution was measured on [3 H] channel of Charm analyzer in
unbound tracer is separated from bound tracer–binder complex, counts per minute (cpm), and compared with control point.27
and analyzed in a scintillation counter for 1 minute to give a
resulting count. The amount of [14 C] or [3 H] bound to the cells Tetracyclines
is measured in counts per minute (cpm). The higher the count, Sample preparation for tetracyclines was restricted to a simple
the less drug contamination in the sample, and vice versa. The dilution step (1 : 10) by dissolving 1 g of honey in 9 mL of MSU
result is simplified to a present/absent result using a control extraction buffer supplied with the test kit. The receptor tablet was
point. The control point is the cutoff number between a negative suspended in 300 µL of distilled water, 4 mL of the diluted honey
and a positive result. For establishing a control point, six negative solution added, and the mixture incubated at 35 ◦ C for 5 min. The
controls (known negative honey samples) have to be run. Then the samples were then homogenized, centrifuged (2000 × g, 10 min.),
results are averaged and 25% subtracted. This value indicates the and the supernatant discarded. Finally, the residue was dissolved
control point. Test results greater than the control point indicate with 300 µL distilled water and 3 mL scintillation fluid (Opti-fluor).
a negative sample, while results less than or equal to the control Immediately, the solution was measured on [3 H] channel of the
point indicate the sample is presumptive positive and needs to be Charm analyzer (cpm), and compared with control point.27
retested.14,27
Chloramphenicol
Liquid chromatography system The procedure is recommended for honey samples with a Pfund
The liquid chromatography consisted of a Model Prostar 240 colour index of 83–114 mm (e.g. heather honey). About 10 g
LC pumping unit; Model 410 Autosampler fitted with a 50 µL of honey was weighed in a 50 mL centrifuge tube, and 30 mL
loop; 330 photodiode array detector (PAD), and 363 fluorescence of the supplied MSU extraction buffer added. The solution was
detector (all from Varian Chromatography Division, Palo Alto, thoroughly mixed for 1 min., and adjusted the pH to 7.5 with
USA). Chromatographic data from LC were processed on Star M2 buffer included in the test kits. Then 5 mL of the solution
Chromatography WorkStation software, version 6.20, for data was loaded onto a CH cartridge Varian (500 mg, 6 mL) that had
processing. LC-MS/MS experiments were performed using a been previously conditioned with 5 mL of 100% methanol. The
Finnigan MAT TSQ 70 triple quadrupole MS/MS system (Finnigan samples were washed with 5 mL of 20% methanol, and the
MAT, San José, CA) equipped with Finnigan MAT thermospray analyte was extracted with 2 mL of 75% methanol. The receptor
interface. white tablet was suspended in 300 µL of distilled water, 5 mL of
the diluted honey solution added, and the mixture incubated at
50 ◦ C for 3 min. The sample was homogenized and the test tube
Sample preparation for analysis by Charm II test
returned to the incubator at 50 ◦ C for an additional 3 min. The
The honey extract had active reagents added in sequential and green tablet was added, the mixture homogenized and incubated
competitive assay formats at various incubation temperatures at 50 ◦ C for 3 min., then centrifuged at 5000 rpm for 5 min. The
optimized for drug detection. Extraction procedure was as supernatant was discarded. Finally, the residue was dissolved
described in Operator’s manual for sulfa 011 and test for with 300 µL distilled water and 3 mL scintillation fluid (Opti-fluor).
antimicrobial drugs in honey (beta-lactams and tetracyclines), Immediately, the solution was measured on [3 H] channel of Charm
including a new edition for chloramphenicol assay.27,28 The test analyzer (cpm), and compared with control point.27,28
took 12–20 minutes for a multitude of antibiotics. Sulfonamides
assay used a more complex acid hydrolysis and reverse phase Sample preparation for LC-FD
preparation to eliminate p-aminobenzoic acid (PABA), and convert Sulfonamides
carbohydrate-SAs into free form (total time 1 h).
The sulfonamides were extracted from 5 g of homogenized honey
with 5 mL of 0.1 g mL−1 trichloroacetic acid solution. The samples
Sulfonamides were shaken for 10 min. at room temperature on a mechanical
About 5 g of honey was transferred into a 50 mL conical centrifuge shaker and heated for 1 hour at c. 64 ◦ C. The pH was adjusted to
tube and homogenized with 20 mL of 1 mol L−1 HCl. After ho- 6.5 with 1 mol L−1 Na2 HPO4 , pH 12 (c. 2.5 mL), and later 10 mL
mogenization, the solution was incubated at room temperature of acetonitrile and 2.5 mL of dichloromethane were added. The
for 1 h. Then, 2.3 mL of 7.5 mol L−1 NaOH were added to sample mixtures were shaken again for another 10 min. on a mechanical
solution, followed by the measurement of pH (1.0 to 5.0). The pH shaker, and centrifuged at 6000 rpm for 10 min. The supernatant
was adjusted to 7.7–8.0 drop by drop with 0.075 mol L−1 NaOH. was poured into a 25 mL volumetric flask. Re-extract twice and
The filtrate solution was applied to an Bond Elut C18 cartridge made up to volume with acetonitrile. Of the organic extract, 10 mL
(500 mg, 3 mL) (Varian, Harbour City, CA), preconditioned with was evaporated to dryness under a nitrogen stream at 50 ◦ C.
5 mL of methanol and 5 mL of distilled water (twice). After the The extract residue was reconstituted in 1 mL of solution buffer
extract had passed through at a flow rate of 1–2 drops per sec- for derivatization [acetonitrile/H3 PO4 pH 6.0 and citrate buffer
ond, the cartridges were washed with 5 mL of distilled water. The 0.5 mol L−1 pH 3.0) in the ratio (6/4, v/v)], adding 20 µL of internal
analytes were eluted with 1 ml of methanol, and the eluate was standard (5 µg mL−1 of sulfanilamide in methanol), and 0.2 mL of
evaporated to dryness under a nitrogen stream (40–45 ◦ C). The 10 mg mL−1 fluorescamine solution (50 mg fluorescamine in 5 mL
dry residue was reconstituted in 5 mL of Zero Control Standard so- of acetone). The solution was allowed to develop the reaction for
40 min. at room temperature.20,21
65

lution, and stopped with cold ice for 10 min. Finally, the deuterated

J Sci Food Agric 2009; 89: 63–72 


c 2008 Society of Chemical Industry www.interscience.wiley.com/jsfa
www.soci.org J Serra Bonvehı́, A Lacalle Gutiérrez

Tetracyclines Tetracyclines. LC was performed with a Chromsep 5 C8 (100 ×


The tetracyclines were extracted from 5 g of homogenized honey 4.6 mm i.d.; 5 µm particle size) column and a guard column
with 25 mL of succinic buffer solution at pH 4 (5.9 g succinic Chromsep 5 C8 (4.0 × 2 mm i.d.) (Varian). The mobile-phases were
acid in 1 L distilled water adjusting pH 4 with 1 mol L−1 NaOH), (A) 1.22 g sodium 1-decanesulfonate dissolved in 1 L of 0.01 N
and this mixture was vortexed to complete dissolution, before oxalic buffer solution adjust to pH 2.25–2.5 with 1 N NaOH
being centrifuged at 3500 rpm for 15 min. at room temperature. and (B) 1.22 g sodium 1-decanesulfonate dissolved in 1 L of 0.01
The Oasis HLB solid phase extraction (200 mg per 6 mL) (Waters N oxalic buffer solution/acetonitrile/methanol (2 : 1:1, v/v). The
Chromatography Division, Milford, MA) was conditioned with column was eluted by using a linear gradient: initial time, 32% B;
3 mL acetonitrile and 3 mL distilled water. The supernatant of the 0–15 min., 100% B; 15–17 min., 32% B; 17–25 min., 32% B. The
centrifuged extract was poured into the reservoir and the solution mobile-phase flow rate was of 0.8 mL min−1 ; Injection volume,
passed through the SPE cartridge for 12–15 min. The samples 50 µL. Fluorescence detection was carried out after post-column
were washed with distilled water (3 mL) for 1–2 min., and the TC-s addition of magnesium acetate [(6 g L−1 ) in 0.1 mol L−1 boric acid
were eluted with 5 mL of extraction solution [oxalic buffer solution buffer pH 9], and the derivation reagent was delivered at a flow
(0.01 mol L−1 oxalic acid adjusted to pH 2.25–2.5 with 1 mol L−1 of 0.3 ml min .−1 . The settings of the fluorescence detection were:
NaOH), and acetonitrile in the ratio (30/8, v/v)].22,29 λexc 385 nm, λem 500 nm, sensitivity 0.3, slit 7 nm [Model 363
ProStar].22,29
Sample preparation for LC/MS-MS confirmation
Sulfonamides and tetracyclines
About 10 g honey was dissolved in 20 mL of 2 mol L−1 HCl and the LC-MS/MS system
samples were allowed to stand for 30 min. at room temperature. Sulfonamides and tetracyclines. Chromatographic separation was
After incubation, 40 mL of 0.3 mol L−1 citric acid was added. The pH performed on a column Nucleosil 100-5, C18 HD (50 × 2 mm i.d.,
was adjusted to 3.4–4.5 with ammonia solution (7.14 mol L−1 ) in 5 µm particle size) equipped with 5 µm Nucleosil 100, C18 HD
20 mL honey filtrate. The solution was passed for 10–15 min. over guard column (4.0 mm × 2 mm i.d.). Column temperature set at
SPE cartridge (Waters Oasis HLB 200 mg per 6 mL) preactivated 25 ◦ C. The mobile-phases were (A) 50 mL acetonitrile and 3 mL
with 3 mL of acetonitrile and 2 mL of distilled water. Later the formic acid (98%) l L distilled water, and (B) 3 mL formic acid
sample was washed with 3 mL of distilled water (three times), dried, in 1000 mL graduated flask and adjusted with acetonitrile. The
and the retained analytes were eluted with 3 mL of acetonitrile in column was eluted by using a linear gradient: 0–10 min., 30% B;
a graduated flask. The extract was then evaporated at 40 ◦ C under 10–12 min., 30% B; 12–12.1 min., 0% B; 12.1–19 min., 0% B. The
nitrogen flow to a small volume, and diluted by adding 0.5 mL solvent flow rate was 0.2 mL min .−1 ; injection volume, 10 µL. A
of mobile phase A, and the graduate flask reweighed. Finally, the tandem mass spectrometer equipped with electrospray ionization
solution was transferred without filtration into a vial.8 (ESI) was used for detection. The ESI source operated in the positive
ionization mode at 90 ◦ C, and desolvation was performed at 250 ◦ C
Chloramphenicol with a desolvation gas flow of 560 L h−1 and a cone gas flow of
About 1 g honey was dissolved in 1 mL of distilled water at 60 ◦ C 50 L h−1 . For desolvation and nebulization, high-purity nitrogen
for 5 min. Then, 5 mL acetonitrile was added, and the mixture was used, and argon (99.998%) served as the collision gas. MS/MS
shaken vigorously for 10 s, centrifuged for 3 min. at 3500 rpm, and was performed in both the daughter scan and neutral loss scan
the supernatant was transferred to glass-stoppered tube. Then modes, applying multi-reaction monitoring (MRM) using the two
1 mL distilled water and 5 mL acetonitrile were added, and the most intense and specific fragment ions.8
mixture was shaken vigorously for 6 s, recentrifuged for 2 min. at
3500 rpm, and the upper layer (water) discarded. The organic layer
was evaporated to dryness under a nitrogen stream (60 ◦ C), and the Chloramphenicol. Chromatographic separation was performed
residue was dissolved in 100 µL of methanol, and 3 mL of distilled on a column Hypersil-Keystone Beta Basic C18 (150 × 2.1 mm i.d.,
water and 0.5 mL of buffer solution 0.5 mol L−1 Na2 HPO4 .2H2 O 5 µm particle size) (Thermo Electron, Madison, WI). The mobile-
(pH 6 adjusted with 0.85 g mL−1 phosphoric acid) added.30,31 phases were (A) water and (B) methanol. The column was eluted
by using a linear gradient: initial time, 30% B; 0–3 min., 95%
Instrumental analysis B; 3–5 min., 95% B; 5–5.1 min., 30% B. The solvent flow rate
LC-FD detection was 0.25 mL min .−1 ; injection volume, 20 µL. MS/MS conditions:
Sulfonamides. The separation was obtained with a Nucleosil Ionization mode polarity: ESI, negative; source temperature,
100-5 C18 AB (250 × 4.6 mm i.d., 5 µm particle size) column 350 ◦ C; scan event (m/z 152–256) and collision energy (m/z
provided with a Nucleosil 100-5 C18 AB guard column (8×4.6 mm) 157–272).30,31
(Varian, Middelburg, The Netherlands). The mobile-phases were
(A) 0.02 mol L−1 H3 PO4 and (B) CH3 OH/Acetonitrile (1 : 1, v/v). The
column was eluted by using a linear gradient: initial time, 40% Standards solutions
B; 0–20 min., 40% B; 20–25 min., 50% B; 25–39 min., 50% B; Sulfonamides
39–40 min., 55% B; 40–50 min., 55% B; 50–51 min., 40% B; Stock solutions of sulfonamides and sulfanilamide (SA) as internal
51–58 min., 40% B. The mobile phase flow rate was of 0.6 mL standard at the concentration of 1 mg mL−1 were prepared in
min .−1 , injection volume, 50 µL. Use of photodiode array detector methanol, and stored protected from light at 4 ◦ C. A stock standard
coupled on line with fluorescence detection. The settings of the mixture (10 µg mL−1 ), was prepared in distilled water. Working
fluorescence detection were λexc 405 nm, λem 496 nm, sensitivity standard solutions with concentrations in the expected ranges
0.3, and slit 7 nm (Model 363 ProStar); the maximum UV absorption were prepared from this stock standard solution at 10, 20, 50, 100
of the SA-s were found to be in the range of 260–290 nm.20,21 and 200 ηg mL−1 .
66

www.interscience.wiley.com/jsfa 
c 2008 Society of Chemical Industry J Sci Food Agric 2009; 89: 63–72
Antibiotic and sulfonamide residues in Basque Country honey www.soci.org

Tetracyclines samples with the SA-s and TC-s described at three concentrations
The tetracyclines were dissolved in 100 mL of methanol (21.6 mg ranging from 50 to 150 µg kg−1 (n = 5 for each concentration).
tetracycline hydrochloride, 22.2 mg chlortetracycline hydrochlo- The following equations were used:
ride and 21.6 mg oxytetracycline hydrochloride) for a final con-
centration of 200 µg mL−1 . The stock solution was diluted to Accuracy (%) = [(measured concn − actual concn)/
appropriate concentration (50 µg mL−1 ) with methanol. A second actual concn] × 100
standard solution (2.5 µg mL−1 ) was prepared daily by dilution
Precision (%) = (standard deviation/mean concn) × 100
of standard solutions with extraction solution [oxalic buffer solu-
tion/acetonitrile (30/8, v/v)]. A calibration curve in the range 10 to
300 ηg mL−1 was prepared daily. Standard solution of the 3-epi According to the guidelines, limit of detection (LOD) and limit of
metabolites was also prepared, but they were only analyzed for quantification (LOQ) are expressed as the decision limit (CCα) and
identification. detection capability (CCβ), respectively. The decision limit is the
lowest concentration level of the analyte that can be detected
in a sample with a an error probability of 1% that a sample is
Chloramphenicol (CAP) noncompliant. The detection capability is the lowest concentration
A stock solution of CAP was prepared by dissolving 25 mg of at which a method is able to detect truly contaminated samples
chloramphenicol in 25 mL of methanol (stable for at least 3 months with an error probability of β (β = 5% for banned compounds).
at 4 ◦ C). The stock solution was diluted with appropriate volumes Blank material was fortified at 5 different concentrations (n = 20)
of methanol/water (3/7, v/v) to create working standards for and the standard error of the y intercept was calculated.
calibration curve (0.2, 0.5, 1, 2, 5, and 10 ηg mL−1 ). The decision limits (CCα = 2.33 × standard error of the y
intercept) and the detection capabilities [CCβ = CCβ + (1.64 ×
Method validation standard deviation of 20 spikes at CCα)] were determined.24
Validation of the LC method for sulfonamides and tetracyclines
determination Charm test
Method validation was carried out according to the criteria The validation method for each antibiotic and sulfonamide is
described in the Council Directive 96/23/EC.24 The parameters incorporated as an appendix to the Cham II manuals for honey. The
taken into account were the following: response linearity, decision decision limit was obtained using negative honey with established
limit, detection capability, trueness, and precision. Usually the control point, spiked with the sulfonamides, tetracyclines and
quantification of drug residues is performed using a matrix- chloramphenicol at different concentrations for obtaining results
matched calibration curve made from fortified blank samples less or equal to this control point.
prepared in the same matrix as the real samples. A calibration curve
was established by measuring sulfonamides and tetracyclines
peak areas over a five-fold range of concentrations in fortified Melissopalynological analysis
blank honeys. The slope, intercept, and correlation coefficient The analysis of floral origin was carried out in accordance with the
were calculated by linear regression analysis. Standard calibration methods of the International Commission of Bee Botany (ICBB)
curves (four replicates each) for the analytes were plotted. These of the International Union of Biological Sciences (IUBS),32 which
calibration curves were prepared repeatedly during the study with involved 1200 pollen grains per sample, as well as honeydew
highly reproducible results. In order to improve the accuracy, the indicators.33 The microscope preparations were made without
internal standard quantification was applied to detect sulfonamide acetolysis to preserve all the components in the sediments
residues at low levels. Internal standard (IS) was a difficult extracted. Pollen was observed at ×400–1200 magnifications.
process, as there were no regions of the chromatogram that were
completely free of extract-born interferences. The sulfanilamide
(SLN) was selected based on the relative lack of interferences, RESULTS AND DISCUSSION
appropriate retention time, low cost, and well-characterized purity, Charm test
and is the compound most frequently referred to as suitable for The results of drug contamination in autochthonous honey
this purpose.20,21 Furthermore, among the samples analyzed, none samples (Table 1) indicated a low incidence of SA-s and TC-s
was found for SLN. However, the selection of a possible internal residues in honey from Basque Country market. Of a total of 567
standard has to be done with great care and, if is possible, an samples, 68 samples (12%) were positive for sulfonamides, and
adequate stable isotope labeled internal sulfonamide is preferred. 24 samples (4%) for tetracyclines. The CCα was about 4 µg kg−1
The rations of the peak area (A) of each SA-s to that of the internal (SOZ), 5 µg kg−1 (SDM, SMR, SDZ, SPZ, and SMP), 6 µg kg−1 (SXZ),
standard (IS) were plotted against concentration described in 8 µg kg−1 (SDX), 9 µg kg−1 (STZ), 10 µg kg−1 (SMZ), 15 µg kg−1
calibration standard solutions. The precision study was evaluated (SML), and 50 µg kg−1 (SAA). For tetracyclines, the CCα was
with the SA-s (STZ, SML, SMZ, and SLN as internal standard), and 5 µg kg−1 (TC), 10 µg kg−1 (OTC and CTC), and 0.3 µg kg−1 for
TC-s (TC, OTC, and CTC) most used in veterinary practice in Spain. CAP. In general, the frequency of residues of antibiotics in honey
The recovery and repeatability of the procedures were evaluated from local beekeepers was low.15,20,34,35 No residues of CAP were
by the analysis of six spiked samples with these SA-s and TC-s found. However, positive results were obtained in the majority of
at 25, 50, 100, and 150 µg kg−1 based on the linearity range, the samples of heather (Erica vagans) honey with the Charm II test
on three different days (n = 24). Repeatability was evaluated for sulfonamides, but could not be confirmed by chromatographic
by applying the whole extraction procedure six times to the procedures. Only 19 of 57 samples presumptive contaminated
same sample. Reproducibility was determined by analyzing each with SA-s were confirmed. This divergence has been induced by
sample of honey fortified on five different days over about a the presence of a natural compound detected in this honey, which
67

month. Precision and accuracy were determined using spiked was not confirmed as residue by LC-MS/MS detection, but at the

J Sci Food Agric 2009; 89: 63–72 


c 2008 Society of Chemical Industry www.interscience.wiley.com/jsfa
www.soci.org J Serra Bonvehı́, A Lacalle Gutiérrez

same time present affinity by a microbial cell containing a specific


Table 2. Absolute recoveries of sulfonamides and tetracyclines from
receptor incorporated in the Charm test for sulfonamide (and spiked samples (ηg g−1 )
not antibodies). Also honey with very high hydroxymethylfurfural
content can lead to false positive results. Component Component Rec.
Component added found (%)

LC-FD method for sulfonamides STZ 25.1 18.6 74.1


51.4 37.6 73.2
Numerous analytical methods have been applied for the identifi-
101.1 77.2 76.4
cation and quantification of SA-s in honey, including colorimetric
150.7 122.5 81.3
procedures, derivative spectrophotometry, and chromatographic
SMZ 25.4 18.1 71.3
methods (TLC/LC), after a simple extraction and clean-up.16,17,20,36
52.4 37.1 70.8
In our laboratory, the LC-FD method has been developed to verify
103.1 74.3 72.1
and quantify trace drugs in this biological matrix, using pre-column
149.7 110.6 73.9
derivatization with fluorescamine at pH 3. This derivatization tech-
SML 25.2 17.9 71
nique is preferred because the reagents are easily prepared, the
50.3 36.0 71.6
reaction occurs at room temperature, and the fluorescamine and
102.1 74.9 73.4
reaction byproducts are not fluorescent. The disadvantage of this
151.2 120.2 79.5
method is that the derivatives are not stable and they must be
OTC 25.1 22.2 88.4
prepared and analyzed one at a time. The derivatization was es-
50.4 44.4 88.1
tablished in 40 min. at room temperature (25 ◦ C). The CCα for STZ
101.7 91.8 90.3
and SMZ was 10 µg kg−1 , and 12 µg kg−1 for SML. The CCβ was
150.4 137.9 91.7
15 µg kg−1 for STZ and SMZ, and 17 µg kg−1 for SML. When IS was
TC 25.3 22.7 89.7
incorporated in standard calibration and sample determination,
50.7 45.3 89.4
calculated peak response ratios, ranges of the recoveries, and
100.3 91.4 91.1
coefficients were highly improved compared to those mentioned
151.1 139.9 92.6
in the literature.20,21 Recover rates obtained were 73.2–81.3% for
CTC 25.4 21.9 86.2
STZ, 70.8–73.9% for SMZ, and 71–79.5% for SML (Table 2). The
51.1 44.2 86.5
precision of the method was determined as repeatability (RSDr )
101.4 89.8 88.6
and reproducibility (RSDR ) at three concentrations for STZ, SMZ,
150.2 135.3 90.1
and SML (Table 3). The RSDr and RSDR obtained were lower than
2.10%, with a mean of 1.92%, and 2.54% with a mean of 2.38%, STZ: sulfathiazole; SMZ sulfamethazine; SML sulfamethizole; OTC:
respectively. The results were analyzed according to an outlier oxytetracycline; TC: tetracycline; CTC: chlortetracycline.
n = 24 determinations.
test (Cochran and Grubbs test). As expected, the SA-s gave linear
response from 0 to 200 µg kg−1 . The correlation coefficient should
be >0.995, and linear regression of the amount injected vs peak
area/height gave a correlation coefficients of 0.9957 for STZ, 0.9971
for SMZ, and 0.9997 for SML. These results indicate that the method
had satisfactory accuracy.37 Our studies showed the need to adjust
the initial gradient conditions in order to elute all SA-s, providing
a satisfactory capacity factor for the first eluting SA-s (STZ) (Fig. 1).
The retention time (RT) of the SA-s referenced ranged from 10 to
30 min. on the basis of five parallel determinations over 5 days,
the precision relative standard deviation (SD) of the RT was 2.7%
for STZ, 3.1% for SMZ, and 4.12% for SML, whereas the precision
of peak area values were 3.57%, 3.83%, and 4.12% for STZ, SMZ,
and SML, respectively. Of the 68 presumptive positive samples,
19 were confirmed as such, with sulfathiazole residues at levels
from 20 to 210 µg kg−1 . Of the 57 presumptive positive samples of
heather honey, residues of sulfathiazole were confirmed in only 8 Figure 1. LC chromatograms of blank honey and spiked samples with the
(Table 1), showing the detection in these honeys of a natural com- sulfonamides most used in Spain (0.050 µg mL−1 ). Peaks: sulfanilamide
(intern standard) (1); sulfathiazole (2); sulfamethizole (3); sulfamethazine
pound not confirmed as residue. Honey samples fortified with STZ (4).
(2000 µg kg−1 ) were reduced in 6 weeks to values of 80 µg kg−1 ,
whereas if the honey was fortified with 1000 µg kg−1 , and stored
at 4 ◦ C, the residue content was still 85% of the initial amount after
solvents, and SPE elution volumes for TC-s. When phenyl columns
11 months; this is in agreement with Belliardo.38
were used, some interference could be observed in the RT of the
TC-s. Under the experimental conditions reported here, the SPE
LC-FD method for tetracyclines was applied through Waters Oasis HLB (200 mg per 6 mL). This
The TC-s were separated on a reverse-phase C8 column and polymeric adsorbent does not contain a silanol backbone, without
reacted post-column with magnesium acetate in boric acid buffer the interference of TC-s interacting too strongly with the silanols
(pH 9), at room temperature. To obtain optimum conditions for of a silica-based cartridge.29 Experimental results indicated that
SPE clean-up, several factors were investigated, including the it was not necessary to do a second cleanup through a COOH
68

following: type of clean-up cartridge, washing solvents, eluent solid-phase extraction with this methodology, as adequate values

www.interscience.wiley.com/jsfa 
c 2008 Society of Chemical Industry J Sci Food Agric 2009; 89: 63–72
Antibiotic and sulfonamide residues in Basque Country honey www.soci.org

Table 3. Analytical Relative Standard Deviation (RSD) of sulfonamides


STZ STZ SMZ SMZ SML SML
added found RSD added found RSD added found RSD
(ηg g−1 ) (ηg g−1 ) (%) (ηg g−1 ) (ηg g−1 ) (%) (ηg g−1 ) (ηg g−1 ) (%)

Intraday
60.1 43.9 ± 0.81 1.85 59.4 41.8 ± 0.87 2.08 60.3 42.9 ± 0.81 1.89
59.7 45.5 ± 0.83 1.82 60.5 43.6 ± 0.91 2.09 59.7 43.7 ± 0.83 1.90
60.4 49 ± 0.87 1.78 59.9 43.7 ± 0.87 1.99 60.6 47.8 ± 0.91 1.90
Interday
60.1 44 ± 0.97 2.21 59.4 41.7 ± 1.04 2.49 60.3 42.8 ± 1.03 2.41
59.7 45.4 ± 1.01 2.22 60.5 43.4 ± 1.09 2.51 59.7 43.8 ± 1.09 2.49
60.4 48.9 ± 1.07 2.19 59.9 43.9 ± 1.11 2.53 60.6 48.1 ± 1.12 2.33

n = 18 determinations.

Table 4. Analytical Relative Standard Deviation (RSD) of tetracyclines


OTC OTC TC TC CTC CTC
added found RSD added found RSD added found RSD
(ηg g−1 ) (ηg g−1 ) (%) (ηg g−1 ) (ηg g−1 ) (%) (ηg g−1 ) (ηg g−1 ) (%)

Intraday
56.4 49.6 ± 0.71 1.43 53.7 47.9 ± 0.79 1.65 59.7 51.7 ± 0.83 1.61
55.7 50.2 ± 0.68 1.36 54.3 49.4 ± 0.71 1.44 58.4 51.6 ± 0.78 1.51
56.9 51.9 ± 0.75 1.45 52.6 48.6 ± 0.75 1.54 60.3 54.4 ± 0.86 1.58
Interday
56.4 49.7 ± 1.03 2.07 53.7 48 ± 0.98 2.04 59.7 51.5 ± 1.07 2.08
55.7 50.1 ± 0.98 1.96 54.3 49.2 ± 1.05 2.13 58.4 51.5 ± 1.11 2.16
56.9 52 ± 1.08 2.08 52.6 48.5 ± 1.09 2.25 60.3 53.9 ± 1.15 2.13

n = 18 determinations.

of recovery and repeatability were achieved, without the retention with tetracycline was detected in all 24 presumptive positive
of TC-s on the COOH column.18,39 Thus, the TC-s were eluted on samples at levels from 15 to 920 µg kg−1 (Table 1). A further
Oasis cartridge with oxalic buffer (pH 2.25–2.5) and acetonitrile complication in the determination of TC-s, in particular CTC, is the
in ratio 30 : 8 (v/v). The eluent was also assayed at 3, 4, 5, and fact that it can rapidly isomerizes, even under mild conditions,
6 mL. The results show that 3 mL was not enough volume to to form 4-epimers.39,40 This paper also reviews the stability of
elute all TC-s (OTC 50%, TC 71%, and CTC 50.2%), indicating the main tetracyclines utilized in Spain. At 4 ◦ C and 20 ◦ C, OTC
that the volume recommended in the literature does not have standard solutions were very stable after 11 weeks, with only 4%
good selectivity. However, with 5 mL, all TC-s were eluted (all of epimeric forms. In contrast, CTC and TC epimerized rapidly
>85%), and no improvement was observed in a further elution after 11 weeks at 20 ◦ C, and low levels have been reported after
with 6 mL. Recoveries for the different TC-s were 88.1–91.7% (for 11 weeks at 4 ◦ C, which may lead to false quantification of TC-s, by
OTC), 89.4–92.6% (for TC), and 86.2–90.1% (for CTC) (Table 3). summing the areas of TC-s and their respective epimers without
Several mobile-phases were tested for efficient elution of TC-s: the appropriate correction factors.39 Thus, TC-s epimer values were not
best results were obtained with gradient program with excellent determined. Similar results were obtained with standard solutions
elutions for all analytes – the peak co-eluting with tetracycline stored in the dark.
disappeared and the interference was decreased significantly.
Under these conditions, the TC-s identified were successfully
separated within 16 min., as shown in chromatogram (Fig. 2). The LC-MS/MS confirmation
RSDr , RSDR , and recoveries of spiked samples were determined at For the identification of positive samples, according to the
three concentrations for TC, OTC, and CTC (Table 4). The RSDr and European Community Decision 2002/657/EC,24 the follow criteria
RSDR obtained were lower than 1.66% and 2.26%, respectively. The were used: (a) the signal-to-noise ratio (S/N) of the characteristic
correlation coefficient should be >0.995: linear regression of the ions selected must be ≥3; (b) the differentiation of the retention
amount injected vs peak area/height gave a correlation coefficient time of analyte and corresponding standard should be within
of 0.9954 for OTC, 0.9981 for TC, and 0.9984 for CTC. The CCα ±2.5%; (c) the allowable deviation of the relative abundance of the
values were 2 µg kg−1 for CTC, 4 µg kg−1 for TC, and 5 µg kg−1 characteristic ions of the target matter and that of the characteristic
for OTC. The CCβ values were 5 µg kg−1 for CTC, 9 µg kg−1 for ions of the corresponding standard should be within ±20% to
TC, and 12 µg kg−1 for OTC. Method precision and accuracy were ±50%, depending on the relative ion intensities. The decision limit
4.90% and 29.7% for SA-s, and 4.54% and 13.9% for TC-s (Table 5). (CCα) varies between 0.8 and 12 µg kg−1 for the 16 sulfonamides
The results demonstrate that the applied analytical procedure referenced in European Union (SML, SMP, and SDX, 0.8 µg kg−1 ;
gives enough guarantees for quality control.29,37 Contamination STZ, SMR, SMZ, SCP, SDM, SOZ, and SXZ, 1 µg kg−1 ; SGD and
69

J Sci Food Agric 2009; 89: 63–72 


c 2008 Society of Chemical Industry www.interscience.wiley.com/jsfa
www.soci.org J Serra Bonvehı́, A Lacalle Gutiérrez

Figure 2. LC chromatograms of blank honey and spiked samples with Figure 3. LC chromatograms of unknown peak solely detected in heather
oxytetracycline (OTC), tetracycline (TC), and chlortetracycline (CTC) honey. Peak: unknown (1).
(0.050 µg mL−1 ). Peaks: oxytetracycline (1); tetracycline (2); chlortetra-
cycline (3).
the appropriate control point), 57 samples of a total of 65 samples
(87.7%) were classified as honeys presumptive contaminated
Table 5. Precision and accuracy for determination of sulfonamides with sulfonamides. However, the chromatography (LC-FD and
and tetracyclines LC-MS/MS) showed clearly the presence of a compound not
confirmed as a veterinary drug residue (Fig. 3), with significant
Component Component Error Precision
Component added (ηg g−1 ) found (ηg g−1 ) (%) (%) interactions for individual Charm receptor assay.
To investigate this anomaly, the LC-MS/MS conditions were
STZ 50.7 37.1 ± 1.37 26.8 3.69 optimized to corroborate possible contamination or the presence
100.3 76.4 ± 2.09 23.5 2.74 of a natural compound in heather honey. Sulfonamides are known
149.7 121.3 ± 3.41 18.9 2.81 to possess amphoteric character due to the presence of nitrogen
SMZ 51.1 35.9 ± 1.47 29.7 4.09 functions in their molecules, which because of their positions in
102.1 73.8 ± 2.74 27.7 3.71 the structure can protonate/deprotonate depending on the pH of
150.7 110.2 ± 4.17 26.9 3.78 the medium.8 Usually, the separation of SA-s is performed by LC
SML 50.3 35.9 ± 1.76 28.6 4.90 in a reversed-phase mode under acidic conditions. Mass spectra
101.1 74.4 ± 3.07 26.4 4.12 of all the sulfonamides and tetracyclines were acquired in a full
150.3 110.9 ± 4.39 26.2 3.96 scan mode with cone voltage of 25 V, using positive electrospray
OTC 50.3 44.4 ± 1.67 11.7 3.76 ionization. Cone voltages were optimized for maximum signal
100.7 90.8 ± 2.23 9.8 2.46 intensity of typical ions during an injection of single compounds
149.8 136.9 ± 3.44 8.6 2.51 into the mass spectrometer. The detection of the compounds was
TC 50.1 44.7 ± 1.53 10.8 3.42 divided into reaction time windows with dwell times for single
101.2 92.4 ± 2.83 8.7 3.12 reaction monitoring (SRM), which were set to 0.05 and 0.20 s (per
150.3 138.4 ± 4.37 7.9 3.16 SMR), depending on the intensity of the products ions for the
CTC 50.4 43.4 ± 1.97 13.9 4.54 transition. The most intense fragments in the positive ion mode
100.7 89 ± 2.87 11.6 3.22 were detected at m/z 156, 108, and 92, although the abundance
150.4 135.1 ± 4.57 10.2 3.38 was different depending on the compounds. For confirmatory
n = 15 determinations. purposes, transitions (product ions), cone and collision energy
voltages applied for the analytes are summarized in Table 6 for
each sulfonamide and tetracycline.
As expected, considerable difference was detected between
SDZ, 2 µg kg−1 ; SNL and SAA, 5 µg kg−1 ; SPZ, 9 µg kg−1 ; and SPY, this unknown component and the 16 SA-s according MS/MS
12 µg kg−1 ), and between 1 and 5 µg kg−1 for the tetracyclines conditions, retention times, and by comparison with spectra
(CTC, 1 µg kg−1 ; OTC and TC, 5 µg kg−1 ). While detection capability in Wiley6N and NIST98 libraries. Additionally, when applying
(CCβ) varies between 1.2 and 16.5 µg kg−1 for the sulfonamides the same chromatographical techniques to other multifloral
(SML, SMP, and SDX, 1.2 µg kg−1 ; STZ, SMR, SMZ, SCP, SDM, SOZ, and unifloral honeys (e.g. eucalyptus, sweet chestnut) free from
and SXZ, 1.7 µg kg−1 ; SGD and SDZ, 3.5 µg kg−1 ; SLN and SAA, residues, no peak was detected. For this reason, five ecological
7 µg kg−1 ; SPZ, 12.6 µg kg−1 ; SPY, 16.5 µg kg−1 ), and between 1.6 heather honey lots free of veterinary treatments and collected
and 7.5 µg kg−1 for the tetracyclines (CTC, 1.6 µg kg−1 ; OTC and in the same area of flowering with an official biocontrol label
TC, 7.5 µg kg−1 ). CCα was 0.08 µg kg−1 , and CCβ was 0.1 µg kg−1 were evaluated and submitted to additional assays. The same
for chloramphenicol. A strong matrix effect was found in honeys response and confirmation were observed in all the batches
of selected unifloral origin during the detection of SA-s by Charm of the ecological heather honey analyzed, characterized by the
assay. It is important to note that the interference was restricted presence of the same unknown peak. The same peak has also
and localized only in heather honeys (dominance required for recently been identified in the nectar extracted of Erica vagans
such classification: >45% pollen Erica vagans). Thus, floral origin (data not published). Apparently, this unknown peak could be
of the samples of this honey were confirmed, all characterized assigned as a natural compound characteristic of this kind of
for the strong presence of Erica vagans (pollen 67–84%; mean honey. This indicates a difference in response between the Charm
74 ± 5%). As mentioned previously, using the Charm II test (with
70

assay and the chromatographical procedures applied on this

www.interscience.wiley.com/jsfa 
c 2008 Society of Chemical Industry J Sci Food Agric 2009; 89: 63–72
Antibiotic and sulfonamide residues in Basque Country honey www.soci.org

contain tetracycline. Finally, CAP was not detected in any of the


Table 6. MRM conditions for detecting sulfonamides and tetracy-
clines by tandem MS samples analyzed.

MRM Cone Collision


transitions voltage energy
Component (m/z) (V) (eV)
CONCLUSIONS
Residues of veterinary drugs were found in a very limited number
Sulfacetamide 215 > 156 17 20 of honey samples. No residues of chloramphenicol were found.
215 > 108 17 33 The Charm test was highly acceptable for detecting tetracyclines
Sulfapyridine 250 > 156 26 24 in honeys. However, it needs correction for the detection of
250 > 108 26 35 sulfonamides in heather (Erica vagans) honeys. The applied
Sulfadiazine 251 > 156 24 22 LC-FD and LC-MS/MS procedures allowed good resolution of
251 > 92 24 38 sulfonamide and tetracycline residues. The precision and accuracy
Sulfathiazole 256 > 156 25 22 of the system and sample preparation validation were acceptable,
256 > 108 25 34 with a high degree of reproducibility. It is therefore concluded
Sulfamerazine 265 > 156 28 25 that either of the two chromatographic methods presented here
265 > 172 28 28 could be used to determine drug residues in honey with a high
Sulfamethoxazole 254 > 156 20 18 degree of reliability. Also both procedures detected an unknown
254 > 108 25 35 natural compound in heather honey with a high affinity by the
Sulfamethizole 271 > 156 20 21 Charm analyzer.
271 > 108 20 35
Sulfamethazine 279 > 186 31 28
279 > 204 31 24 REFERENCES
Sulfadimethoxine 311 > 156 31 29 1 Hansen H and Brodsgaard CJ, American foulbrood: a review of its
311 > 218 31 28 biology. Bee World 80:5–23 (1999).
Sulfanilamide 173 > 156 20 9 2 Otten Ch, A general overview on AFB and EFB pathogen, way of
infection, multiplication, clinical symptoms and outbreak. Apiacta
173 > 92 20 30 38:106–113 (2003).
Sulfaguanidine 215 > 156 20 15 3 Sáenz Z, Zarazaga M, Brinas L, Lantero M, Ruiz-Larrea F and Torres C,
215 > 108 24 18 Antibiotic resistance in Escherichia coli isolates obtained from
Sulfachlorpyrazine 265 > 156 20 22 animals, foods and humans in Spain. Int J Antimicrob Agents
18:353–358 (2001).
285 > 126 20 36
4 Wallner K, Sulfonamide residues in German honey – The actual
Sulfamethoxipyridazine 281 > 156 20 26 situation. Apidologie 34:489 (2003).
281 > 126 20 31 5 EEC, Establishment of Maximum Residue Levels of veterinary medical
Sulfachloropyridazine 285 > 156 20 22 products in foodstuffs of animal origin, Council Regulation (EEC)
285 > 108 20 36 No. 2377/90, Off J Eur Commun L224/1 (1990).
6 EEC, Commision decision 2003/181/EC, March 13, 2003, amending
Sulfadoxine 311 > 156 20 29 decision 2002/657/EC as regards the setting of minimum required
311 > 218 20 28 performance limits (MRPLs) for certain residues in food animal
Sulfisoxazole 268 > 156 15 21 origin. Off J Eur Communities L71, pp. 17–18 (2003).
268 > 113 20 25 7 Serra Bonvehı́ J, Soliva Torrentó M and Muntané Raich J, Invertase
activity in fresh and processed honeys. J Sci Food Agric 80:507–512
Tetracycline 445 > 410 18 12
(2000).
445 > 427 18 18 8 Kaufmann A, Roth S, Ryser B and Widmer M, Quantitative LC/MS-MS
Oxytretracycline 461 > 426 25 17 determination of sulfonamides and some other antibiotics in honey.
461 > 443 25 10 J AOAC Int 85:853–860 (2002).
Chlortetracycline 479 > 444 23 18 9 Hornitzky MAZ, Commercial use of gamma radiation in the
beekeeping industry. Bee World 75:135–142 (1994).
479 > 462 23 16 10 Tetrasensor, Rapid test. Assay for a panel of tetracycline residues in honey.
Unisensor, S.A. (info@unsensor.be), Liège, Belgium (2003).
11 Shet HB and Sporns P, Enzyme immmunoassay for screening of
sulfathiazole in honey. J AOAC 73:871–874 (1990).
12 Masher A, Lavagnoli S and Curatolo M, Determination of residual
honey. Comparative data on biological activities (antibacterial, oxytetracycline in honey with an immunoassay kit. Apidologie
antioxidant, etc.), and full characterization of this compound are 27:229–233 (1996).
required. 13 Heering W, Usleber E, Dietrich R and Märtlbauer E, Immunochemical
As a first step, appropriate correction response has been screening for antimicrobial drug residues in commercial honey.
Analyst 123:2759–2762 (1998).
established in this honey for detecting sulfadrug residues by 14 Salter R, Charm II system – comprehensive residues analysis system
Charm II system. These differences are mainly due to the plant for honey. Apiacta 38:198–206 (2003).
source and to the different chemistry of the honey. This result 15 Edder P and Corvi Cl, Utilisation du Charm II test pour le dépistage
emphasizes that the floral origin and habitats are an important des résidus d’antibiotiques dans les denrées alimentaires d’origine
animal. Mitt Lebensm Hyg 1:218–228 (2001).
factor to be taken into account for the appropriate detection of 16 Barry CP and MacEachern GM, Reverse phase liquid chromatographic
trace residues by screening procedures. Until now, no studies determination of sulfathiazole residues in honey. J AOAC Int 66:4–7
have been performed to find out if other heather honeys of (1983).
different geographical origin would give similar results. Only in 17 Horie M, Saito K, Nose N and Nakazawa H, Simultaneous
determination of sulfonamides in honey by liquid chromatography.
8 of the samples presumptive contaminated with sulfonamides J AOAC Int 75:786–789 (1992).
was the presence of sulfathiazole confirmed, while all 24 samples 18 Oka H and Patterson J, Chemical analysis of tetracyclines. in Chemical
71

presumptive contaminated with antibiotics were confirmed to Analysis for Antibiotics used in Agriculture, ed. by Oka H, Nakazawa N,

J Sci Food Agric 2009; 89: 63–72 


c 2008 Society of Chemical Industry www.interscience.wiley.com/jsfa
www.soci.org J Serra Bonvehı́, A Lacalle Gutiérrez

Harada K and McNeil J-D. AOAC Publ, Gaithersburg, MD, pp. 30 Hormazábal V and Yndestad M, Simultaneous determination of
333–405 (1995). chloramphenicol and ketoprofen in meat and milk and
19 Vinas P, Balsalobre N and Hernández-Córdoba LE, Liquid determi- chloramphenicol in egg, honey, and urine using liquid
nation with ultraviolet absorbance detection for the analysis of chromatography-mass spectyrometry. J Liq Chrom Rel Technol
tetracycline residues in honey. J Chromatogr A 75:786–789 (1992). 24:2477–2486 (2001).
20 Diserens JM and Savoy-Perroud MC, Determination of sulfonamide 31 Yuan J, Jin ZX, Chong YS, Tao D, Hui LC, Bin W, et al, Determination of
residues in honey. Quality & Safety Assurance Department, Nestlé chloramphenicol in royal jelly by liquid chromatography/tandem
Research Center, Lausanne, Switzerland (2002). mass spectrometry. J AOAC Int 89:1432–1436 (2006).
21 Edder P, Cominoli A and Corvi C, Dosage de résidus de sulfamides 32 Louveaux J, Maurizio A and Vorwhol G, Methods for melissopalynol-
dans des denrées alimantaires d’origine animale (foies, rognoms, ogy. Bee World 59:139–157 (1979).
viandes, poissons, oeufs, lait) par HPLC avec prédérivatisation et 33 Vergeron Ph, Interprétation statistique des résultats à matière
détection fluorimétrique. Trav Chim Aliment Hyg 88:554–569 (1997). d’analyse. Ann Abeille 7:384–394 (1964).
22 Kaufmann A, Pacciarelli B, Prijic A, Ryser B and Roth S, Bestimmung 34 Reybroeck W, Residues of antibiotics and sulphonamides in honey on
von rückständen von tetracyclinen in lebensmitteln. Mitt Lebensm the Belgian market. Apiacta 38:23–30 (2003).
Hyg 90:167–176 (1999). 35 Morlot M and Beaune P, An experience with Charm II system. Apiacta
23 Pang GF, Cao YZ, Zhang JJ, Jia GQ, Fan CH, Li XM, et al, Determination 38:226–234 (2003).
of 16 sulfonamides in honey by liquid chromatography/tandem 36 Nagaraja P, Yathirajan HS, Sunitha KR and Vasantha RA, A new,
mass spectrometry. J AOAC Int 88:1304–1311 (2005). sensitive, and rapid spectrophotometric method for the
24 EC, Implementing Council Directive 96/23/EC concerning the determination of sulfa drugs. J AOAC Int 85:869–874 (2002).
performance of analytical methods and the interpretation of results, 37 Cuadros Rodrı́guez C, Garcı́a Campaña AM, Alés Romero F, Jiménez
EC Decision 2002/657, Off J Eur Commun L221/8 (2002). Linares C and Román Ceba M, Validation of an analytical
25 Schwaiger I and Schuch R, Bound sulfathiazole residues in instrumental method by standard addition methodology. J AOAC
honey – need of a hydrolysis step for the analytical determination Int 78:471–476 (1995).
of total sulfathiazole content in honey. Dtsch Lebensm Rundsch 38 Belliardo F, Determination of sulphonamide residues in honey by
96:93–98 (2000). high-pressure liquid chromatography. J Apic Res 20:44–48 (1991).
26 B.O.P.V. (Boletı́n Oficial del Paı́s Vasco) N◦ 207 ZK/1993 (28 October), 39 Bogialli S, Curini R, Corcia A, Laganà A and Rizzuti G, A rapid
Label Vasco de Calidad Alimentaria de la Miel, Dpto. Agricultura y confirmatory method for analyzing tetracycline antibiotics in
Pesca, Vitoria, pp. 9592–9606 (1993). bovine, swine, and poultry muscle tissues: matrix solid-phase
27 Charm Sciences, Charm II 6600/7600 Analyzer Operating Instructions. dispersion with heated water as extractant followed by liquid
Version 2.13. Charm Sciences, Inc., St Lawrence Mass, USA (2001). chromatography-tandem mass spectrometry. J Agric Food Chem
28 McMullen SE, Lansden JA and Schenck FJ, Modifications and 54:1564–1570 (2006).
adaptions of the Charm II rapid antibody assay for chloramphenicol 40 Khong SP, Hammel YA and Guy PhA, Analysis of tetracyclines in
in honey. J Food Protect 67:1533–1536 (2004). honey by high-performance liquid chromatography/tandem mass
29 Pena A, Pelantova N, Lino CM, Silveira MIN and Solich P, Validation of spectrometry. Rapid Commun. Mass Spectrom 19:493–502 (2005).
an analytical methodology for determination of oxytetracycline and
tetracycline residues in honey by HPLC with fluorescence detection.
J Agric Food Chem 53:3784–3788 (2005).
72

www.interscience.wiley.com/jsfa 
c 2008 Society of Chemical Industry J Sci Food Agric 2009; 89: 63–72

Potrebbero piacerti anche