Hdb.ugent.be

J. Agric. Food Chem. 2007, 55, 8359-8366
Validation of the Tetrasensor Honey Test Kit for the
Screening of Tetracyclines in Honey
WIM REYBROECK,*,† SIGRID OOGHE,† HUBERT DE BRABANDER,‡ AND Technology and Food Unit, Institute for Agricultural and Fisheries Research, Brusselsesteenweg 370, 9090 Melle, Belgium, and Faculty of Veterinary Medicine, Laboratory of Chemical Analysis, Ghent University, Salisburylaan 133, 9820 Merelbeke, Belgium Regarding anti-infectious agents, no maximum residue limits are fixed for honey in the Europeanlegislation. Discussions are being conducted in order to set working limits at the European level; forexample, for tetracyclines, 20 µg/kg was proposed. The Tetrasensor Honey test kit is a receptor-based assay using dipsticks for a rapid screening (30 min) of honey on the presence of tetracyclines.
The test was validated according to Commission Decision 2002/657/EC. The test detects tetracycline,oxytetracycline, chlortetracycline, and doxycycline in honey in a specific and sensitive way. Dependingon the type of tetracycline, detection capabilities (CC ) between 6 and 12 µg/kg were obtained (4-7µg/kg for dried dipsticks). The test is rugged and participation with the test in an international ringtrial gave compliant results. It can be concluded that the Tetrasensor Honey test kit is a simple andreliable test that can even be used at the production site.
KEYWORDS: Tetracyclines; honey; screening; rapid tests; residues
INTRODUCTION
chloramphenicol, furazolidone, streptomycin, and tylosin in Honey is generally considered as a natural and healthy honey after administration to bee colonies resulted in similar product. The addition of additives or conserving agents to honey is not allowed. However, in recent years, the problem of residues Reliable screening methods are needed in order to check of antibiotics in honey has been mentioned in some publica- honey for the presence of antibiotics. In general, Charm II tions (1, 2). Antibiotics, for example, tetracyclines, are used in receptor assays are used, but the rate of false-positive results apiculture for the treatment of bacterial brood diseases like could be relatively high (1, 11, 12). A new rapid receptor-based American foulbrood (Paenibacillus larVae subsp. larVae) (3, 4) screening test for the detection of tetracyclines in honey was and European foulbrood (Melissococcus pluton) (5, 6). This developed by Unisensor s.a. (Liège, Belgium), namely, the practice is illegal in Europe. However, oxytetracycline is used in Great Britain in the statutory treatment of European foulbrood The use of antibiotics in apiculture is not authorized in the since this is considered by the authorities as within the cascade European Union. No maximum residue limits are fixed for system for veterinary medicines under minor uses and minor tetracyclines in honey in the European legislation (EEC Regula- species (6). The intensive use of tetracyclines in professional tion 2377/90 and modifications (13)). Some member states beekeeping in the United States and South America resulted in established action limits in order to make the situation more tetracycline-resistant Paenibacillus strains (7, 8).
clear for honey producers, traders, and food inspectors. In Tetracyclines are broad-spectrum bacteriostatic antibiotics Belgium, action limits for residues of antibiotics and sulfona- with a long history in veterinary medicine and are used for the mides in honey were introduced in 2002, taking into account treatment and control of a wide variety of bacterial infections.
analytical possibilities and available toxicological data. During When used in beekeeping, important concentrations up to a the first period of 6 months, the action limit for the group of milligram per kilogram level could be found in the honey of tetracyclines was preliminarily set at 50 µg/kg. Since July 1, the treated hives (6, 9) with a slow depletion and degradation 2002, this value has been fixed at 20 µg/kg. France applies a (a half-life time for oxytetracycline of 9-44 days (6) and a half- nonconformity limit for tetracyclines in honey of 15 µg/kg, the life time of 65 days for tetracycline in honey from supers (9)).
reporting limit in Great Britain (Central Science Laboratory, The research conducted by Adams et al. on the fate of DEFRA, GB) is 50 µg/kg, while the tolerance levels inSwitzerland are 20 µg/kg. Discussions are being held at theEuropean level to set working limits for residues of antibiotics * Corresponding author. Tel.: +32 9 272 30 11. Fax: +32 9 272 30 in honey. The community reference laboratory proposed 20 µg/ 01. E-mail: wim.reybroeck@ilvo.vlaanderen.be.
† Institute for Agricultural and Fisheries Research.
kg as the recommended concentration for the screening of J. Agric. Food Chem., Vol. 55, No. 21, 2007 At the present, it is generally accepted that the screening level for tetracyclines in honey should lie within the range of 10-20µg/kg. To adapt their screening test to this level, Unisensorimproved the sensitivity of the Tetrasensor Honey test kit in2004. All data in this study are based on test kits with animproved detection capability (second generation). The kits ofthis generation carry "Detection limit at 10 µg/kg" on the label.
The aim of this work was to perform a validation study of the Tetrasensor Honey (second generation) screening methodon the basis of validation criteria set in Commission Decision2002/657/EC (15). Since the Tetrasensor Honey is a qualitativescreening kit for tetracyclines, only the following parameters Figure 1. Visual interpretation of Tetrasensor Honey dipsticks.
had to be investigated for validation purposes: the specificityof the test kit, the detection capability, and the test ruggedness For the instrumental reading, the intensity of the color formation is measured, and the result is expressed as the ratio of the color intensityof the test line to the color intensity of the control line. Honey sampleswith a ratio of g 1.40 are free of tetracyclines ("neg"); honey samples MATERIALS AND METHODS
with a ratio of g 0.90 and < 1.40 are slightly contaminated ("low Reagents and Standards. The tetracycline (T3383), oxytetracycline
pos"), and honey samples with a ratio of < 0.90 are more heavily (O5875), chlortetracycline (C4881), doxycycline (D9891), penicillin contaminated ("pos"). When testing honey samples in a routine, samples G (PENNA), cephapirin (C8270), sulfadiazine (S8626), neomycin giving a ratio g 1.40 are considered free from residues of tetracyclines; (N1876), and erythromycin (E6376) were all from Sigma-Aldrich samples giving a ratio < 1.40 are considered suspect for the presence (Bornem, Belgium). The enrofloxacin (17849) was from BioChemika Standard stock solutions of 100 mg/L were made in water and kept below 4 °C. Dilutions of 1 and 0.1 mg/L were freshly prepared on a RESULTS AND DISCUSSION
The most recent EU legislation concerning residue analysis The Tetrasensor Honey kits were from Unisensor s.a. (Liège, (EEC Regulation 2377/90 and modifications (13)) was used as Belgium). In general, lot TH00616-042405/4, expiration date November a guideline for the validation of the method.
23, 2005, was used for the evaluation study; for some parts, such as Stability of Tetracyclines in Honey. Tetracycline is rather
the study of batch-to-batch differences and the stability of the reagents, stable in honey so long as the honey is stored in the dark, since lot TH00624-041907/2, expiration date January 19, 2006, was also used.
The Charm II Tetracyclines Honey kits were from Charm Sciences tetracyclines are light-sensitive. As part of a collaborative trial, Martel et al. implemented a stability study by storing honey A mixture of different honey samples of known (organic) origin with an analyte (tetracycline) for 2 months at 4 °C. No loss of and of different compositions (liquid and solid, flower and honeydew, analyte content could be observed (16). Münstedt et al. spiked Belgian and imported) was used as blank honey. Each honey from the honey with 500 µg/kg of oxytetracycline, chlortetracycline, and blank mixture was tested individually as negative with the Charm II tetracycline. Its high-performance liquid chromatography (HPLC) Tetracyclines Honey (detection capability for tetracycline, oxytetracy- analysis after 10 months of storage at ambient temperatures still cline, chlortetracycline, and doxycycline in honey e 10 µg/kg).
showed more than half of the original concentration of chlor- Material. For the instrumental reading, a QuantiSensor (Matest
tetracycline and tetracycline, but no detectable oxytetracycline, Systemtechnik GmbH, Mössingen, Germany), a small reader device proving an instability of oxytetracycline in honey (17).
with specially designed QuantiSensor software (release 345, version In the study about the false-positive rate for a new Charm II 2003), was used. A QuantiSensor Control dipstick (batch 051307/01,expiration date July 13, 2008) was used daily to check whether the Tetracyclines Honey kit with adapted sensitivity, the incurred samples of the ring trial (11) were retested after 1 year of storage Test Protocol and Interpretation of the Results. For liquid and
in the dark in a cool room by liquid chromatography-mass semisolid honey, there is no sample preparation requested. Solid honey spectrometry/mass spectrometry (LC-MS/MS), and nearly iden- can be made liquid by heating in a glass test tube in a water bath at 37 tical concentrations of tetracycline were measured (data not °C. The lid of the plastic vial is filled with honey so that a correct amount of honey (around 600 mg) is diluted with the buffer content of Test and Reader Repeatability. A blank and four incurrent
the vial (1.8 mL). A total of 200 µL of diluted honey sample is added positive (import table honey) honey samples were analyzed 20 to the lyophilized receptor present in a glass vial and incubated at room times. The color of the test line was evaluated at the end of temperature (20 ( 5 °C) for 15 min. During this first incubation period, each assay (wet dipstick) and a second time after 30 min (dry tetracyclines possibly present in the honey bind with the specificreceptor. After 15 min, the dipstick is dipped into the vial, and a second dipstick). The results were used to calculate the test repeatability incubation at room temperature takes place for 15 min. When the liquid on the basis of wet or dry dipstick readings.
passes through the green capture lines, a red color appears. The first To calculate the repeatability of the reader, only dry dipsticks line captures the remaining active receptor, and the second line takes were measured 20 times since the ratio still shifts slowly during a certain amount of the excess reagent that passed through the first the drying of the strips. This was done at three different levels, line. The second line serves as a control line and always has to become namely, for a blank, a low, and a high positive strip. The results visible; otherwise, the test is invalid. This is shown in figure 1.
are shown in Table 1.
Results were read both visually and using the Quantisensor, The test repeatability was good and even improved as the comparing the color intensity of both capture lines. The visual concentration of tetracyclines in the honey increased. In general, interpretation is as follows: when the color of the test line is more the standard deviations of repeatability decreased when the dry intensive than the color of the control line, the honey sample is negative("vis neg"). In all other cases, the honey is contaminated with dipstick readings were considered in comparison to the wet dipstick tetracyclines ("vis pos"). The visual interpretation is always done before reading, except for the blank honey sample. The reader repeatability the instrumental reading in order to prevent an influence on the judging also improved as the concentration of tetracyclines in the honey increased and lower ratio values were obtained.
Validation of the Tetrasensor Honey Test Kit J. Agric. Food Chem., Vol. 55, No. 21, 2007 Table 1. Test and Reader Repeatability (Wet and Dry Dipstick Reading)
of 10-20 µg/kg in steps of 2 µg/kg. The doped samples wereblind coded before analysis. For each investigated tetracycline, test repeatability (n ) 20) (wet dipstick reading) the lowest concentration giving 19 (low) positive test results on 20 test results was determined. When a certain concentration tested negative two times, we directly tested a higher concentra- tion since 19 positive test results on 20 test results was no longer achievable, in order to save time and reagents.
Since the strips were read both visually and by using a reader system, the detection capability was determined for both means test repeatability (n ) 20) (dry dipstick reading) of strip reading. Moreover, the strips were not only readimmediately (wet dipstick reading) but also after 30 min of drying (dry dipstick reading). The results are shown in Tables
First of all, no differences in detection capability were noticed between the visual and the instrumental reading. Second, the detection capabilities obtained from the dry dipstick readingwere lower than those obtained from the wet dipstick reading.
reader repeatability (n ) 20) (dry dipstick reading) So, by postponing the reading, the sensitivity of the test increased. A summary of the detection capabilities is given in Test Ruggedness. Honey is a complex matrix with a large
variety in composition due to different proportions of the possible sources, nectar and/or honeydew, coming from a great r: standard deviation of repeatability.
variety of plants. So it is important to check the robustness of The consistency in visual judging by the technicians was also the Tetrasensor Honey test kit on different unifloral and checked. It needs to be emphasized that the technicians all received training in the reading of dipsticks as part of the Impact of the Nature (Type, Origin, Physical Parameters, accreditation procedure and that they all had very much etc.) of the Honey on the Test SensitiVity. As a starting point, experience in the color interpretation of analogue dipsticks ( eta- we took the detection capability for tetracycline (TC) of 9 µg/ s.t.a.r. and Tetrasensor Tissue). Real negative and positive honey kg (wet dipstick reading, Table 4), since this concentration is
samples were never wrongly classified by any technician; only just at the top of the dose-response curve. Within the group of very occasionally and only for samples giving a borderline result tetracyclines, tetracycline was the most obvious choice since it (both test lines equal in intensity) was a nonconform result is the most frequently detected tetracycline in honey on the between two different persons obtained (data not shown).
Specificity. The specificity or the ability of the method to
When testing different types of honey, a comparison was run distinguish between the analyte being measured (tetracycline to see whether the same test detection capability was obtained.
residues) and other substances was first investigated by spiking The following types of honey were compared in this study: blank honey in duplo with some other relatively high concentra- Belgian honey versus imported honey (Table 5), blossom honey
tion anti-infectious agents (antibiotics and chemotherapeutics).
versus honeydew honey (Table 6), rape (Brassica spp.) honey
The Tetrasensor Honey test kit was used for the analysis. One (high glucose content) versus black locust (Robinia pseudoa- substance was chosen from each of the most important groups: cacia L.) honey (high fructose content) (Table 7), and solid
penicillin G (penicillins), cephapirin (cephalosporins), sulfadi- honey versus liquid honey (Table 8).
azine (sulfonamides), enrofloxacin (quinolones), neomycin Regarding flower honey, no significant differences were (aminoglycosides), and erythromycin (macrolides); spiking was observed between the Belgian and the imported honey, since performed at 100 times the Belgian action limit for tetracyclines all of the honey samples (spiked with 9 µg/kg TC) gave low- () 2 mg/kg). The color of the test line was evaluated directly positive to positive results. For the Belgian honey samples, this at the end of the assay and after 30 min (dry dipstick).
ratio ranged from 0.58 to 1.15; for the imported honey samples, All honey samples doped with sulfonamides or antibiotics it ranged from 0.67 to 1.27 (both wet dipstick readings).
other than tetracyclines provided negative ratios, and visually Regarding honeydew honey, a difference was observed the results were also all interpreted as "vis neg". In general, between the Belgian and the imported honeydew honey. All following the drying of the dipsticks, the ratio values dropped, Belgian honeydew honey samples spiked with 9 µg/kg TC gave but the results all remained negative.
low positive to positive results, while the Spanish honeydew From the results, it can be concluded that the analysis is not honey sample spiked with 9 µg/kg TC gave a negative result disturbed by anti-infectious agents, which are different from (wet dipstick reading). So, the detection capability of 9 µg/kg tetracyclines. The Tetrasensor Honey kit is very specific for the tetracycline was not valid for the Spanish honeydew honey. It is worth noting that the results became "low pos" once the Detection Capability. Another important validation param-
eter is the detection capability for the most important tetracy- Electrical conductivity could be used for differentiation clines [tetracycline (TC), oxytetracycline (OTC), chlortetracy- between honeydew and blossom honeys (except chestnut honey) since electrical conductivity correlates well with the mineral Therefore, starting from the detection capability concentra- content of honey (18). Regarding the composition criteria for tions obtained from the manufacturer, blank honey was spiked honey (19), blossom honey should have an electrical conductiv- with the investigated tetracycline at different concentrations: ity below 0.8 mS/cm, while the electrical conductivity of in the range of 1-10 µg/kg in steps of 1 µg/kg and in the range honeydew and chestnut honey should be higher than 0.8 mS/ J. Agric. Food Chem., Vol. 55, No. 21, 2007 Table 2. Visual and Instrumental Reading of the Testing of Honey Doped with the Most Important Tetracyclines, Wet Dipstick Reading
Table 3. Visual and Instrumental Reading of the Testing of Honey Doped with the Most Important Tetracyclines, Dry Dipstick Reading (after 30′)
Table 4. Detection Capability (CC ) of the Tetrasensor Honey Test Kit
reading). No differences were observed between the solid honey for the Most Important Tetracyclines: Wet Dipstick Reading and Dry and the liquid honey: all honey samples spiked with 9 µg/kg TC gave low positive to positive results. It is worth noting that,while the liquid honey sample 2, which when undoped, gave detection capability (µg/kg) (visual and instrumental reading) an extremely high ratio of 6.55, a (low) positive result was also obtained when it was doped with 9 µg/kg of tetracycline.
Batch-to-Batch Differences and Reagents' Stability Regarding the Detection Capability. It was examined whether the same detection capabilities for the four most important tetracyclines were obtained when using two completely different batches
(reagents and strips). The results are shown in Tables 9 and
10. Lot B, used shortly after production, gave a mean ratio value
cm. In the validation study, we compared the detection of of 3.98 (wet dipstick reading) and 3.42 (dry dipstick reading) tetracycline in blossom and honeydew honeys. No differences were observed between the blossom honey and the honeydew In the first experiment, doped honey samples were tested on honey: all honey samples spiked with 9 µg/kg TC gave low the same day with two different batches, namely lot TH00616- positive to positive results. For the blossom honey samples, this 042405/4 (A) and lot TH000624-041907/2 (B). It is worth noting ratio ranged from 0.58 to 1.15; for the honeydew honey samples, that lot B was used shortly after production and was marked as it ranged from 0.41 to 1.07 (both wet dipstick reading).
"young", while lot A was already several months old. From Among the European unifloral honeys, rape honey (Brassica Tables 9 and 10, it can be concluded that, in this experiment,
spp.) and black locust honey (Robinia pseudoacacia L.) differ differences in detection capability were obtained for tetracycline, in composition to an extreme extent. Rape honey is light incolor and always comes in a crystallized form (solid). Rape oxytetracycline, and chlortetracycline (wet dipstick reading). The honey contains an average of 40.5 g/100 g glucose and 38.3 detection capabilities claimed in Table 2 were not reached with
g/100 g fructose, and the mean fructose/glucose ratio amounts lot B "young". However, the ratio values were close to the cut- to 0.95. Black locust honey is very light in color and flavor off value of 1.40 (the highest ratio value obtained for the wet with 26.5 g/100 g glucose and 42.7 g/100 g fructose and a dipstick reading is 1.79). All ratio values for doped honey fructose/glucose ratio of 1.61 (20).
samples were far below the ratio values for the blank honey No differences were observed in detection capability (CC ) (wet dipstick reading: lot A, mean ratio ) 4.18; lot B, mean in the examination of rape and black locust honeys, despite the ratio ) 3.98). So the differences in test capability between the serious differences in composition (main type of sugar) and in two different batches remained limited.
In this experiment, both tested batches had a different Honey normally becomes solid due to a natural crystallization production date, so it should be further made clear whether the process caused by the high percentage of sugars, especially small differences are related to a different production or related glucose. In the validation study, the detection in honey of both to a different age at the moment of use. So it was decided to liquid and solid forms was studied. For the solid honey samples, store batch B for more than a year at 4 °C and to retest the the ratios obtained ranged from 0.75 to 1.15; for the liquid honey same concentrations of tetracyclines doped in the same blank samples, the ratios ranged from 0.58 to 1.37 (both wet dipstick honey just before the expiration date of the reagents. The results Validation of the Tetrasensor Honey Test Kit J. Agric. Food Chem., Vol. 55, No. 21, 2007 Table 5. Visual and Instrumental Reading of the Testing of Belgian Honey versus Imported Honeya
a vis: visual reading. neg: negative. pos: positive.
Table 6. Influence of the Botanical Origin of the Honey on the Detection Capability: Blossom Honey versus Honeydew Honey, Both of Belgian Origina
identification conductivity (µS/cm) ratio on t ) 0 vis on t ) 0 ratio on t ) 30′ vis on t ) 30′ ratio on t ) 0 vis on t ) 0 ratio on t ) 30′ vis on t ) 30′ a vis: visual reading. neg: negative. pos: positive.
Table 7. Influence of the Type of the Honey on the Detection Capability: Rape Honey versus Black Locust Honeya
a Vis: visual reading; neg: negative; pos: positive.
of this additional testing with the reagents marked as "old" are the validation by comparing direct (wet dipstick) readings of also summarized in Tables 9 and 10.
the test strips and reading after at least 30 min of drying.
From the stability testing data, we remarked a tendency of a Detailed results are provided in the separate tables. The ratio small improvement of the testing capacity of the reagents during values always decreased when the reading was postponed the shelf life. The reagents of lot B, used just before the (longer time for the color formation and the dipsticks become expiration date, gave results comparable to the results obtained dry); so the detection capability of the test increased by postponing the reading. At the same time, throughout the drying Impact of Drying of the Strips. The impact on the test results of the strips, the color formation at both capture lines became more pronounced, which facilitated visual reading.
J. Agric. Food Chem., Vol. 55, No. 21, 2007 Table 8. Influence of a Physical Parameter of the Honey (Form) on the Detection Capability: Solid Honey versus Liquid Honey, Both from Belgian Origina
a vis: visual reading. neg: negative. pos: positive.
Table 9. Batch-to-Batch Differences and Kit Stability Regarding the
Table 11. Tetrasensor Honey Results of an International Proficiency Test:
Detection Capability, Wet Dipstick Reading Control of False Negative Results of Honey Contaminated Naturally withTetracyclinesa concentration positive/ average lowest highest a vis: visual reading. LC-MS results by Jean-Marc Diserens, Lausanne, CH.
Positively screened samples were sent to an external laboratory for confirmation. Out of the data about the concentrations Lot A: TH00616-042405/4. Lot B: TH000624-041907/2.
confirmed in the positively screened honey samples for tetra- Table 10. Batch-to-batch Differences and Kit Stability Regarding the
cyclines, we have no indication that honeys with tetracyclines Detection Capability, Dry Dipstick Reading above the detection limits found in this validation study weremissed in the screening. Moreover, the concentration determined by LC-MS is sometimes far below the detection limit of the concentration positive/ average lowest highest In 2004, our laboratory used the Tetrasensor Honey test kit in an international proficiency test regarding tetracyclines, organized by P. Beaune (Famille Michaud Apiculteurs, Gan, France). Our Tetrasensor Honey results in this proficiency test (11) are shown in Table 11, with exceptions from the 10 blanks,
which were all found to be negative (no false-positive samples).
The blind-coded positive honey samples in the proficiency test, which were naturally contaminated with 4 µg/kg tetracycline or higher, all gave (low) positive results (wet dipstick reading) for the Tetrasensor Honey test kit. Only one honey sample, naturally contaminated with 3 µg/kg tetracycline, gave a negative result (wet dipstick reading); however, when we read the dipstick after 30 min, the same sample already gave a low positive result.
In 2005, our laboratory also participated in another interna- a Lot A: TH00616-042405/4. Lot B: TH000624-041907/2.
tional collaborative trial on antibiotic residues in honey,organized by the Laboratoire d'Etudes et de Recherches sur les Test for False NegatiVe/False PositiVe Results. To investigate Petits Ruminants et les Abeilles de l'AFSSA (Sophia Antipolis, the possibility of false-negative results, naturally incurred honey France) (16). Three samples contained no residues above the samples from our collection were retested using the Tetrasensor limit of detection of the LC-MS reference analysis. These Honey test kit. All samples with a known concentration above samples all tested negative for Tetrasensor Honey. In this the detection capability also gave positive screening results when collaborative trial, sample 6, containing 8.7 µg/kg tetracycline, using the Tetrasensor Honey test kit.
yielded a negative result directly at the end of the test (wet The test, BELAC accredited since the end of 2004, has also reading). However, the result became "low positive" after half been used at T&V-ILVO routinely over the past 3 years.
an hour, following a dry dipstick reading. The other positive Validation of the Tetrasensor Honey Test Kit J. Agric. Food Chem., Vol. 55, No. 21, 2007 samples with concentrations of tetracycline ranging from 19.8 Hullebusch and Veronique Ottoy for practical assistance during to 31.7 µg/kg (LC-MS) were all detected as positive even during wet reading. Five out of seven laboratories using high-performance liquid chromatography with a diode array detectorreported sample 6 as negative; of the 22 laboratories using LC- LITERATURE CITED
MS, two laboratories reported sample 6 as not being detectedand one laboratory as <5 µg/kg.
(1) Reybroeck, W. Residues of antibiotics and sulphonamides in honey So in both trials, no false-negative or false-positive results on the Belgian market. Apiacta 2003, 38, 23-30.
(2) Bogdanov, S. Contaminants of bee products. Apidologie 2006,
were obtained with the Tetrasensor Honey test kit. In the 2004 proficiency test, even concentrations of tetracycline below the (3) Hopingarner, R.; Nelson, K. American foulbrood cleanup rate detection capability of 9 µg/kg yielded "low positive" results using three terramycin treatments. Am. Bee J. 1987, 128, 120-
with the Tetrasensor Honey test kit. This could be explained by the dose-response results as shown in Tables 2 and 3:
(4) Spivak, M. Preventative antibiotic treatments for honey bees. Am. concentrations just below the detection capability could some- Bee J. 2000, 140, 867-868.
times result in a positive result. Another explanation could also (5) Waite, R. J.; Brown, M. A.; Thompson, H. M.; Brew, M. H.
be the time delay between the Tetrasensor analysis and the Controlling European foulbrood with the shook swarm method physicochemical confirmation, which possibly resulted in a and oxytetracycline in the UK. Bee World 2003, 82, 130-138.
degradation of incurred tetracycline in honey to degradation (6) Thompson, H. M.; Waite, R. J.; Wilkins, S.; Brown, M. A.; products with sterical similarity to the parent compound (17).
Bigwood, T.; Shaw, M.; Ridgway, C.; Sharman, M. Effects of Finally, it is worth noting that the Tetrasensor Honey test kit European foulbrood treatment regime on oxytetracycline levels detects not only the parent compound but also the epimers.
in honey extracted from treated honeybee (Apis mellifera) colonies If the data of the screening of 100 table honeys from different and toxicity to brood. Food. Addit. Contam. 2005, 22, 573-578.
countries are considered (21), no false-positive or false-negative (7) Alippi, A. M. Is Terramycin losing its effectiveness against AFB.
results were obtained when the Tetrasensor Honey test kit was Bee Biz 2000, 11, 27-29.
used as screening method, whereas a rate of 8% false-positive (8) Miyagi, T.; Peng, C. Y. S.; Chuang, R. Y.; Mussen, E. C.; Spivak, results was obtained for the same honey when the Charm II M. S.; Doi, R. H. Verification of oxytetracycline-resistant Tetracyclines Honey was used. In this study, two samples with American foulbrood pathogen Paenibacillus larVae in the United a tetracycline concentration below the detection capability tested States. J. InVertebr. Pathol. 2000, 75, 95-96.
positive for Tetrasensor Honey; the presence of tetracyclines (9) Martel, A. C.; Zeggane, S.; Drajnudel, P.; Faucon, J. P.; Aubert, in both samples (4 and 6 µg tetracycline/kg) was confirmed by M. Tetracycline residues in honey after hive treatment. Food. Addit. Contam. 2006, 23, 265-273.
From this work, it can be concluded that Tetrasensor Honey (10) Adams, S.; Heinrich, K.; Caldow, M.; Sharma, A.; Ashwin, H.; is a suitable test kit for the screening of tetracyclines in honey.
Homer, V.; Stead, S.; Fusell, R.; Kelly, M.; Wilkins, S.; The test takes 30 min. In general, no differences were noticed Thompson, H.; Sharman, M. Investigation of the fate of veterinary between a visual and an instrumental reading of the dipsticks.
drugs used in apiculture. Proceedings of the 5th International So a reader system is not required for screening purposes. Since Symposium on Hormone and Veterinary Drug Residue Analysis, no special equipment (incubator, reader, etc.) is required, the Antwerp, Belgium, May 16-19, 2006; pp 27-27.
test can be performed at the production site even by the (11) Beaune, P.; Garcet, C. Report Charm II Tetracycline Ring Trial; 2004; pp 1-11.
(12) Beaune, P.; Diserens, J. M.; Reybroeck, W. Proficiency Testing The test detects the tetracyclines in honey in a specific and of Charm II Tests for Residue Control of honey. Proceedings of sensitive way. Depending on the type of tetracycline concerned, the 39th Apimondia International Agricultural Congress,Dublin, a detection capability between 6 and 12 µg/kg was obtained Ireland, August 21-26, 2005; pp 33-34.
directly after the second incubation period. When reading dry (13) Council Regulation (EEC) No 2377/90 laying down a Community dipsticks, detection capabilities of between 4 to 7 µg/kg were procedure for the establishment of maximum residue limits of obtained, so when the dipstick becomes dry, the detection veterinary medicinal products in foodstuffs of animal origin. Off. J. Eur. Communities: Legis. 1990, L224, 1-8.
The test procedure is very simple, and the test is rugged. No (14) CRLs view on state of the art analytical methods for the national influence on the test capability was noticed, with regard to the residue plans for control of residues. Discussion paper residue geographical or botanical origin or by physical parameters (solid expert meeting, Brussels; 2007; pp 1-15.
versus liquid). Only small problems were encountered with a (15) Commission Decision (EC) N° 2002/657 implementing Council Directive 96/23/EC concerning the performance of analytical We noticed differences in the test capability between two methods and the interpretation of results. Off. J. Eur. Communities: different batches. However, the differences remained limited, Legis. 2002, L221, 8-36.
and when the tests were repeated 16 months later using the (16) Martel, A. C.; Zeggane, S.; Halimi, C. Report Results of second test kit, significant differences were no longer observed.
CollaboratiVe Trial on Antibiotic Residues in Honey-2005; A stability study showed a slight increase of the test sensitivity International Honey Commission: Posieux, Switzerland, 2005; pp No false negative and no false positive results were obtained (17) Münstedt, T.; Rademacher, E.; Petz, M. Chlortetracycline and during two international proficiency tests and a study of 100 oxytetracycline residues in honey after administration to honey- bees. Proceedings of the 4th International Symposium on Hormoneand Veterinary Drug Residue Analysis; Antwerp, Belgium, June4-7, 2002; pp 147-147.
ACKNOWLEDGMENT
(18) Bogdanov, S.; Ruoff, K.; Persano Oddo, L. Physico-chemical The authors thank the company Unisensor s.a. for kindly methods for the characterization of unifloral honeys: a review.
supplying the test kits. The authors wish to thank Kurt Apidologie 2004, 35, 4-17.
J. Agric. Food Chem., Vol. 55, No. 21, 2007 (19) Council Directive 2001/110/EC of 20 December 2001 relating to EurBee 2006, Prague, Czech Republic, September 10-14, 2006; honey. Off. J. Eur. Communities: Legis. 2002, L10, 47-52.
(20) Persano Oddo, L.; Piro, R. Main European unifloral honeys: descriptive sheets. Apidologie 2004, 35, 38-81.
Received for review June 28, 2007. Revised manuscript received July
(21) Reybroeck, W.; Ooghe, S.; Daeseleire, E. Presence of antibiotics 30, 2007. Accepted July 30, 2007.
and sulfonamides in honey and royal jelly on the European market.
Proceedings of the Second European Conference of Apidology

Source: http://www.hdb.ugent.be/HDB/Publications_files/P2007-10%20(WR).pdf