Tfac101224 564.57

Food Additives and Contaminants, Vol. 21, No. 6 (June 2004), pp. 564–571 Contamination of honey by the herbicide asulam and itsantibacterial active metabolite sulfanilamide Sulfathiazole is known to be effective against the a Official Food Control Authority of the Canton of Zurich American foulbrood (Bacillus larvae) (Haseman (Kantonales Laboratorium Zu¨rich), PO Box, CH-8030 Zu¨rich, 1946). A number of other sulfonamides like sulfa- methoxazole and sulfamdimethoxine are also used for Official Food Control Authority of the Canton of Aargau (Kantonales Laboratorium Aargau), Kunsthausweg 24, CH-5000 the same purpose. However, the appearance of resist- ance against some antibacterially active substancesled to the belief that these drugs should be avoidedas much as possible in animal husbandry (Wille 1967).
(Received 8 August 2003; revised 10 February 2004; accepted17 February 2004) In Switzerland, the treatment of bees with antibioticshas not been permitted since 1974 (Bogdanov andFluri 2000, Zentrum fu¨r Bienenforschung 2002).
A number of antibacterial drugs (antibiotics) like Analytical controls were initiated to monitor the sulfonamides, tetracyclines and streptomycin are used for the treatment of bacterial diseases in beehives. Yet,the finding of sulfanilamide residues in some 15 Swiss Honey is a rather complex matrix and the analysis honeys out of some 350 samples could not be explained of veterinary residues by classical liquid chromatog- by such apicultural practice. Bees occasionally collect raphy with ultraviolet light/vis or fluorescence detec- nectar from meadows treated with the herbicide tion methods might be affected by false-positive asulam. Such honey is not only contaminated by findings. It was the use of liquid chromatography asulam, but also by its degradation product sulfanil- coupled with tandem mass spectrometry detection amide. This is the first report that the use of a herbicide (LC-MS/MS) as an analytical tool that permitted causes the appearance of residues of an antibacterial the unambiguous detection and confirmation of a active metabolite belonging to the category of sulfon- number of antibacterial drug residues (Kaufmann amide drugs in food. The relevance of this finding lies in the fact that the use of the herbicide asulam might cause In relation to sulfonamide residues, there were three unacceptable residue levels of sulfanilamide in a prod- relevant sulfonamide findings among the 350 Swiss honey samples analysed in the present authors’ . Group of beehives with a weakened bee popula- tion was treated with sulfathiazole. Bees fromneighbouring hives were able to tap the honeyfrom this weak bee population as a food source.
This phenomenon, termed ‘robbery’, can also beobserved among healthy bee populations. However,the practice significantly increases when a bee Apicultural use of sulfonamides for treatment population weakens or even abandons a beehive.
Hence, depending on the distance, honey fromother beehives showed significant residue levels of Sulfonamides are used against a number of bacterial sulfathiazole (Seiler and Kaufmann 2002).
diseases affecting bees (Frey and Lo¨scher 1996).
. Contamination of a honey with sulfathiazole from another region could be linked to the winter feeding * To whom correspondence should be addressed.
of bees with honey of rather dubious origin. This feeding honey was most likely very old and no Food Additives and Contaminants ISSN 0265–203X print/ISSN 1464–5122 online # 2004 Taylor & Francis Ltd Contamination of honey by asulam and sulfanilamide longer fit for human consumption. It contained might be caused by robbery or winter feeding with some 8000 mg kgÀ1 sulfathiazole, making it a likely contaminated honey/sugar as described above (Seiler and Kaufmann 2002). There exists the possibility . Several honey samples contained sulfanilamide, that environmental pollution, e.g. contamination an exceptional finding that could not be explained caused by agricultural use of pesticides in different cultivations, is carried into the hive by the beesthemselves (Fernandez and Muino 1995, Koch andWeiber 1997).
An accumulation of several sulfanilamide-positivehoney samples originating from different beekeepers The herbicide asulam (methyl-4-sulfanylcarbamate) in the Canton (Province) of Aargau were the startingpoint for comprehensive investigations. As a result of in Switzerland is approved to control the growth of preliminary research, the present authors stated the a variety of broadleaf weeds (Rumex sp., Dryopteris hypothesis that the sulfanilamide residues in honey sp., Pteridium aquilinium) in meadows, pasture, pome probably originated from the metabolic pathway of and stone fruit orchards. It is used in springtime the herbicide asulam and were brought in the bee- (April/May) in significant quantities (1–3 kg haÀ1).
hives from the surroundings by the bees themselves Asulam is known to be degraded into sulfanilamide (Jahresbericht Kantonales Laboratorium Aargau (4-aminobenzenesulfonamide; figure 1) and further metabolites (US Environmental Agency 1995). A soilFlavobacterium sp. was reported to degrade asulam The present study investigates the finding of sulfanil- to sulfanilamide (Walker 1978, Allan and Millward amide residues in Swiss honey and attempts to eluci- 1984). This bacterium can grow on asulam and date the question whether sulfanilamide is illegally sulfanilamide as they act as a carbon and energy used as an antibiotic to fight bacterial bee diseases source (Walker 1978). The time to degrade 50% or if the sulfanilamide residue levels in honey are of the original asulam concentration in treated indirectly due to the agricultural application of soil was less than 7 days (Allan and Millward 1984).
asulam and the honeybees are the crucial vector of Only 2.5% of asulam remained after 15 days (Suzuki and Yaguchi 2001). This degradation pro-ceeds rapidly in topsoil, although adding yeast extractenhanced degradation (Babiker and Duncan 1977).
Besides enzymes, iron(III) aqua-complexes were shown to act as photocatalysts in the degradationprocess of asulam to sulfanilamide (Castastini andSarkha 2002).
The following chemicals were obtained commercially: sulfanilamide (Fluka, Buchs, Switzerland); asulam(Promochem, Wesel, Germany); citric acid mono- Residues of active substances in food like honey hydrate and formic acid 98% (Merck, Darmstadt, not only might be caused by the direct use of such Germany); hydrochloric acid 37%; ammonia 25% drugs by bee-keepers who treat their bees, but also and acetonitrile (Scharlau, Barcelona, Spain).
Figure 1. Degradation of asulam into sulfanilamide.
The materials used were as follows: centrifuge tubes, 0% B. Mobile phase A: 50 ml acetonitrile and 3 ml 250 ml, fluted filter paper (Schleicher & Schu¨ll, formic acid were added into a 1000 ml graduated Dassel, Germany); solid-phase extraction (SPE) flask and filled up to the mark with distilled water.
columns, OASIS HLB 200 mg (Waters Milford, Mobile Phase B: 3 ml formic acid were added into a MA, USA); solid-phase extraction unit, Visiprep 1000-ml graduated flask and filled up to the mark (Supelco, Bellefonte, PA, USA); vortex mixer, vari- able speed; magnetic stirrer, variable speed; pH meter, 0.2 ml minÀ1 and the injection volume was 10 ml.
Metrohm 691 (Herisau, Switzerland); rotary evapora- MS parameters were as follows: capillary voltage, tor, Bu¨chi RE 120 (Flawil, Switzerland).
3.25 kV; extractor voltage, 3 V; source temperature,90C; desolvation temperature, 250C; cone gas flow, 50 l hÀ1; desolvation gas flow, 560 l hÀ1 (nitro-gen); nebulizer gas flow, factory preset value; collision Honey (7.5 g) was weighed into a centrifugation tube cell pressure, 2 Â 10À3 mbar (Argon); multiplier volt- (the spike solution was added if recoveries were to age, 650 V. Compound specific parameters were be determined). It was dissolved in 15 ml 2 mol lÀ1 as follows: asulam: 231>156 transition by applying hydrochloric acid and left for 30 min at room tem- 10 eV collision energy and 32 V cone potential; sulfa- perature. Citric acid solution (30 ml 0.3 mol lÀ1) was nilamide: 173>93 by applying 30 eV collision energy added, mixed and filtered. An SPE cartridge was moistened with 3 ml acetonitrile and rinsed twice Confirmation was based on the monitoring of at with 2 ml distilled water. Honey filtrate (20 ml) was least two MS/MS transition ratios. Sulfanilamide: taken and the pH adjusted to 3.5–4.5 with ammonia.
173>156; cone, 30 V; collision energy: 9 eV; asulam: The next step was to proceed immediately. The neutralized honey solution was transferred into the A number of samples were analysed by a pseudo- reservoir above the SPE cartridge and the solution MS/MS/MS approach (cone-induced fragmentation).
sucked through the cartridge within 10–15 min. Thecartridge was rinsed three times with 3 ml distilledwater and allowed to run dry for about 4 min. The cartridge was eluted with 3 ml acetonitrile into asmall, previously weighted conical flask. The solutionwas evaporated in a rotary evaporator (40C) to a The method underwent validation. Sulfanilamide was small volume. Mobile phase A (0.5 ml) was added validated in three different honey matrices (six levels, to the remaining liquid. It was mixed with a vortex three repetition each), while asulam was validated mixer and the conical flask was weighed. The extract only for one honey matrix (blossom honey).
was transferred without filtration into a high- r2 were between 0.984 and 0.994. The recoveries performance liquid chromatography vial.
(not corrected for signal suppression effects) variedbetween 45 and 55%. The MS signal was observed to Liquid chromatography coupled with tandem mass be suppressed by the matrix to a level between 54 and 78% of the sulfanilamide peak area obtainedby injecting a pure standard. Hence, recoveriescorrected for signal suppression were found between LC-MS/MS was performed by means of a type 70 and 83%. The r2 for asulam was 0.989; recovery Agilent Model 1100, binary pump, autosampler (not corrected for signal suppression effects) was (Agilent, Waldbronn, Germany) coupled to an elec- 61%. MS signal suppression was 85%. Hence, the trospray ionization (ESI) coupled with tandem mass recovery corrected for signal suppression was 72%.
spectrometry detection, Quattro LCZ with electro- The limit of detection was 1 mg kgÀ1 (sulfanilamide) spray interface and MassLynx software (Micromass, and 0.8 mg kgÀ1 (asulam). These limits were calculated Manchester, UK) using a Nucleosil 100-5, C18 HD, based on a 3:1 s/n ratio. Intra- and interday preci- 50 Â 2 mm, 5 mm plus guard column (Macherey- sions were 6.2 and 9.8% relative standard deviation (RSD) for sulfanilamide and 5.4 and 10.2% RSD for A linear gradient was employed: 0–10 min: 0–30% B; asulam (referring to a blossom honey sample spiked 10–12 min 30% B; 12–12.1 min: 30–0% B; 12.1–19 min Contamination of honey by asulam and sulfanilamide origin of the contaminated samples. The dismayedbeekeepers clearly denied the use of any drugs in their Some 350 Swiss honey samples were officially collected from either retail trade or directly from If sulfanilamide was used for the treatment of bees, apiculturists. Among the 15 sulfanilamide contami- honey samples from these producers should show nated honey samples, four corresponding beekeepers very high antibacterially effective concentrations of were selected for further investigations, in the course sulfanilamide. However, unlike the mentioned cases of which additional samples were obtained from them concerning sulfathiazole with measured levels up to the following year, consisting of honey collected by above 10 mg kgÀ1 (Seiler and Kaufmann 2002), only the bees particularly in spring (April/May) and in small amounts of sulfanilamide in the mg kgÀ1 range were found. Such levels are more easily explained All samples were analysed by the official Food by unintentional contamination than by a prohibited Control Authority of the Canton of Zu¨rich by the LC-MS/MS method described above. The analytical Further inquires by the Official Food Control method permitted the detection of 16 different sulfo- Authority of the Canton of Aargau, focused on other namides and three tetracyclines. Additional technical possible sources of sulfanilamide, finally referred details have been published by Kaufmann and to the following theory as a ‘working hypothesis’ Guggisberg (2002). Positive samples were confirmed (Jahresbericht Kantonales Laboratorium Aargau by monitoring the ratio of at least two analyte-specific Sulfanilamide contamination in honey might be closelyinterrelated to the agricultural use of the herbicideasulam, because the active agent is known to bedegraded into sulfanilamide and asulam-containing products are usually applied in cultures and plantationsforaged by honeybees.
Consequently, this hypothesis was tested by analysing Fifteen honey samples officially collected from the sulfanilamide-positive honey samples for possible local market contained measurable concentrations asulam residues. It was shown that the described of sulfanilamide in the range 3–227 mg kgÀ1, whereby sample preparation used for the determination of in four cases, the sulfanilamide content was above sulfonamides in honey could be used as well for the the maximum permitted residue level of 50 mg kgÀ1, sample clean up of asulam. The substance survived according to the ordinance on foreign substances the hydrolysis step and was also concentrated effi- and constituents in foods (Swiss Federal Office of ciently on the SPE material. Therefore, all samples Public Health 2002). Figure 2 shows a typical chro- containing sulfanilamide residues were (re)analysed matogram of a honey containing sulfanilamide and In all the 15 sulfanilamide-positive honey sam- In the beginning, the positive sulfanilamide findings ples, asulam was detected in concentrations of were not easy to explain. Initially, sulfanilamide 1–200 mg kgÀ1 (table 1). One sample contained asulam was the first discovered drug belonging to the group residue levels reaching the provisional maximum of sulfonamides. Chemical modifications of the side residue level of 200 mg kgÀ1, which was fixed by the chain of the molecule produced a large number of Swiss Federal Office of Public Health (January 2002) derivatives exhibiting clearly stronger antibacterial after discovering the unexpected asulam contami- activities. Hence, it was not evident why an ‘old’, nation in honey. These results showed a clear correla- not very potent drug, which is therefore not anymore tion between sulfanilamide and asulam content easily obtainable, should be used in apiculture.
(figure 3). Note that among the 350 honeys from Since a number of sulfanilamide-positive honey the local market, there were no samples where either samples originated from the Canton Aargau, the only asulam or only sulfanilamide was detected.
local Food Control Authority questioned the apicul- Hence, there is a clear relationship between these turists involved and investigated in detail the apicul- two substances. All positive honeys originated from tural and agricultural situation surrounding the locations where spring flowers in meadows and pas- Figure 2. Asulam and sulfanilamide NRM traces of a standard solution and a contaminated honey sample.
tures for forage production represented a major With the additional samples from affected bee- keepers, it was discernible that honey collected bythe bees in spring (April/May) contained substantially higher levels of sulfanilamide and asulam than honey from the same beehives subsequently collected in mid-summer (June/July). Typical examples are summarized in table 2. This finding corresponds to the period where meadows and pastures are treated The use of asulam for the treatment of integrated production cultures has to be sanctioned by the Official Center of Crop Protection. Yet, no such application has been received from farmers in the neighbourhood of the contaminated beehives. How- ever, it is possible that non-integrated produced meadows were treated with asulam or that asualm Contamination of honey by asulam and sulfanilamide Figure 3. Asulam/sulfanilamide relationship among sulfanilamide-positive honey samples.
was used without official permission. Nevertheless, of up to five different sulfonamides. However, the the main proposition of the present working hypoth- present authors could never detect asulam/sulfanil- esis was substantiated by the preliminary result of a amide in foreign honey. This might be explained project initiated by the Swiss Bee Research Centre by the nature of the final product. Swiss honey is following the unexpected and mysterious sulfonamide a high-priced product, which is commonly sold contamination in Swiss honey. Experimentally con- directly by the apiculturist and it is rather seldom trolled application of asulam in pasture in springtime that such honey is blended with honey from other (April) caused analytically measurable sulfanilamide producers. On the other hand, imported honey and asulam residues in nearly all blossom honey sam- mostly comprises mixtures consisting of products ples harvested in neighbouring beehives (Bogdanov from different producers, possibly even from differ- ent countries or continents. Hence, the blending ofa few strongly contaminated honeys with a majority More than 50% of all foreign honey sold in local of uncontaminated honeys may produce products markets contained traces of at least one antibacter- with a number of different antibiotic residues. The ial drug (Jahresbericht Kantonales Labor Zu¨rich possible low level contamination of some honeys 2001a). Some samples contained small amounts Sulfanilamide and asulam content in honey The results and findings of these investigations are in agreement with the present main hypothesis: thatsulfanilamide contamination in honey might be closely related to the agricultural use of the herbi- cide asulam in springtime. This rather unexpected finding should be investigated and clarified in detail by means of further experimental studies and pro- jects. In Switzerland, as a consequence of these find- ings has been that the approval of asulam for agricultural use, especially with regard to the treat-ment area, has to be modified by the authorizingadministration. For instance, restricting the use ofasulam to application in autumn would probably by sulfanilamide and/or asulam is therefore most resolve the problem to do with the contamination of likely diluted to an extent to make them undetect- honey. However, asulam is used world wide in vari- able by present analytical capabilities. On the other ous geographical environments to treat different hand, in respect of botanical origin a considerable ‘crops’, ranging from potato and sugarcane to number of foreign products available in Switzerland Christmas tree cultures. Therefore, different circum- are specialty honeys, e.g. acacia, chestnut, lemon, stances might require appropriate precautions and lavender or forest trees. These trees/flowers are measures to prevent a contamination of the food probably less likely to be treated with asulam than chain by its metabolite sulfanilamide.
There also is the possibility that the source of sulfa-nilamide is related to the manure of animals (pig/ cows) treated with sulfonamides. This vector cannotbe completely ruled out; however, it is consideredto be rather unlikely. The Official Food Control Authority of the Canton of Zu¨rich screens yearly an Mueller, Official Center of Crop Protection of average of about 800 cow/pig urine samples for the the Canton Aargau, concerning the agricultural presence of different veterinary drug residues. Positive use of asulam are greatly appreciated. The authors results were repeatedly found for tetracyclines and the gratefully acknowledge the helpfulness, courtesy sulfonamides. Sulfadimidin and sulfathiazole residues and patience of the affected beekeepers during the were rather common, but sulfanilamide was never detected (Jahresbericht Kantonales Labor Zu¨rich2001b). Although in some cases total sulfonamideconcentrations up to 20 mg kgÀ1 liquid manure could be found on farms where medicinal feed had beenapplied (Haller et al. 2002), the amount of sulfa drugs Allan, E., and Millward, L., 1984, Sulfanilamide as a minor soil spread by contaminated manure on meadows is most degradation product of asulam. 11th Workshop on Chemistry likely significantly below the sulfanilamide quantities and Biochemistry of Herbicides, pp. 107–112.
on meadows resulting from the degradation asulam Babiker, A., and Duncan, H., 1977, Influence of soil depth on asulam adsorption and degradation. Soil Biology and Biochemistry, 9, sprayed in the scale of 1–3 kg haÀ1. Moreover, manure is spread almost throughout the year on mowed Bogdanov, S., 2003, Projekt ‘Antibiotikaru¨cksta¨nde in Honig aus meadows, in contrast to asulam which is applied in landwirtschaftlichen Verschmutzungsquellen’, Stand Februar2003: Situationsanalyse. Internal Report (Bern: Swiss Bee April/May shortly before the flowering of the pasture- land or after the end of August. The preferential appearance of sulfanilamide in spring blossom honey Antibiotikaru¨cksta¨nde. Schweizerische Bienenzeitung-Z., 123,407–410.
could therefore not be explained by manure as the Castastini, C., and Sarkha, M., 2002, Iron (III) aquacomplexes source of sulfanilamide contamination.
as effective photocatalysts for the degradation of pesticides Contamination of honey by asulam and sulfanilamide in homogeneous aqueous solutions. Science of the Total Foreign Substances and Constituents in Foods [Verordnung Fernandez-Muino, M. A., Sancho, M. T., Munategui, S., 817.021.23 (Bern: Swiss Federal Office of Public Health) [avail- Huidobro, J. F., Simal-Lozano, J., 1995, Nonacaricide pesti- able at: http://www.admin.ch/ch/d/sr/c817_021_23.html).
cide residues in Honey: Analytical methods and levels found.
Seiler, K., and Kaufmann, A., 2002, Kontamination von Honig mit Journal of Food Protection, 58, 1271–1274.
Sulfathiazol durch Ra¨uberei unter Bienen. Mitteilungen aus Frey, H.-H., and Lo«scher, W., 1996, Lehrbuch der Pharmakologie Lebensmitteluntersuchung und Hygiene Lebensm. Hyg., 93, und Toxikologie fu¨r die Veterina¨rmedizin (Stuttgart: Erdinand Suzuki, T., and Yaguchi, K., 2001, In vitro pesticide degradation Haller, M. Y., Mu«ller, S. R., McArdell, C., Alder, C., Suter, in turfgrass soil incubated under open and sealed conditions.
M. J.-F., 2002, Quantification of veterinary antibiotics (sulfo- Journal of Environmental Quality, 30, 18–23.
namides and trimethoprim) in animal manure by liquid chro- US Environmental Protection Agency, Office of Prevention, matography-mass spectrometry. Journal of Chromatography A, Pesticides and Toxic Substances, 1995, Reregistration Haseman, L., 1946, American foulbrood. Journal of Economic Ento- (Washington, DC: US EPA) (available at: http://www.epa.
Jahresbericht des Kantonalen Laboratoriums Aargau, 2000, 61–62.
Walker, N., 1978, A soil Flavobacterium sp. that degrades sulpha- Jahresbericht des Kantonalen Laboratoriums Aargau, 2001, 63.
nilamide and Asulam. Journal of Applied Bacteriology, 45, Jahresbericht des Kantonalen Laboratoriums Zu¨rich, 2001, 37–38.
Jahresbericht des Kantonalen Laboratoriums Zu¨rich, 2001, 53–54.
Wille, H., 1967, Was ist von der Sanierung der bo¨sartigen Faulbrut Kaufmann, A., and Guggisberg, D, 2002, Quantitative LC/MS-MS mit Heilmitteln zu halten? Schweizerische Bienenzeitung-Z., determination of sulfonamides and some other antibiotics in honey. Journal of the AOAC, 85, 853–860.
Koch, H., and Weiber, P., 1997, Exposure of honeybees during Milchwirtschaft Liebefeld, 2002, Richtlinien zur pesticide application under field conditions. Apidologie, 28, Beka¨mpfung der Bienenkrankheiten, Ausgabe October (Bern: Zentrum fu¨r Bienenforschung, Eidg. Forschungsanstalt fu¨r Swiss Federal Office of Public Health, 2002, Ordinance on

Source: https://www.ag.ch/media/kanton_aargau/dgs/dokumente_4/verbraucherschutz_1/lebensmittelkontrolle/lebensmitteluntersuchungen/Contamination_of_honey_by_the_herbicide_asulam_and_its_antibacterial_active_metabolite_sulfanilamide.pdf