twitter

Wednesday, 6 July 2016

Food Chemistry
Volume 178, 1 July 2015, Pages 156–163
Analytical Methods

Rapid immunochemical analysis of the sulfonamide-sugar conjugated fraction of antibiotic contaminated honey samples

Dedicated to the memories of Francisco Sanchez-Baeza and Alejandro Muriano.
  • a Nanobiotechnology for Diagnostics group (Nb4D), IQAC-CSIC, Spain
  • b CIBER de Bioingenieria, Biotemateriales y Nanomedicina (CIBER-BBN), Barcelona, Spain1
  • c UNISENSOR, Liege Wandre, Belgium2
  • d Nestle Research Center, Lausanne, Switzerland3
Received 7 August 2013, Revised 21 October 2014, Accepted 3 January 2015, Available online 13 January 2015
doi:10.1016/j.foodchem.2015.01.037
Get rights and content

Highlights

•
Rapid immunochemical screening for sulfonamide contaminated honey analysis.
•
Sulfonamides conjugated to sugars have a very low SA immunoreactivity (<5%).
•
Sulfonamide antibiotics must be first released before the analysis.
•
The whole analytical procedure is performed in less than 2 h for more than 100.
•
Detectability below action limits established in EU countries for nine sulfonamides.

Abstract

A rapid high-throughput immunochemical screening (HtiS) procedure for the analysis of the sulfonamide (SA)-sugar conjugated fraction of antibiotic contaminated honey samples has been developed. Studies performed with this matrix have indicated that sulfonamide antibiotics are conjugated to sugars rapidly and quantitatively, providing samples with very low SA immunoreactivity. Therefore, sulfonamides must be first released before the analysis, and for this purpose, a simple and fast sample preparation procedure has been established consisting of hydrolyzing the sample for 5 min, adjusting the pH and buffering the sample prior to the immunochemical analysis. Under these conditions, honey samples could be directly analyzed without any additional sample treatment, other than dilution. Recovery values of the whole analytical procedure were greater than 85%. The analysis of the same samples without the hydrolysis provided recovery values below 5%. Selectivity studies performed in hydrolyzed honey samples revealed that nine relevant sulfonamide antibiotics can be detected with limit of detection (LOD) values below the action limits established by some EU countries (Belgium, 20 μg kg−1, United Kingdom or Switzerland, 50 μg kg−1).

Keywords

  • Sulfonamide antibiotics;
  • Immunoassay;
  • Magnetic particles;
  • Honey;
  • Sulfonamide-sugar conjugates;
  • Hydrolysis;
  • Class-selective antibodies

1. Introduction

Honey has been traditionally considered a natural and healthy product. However, recently, the presence of antibiotic residues in this nutrient has been reported (Bogdanov, 2005). Contamination of natural honey with antibiotics may occur after the direct treatment of bees against bacterial brood diseases, such as American foulbrood (AFB) or European Foul Brood (EFB) (Serra Bonvehí & Gutiérrez, 2008). Residues could also originate from the increasing use of antibiotics to treat bacterial infections of orchard plants and trees (McManus & Jones, 1994). Thus, important fruit-tree diseases such as Pseudomonas blossom blast are treated with antibiotics, mainly during blossom ( Spotts & Cervantes, 1995). Contamination of the blossom with high concentrations of antimicrobials implies the risk of a carry-over of the residues into the honey ( Heering et al., 1998 and Wan et al., 2005)
Sulfonamides are among the antibiotics most frequently found in honey (Reybroeck et al., 2010 and Wang et al., 2006). The majority of the sulfonamides currently used show a relatively long half-life, which may result in serious health problems in humans, such as allergic or toxic reactions (Sensderson, Naisbitt, & Park, 2006). Moreover, it has been reported that, in honey, sulfonamides tend to bind sugars via the formation of N-glycosidic bonds through their aniline group (Sheth, Yaylayan, Low, Stiles, & Sporns, 1990). Although governmental and regulatory agencies have established maximum residue limits (MRLs) for sulfonamides residues in different food commodities to safeguard public health (Commision Regulation, 1990 and Food and Durg Regulations, 1991), no MRLs have been established for honey in Europe, since the use of antimicrobials to treat honeybees is not authorized. Nevertheless, since in many other countries, these practices are legal, problems arise regarding imports of honey into the EU, which calls for reliable, rapid and high-throughput screening (HTS) analytical methods to ensure that imported honey samples placed in the EU market will comply with the EU rules. As stipulated in Annex II of Council Directive, 2001/110/EC, honey must be free from organic or inorganic foreign matter to its composition. In the absence of either EU MRLs, some countries within the European Union have established their particular action limits (recommended target concentrations, non-conformity or tolerance levels) for these antibiotics (Bernal et al., 2009 and Reybroeck et al., 2012). As an example, Belgium and the United Kingdom have set up tolerance levels of 20 and 50 μg kg−1, respectively, for total sulfonamides in honey, and Switzerland has set a tolerance level of 50 μg kg−1, referring to the sum of sulfonamides and their metabolites.
High-performance liquid chromatography followed by tandem mass spectrometry or fluorescence detection are the most commonly used techniques for the analysis of sulfonamides in honey (Maudens et al., 2004 and Sheridan et al., 2008). Alternatively, immunochemical methods could complement the screening of antibiotic residues, based on their simplicity, low cost and high throughput capabilities (Heering et al., 1998, Pastor-Navarro et al., 2007, Serra Bonvehí and Gutiérrez, 2008 and Tafintseva et al., 2009). Despite these advantages, current immunochemical analytical procedures reported for the detection of sulfonamide residues in honey samples still involve complex extraction and clean up processes using organic solvents (Heering et al., 1998, Pastor-Navarro et al., 2007 and Tafintseva et al., 2009), which limit their use as first action screening methods. Moreover the necessity to release the sulfonamides from the sugar conjugates is rarely discussed. Thus, blocking of the aniline group by the sugar, could result in a decrease of the recognition of the sulfonamides by the antibody and in an underestimation of the concentration of these residues in the sample. Eventually, chemical methods can be used to release the antibiotic prior the analysis. Hence, strong acids may be employed to break the imine bond formed between the sugar and the sulfonamide (Schawaiger & Schuch, 2000), but usually these procedures yield complex samples that have to be purified prior the analysis. Sheridan et al. (2008) used a multi-screening approach to monitor 14 sulfonamide compounds and chloramphenicol applying acidic hydrolysis (1 h, room temperature, RT) to liberate the sugar-bonded sulfonamide, but the sample had to subsequently be purified by solid-phase extraction (SPE) to remove potential interferences with an absolute recovery of around 60%. Similarly, Wang et al. (2012) describe an acidic hydrolysis step (1 h, RT) followed by liquid–liquid extraction (LLE) prior LC–ESI-MS/MS analysis.
The advantages of using antibody-derivatized magnetic particles to simplify sample treatment procedures are well recognized. A significant number of immunoassays and immunosensors have exploited the benefits of using magnetic particles biofunctionalized with either antigens or antibodies, as it can be seen in recently published papers (Baniukevic et al., 2013, Font et al., 2008, Lermo et al., 2009, Orlov et al., 2013 and Xu et al., 2012), and reviews (Aguilar-Arteaga et al., 2010, Kuramitz, 2009, Pedrero et al., 2012 and Zhang and Zhou, 2012). Few years ago, we also reported their use on a direct ELISA (Font et al., 2008) and an electrochemical immunosensor (Zacco et al., 2007) to directly detect sulfonamides residues in milk samples without any sample treatment, although the sulfonamide selectivity profile of such immunochemical approaches was very narrow. Later on, we reported a broad-selectivity microplate-based indirect ELISA able to detect up to 10 different sulfonamide antibiotic congeners in different biological samples (Adrian, Font et al., 2009 and Adrian, Gratacós-Cubarsí et al., 2009). The antibodies used were raised against an immunizing hapten maximizing recognition of the common epitope of this antibiotic family, which is the aniline group. Based on this previous knowledge, we report here the development of a rapid and efficient broad-selectivity immunochemical procedure to quantify the sulfonamide-sugar conjugate fraction of honey samples, involving a quick hydrolysis step prior the immunochemical analysis.

2. Materials and methods

2.1. Chemicals and immunochemicals

All the sulfonamides used in this work were supplied by Riedel-de Haën (Seelze, Germany). Hapten SA1 (5-[6-(4-amino-benzenesulfonylamino)-pyridin-3-yl]-2-methyl-pentanoic acid) and hapten SA2 (5-[4-(amino) phenylsulfonamide]-5-oxopentanoic acid) were prepared as previously described (Adrian, Font et al., 2009 and Font et al., 2008). Ovalbumin (OVA) and the secondary antibody peroxidase conjugate (antiIgG-HRP) were purchased from Sigma Chemicals Co. (St. Louis, Missouri). Tosyl-activated paramagnetic beads (MB-Tosyl, from now on named “magnetic beads”) were obtained from Dynal Biotech ASA (Oslo, Norway). Stock solutions of different sulfonamides were prepared in dimethylsulfoxide (DMSO, Merck, Darmstadt, Germany) (10 mmol L−1) and stored at 4 °C. The Bradford reagent (BIO-RAD protein assay cat no. 600–0005) was purchased from BIO-RAD laboratories GmbH (Munich, Germany). The immunoreagents used in this study (As155 and SA2-OVA) were produced as previously described (Adrian, Font et al., 2009 and Font et al., 2008). As155 was raised against hapten SA1 (5-[6-(4-aminobenzenesulfonylamino)pyridin-3-yl]-2-methylpentanoic acid) coupled to HCH (horseshoe crab hemocyanin). The SA1 hapten was designed to maximize recognition of the common aniline group of the sulfonamide antibiotic family. SA2-OVA is 5-[4-(amino)phenylsulfonamide]-5-oxopentanoic acid covalently coupled to ovalbumin. Antigen-derivatized magnetic particles (SA2-OVA-MP) were prepared following manufacturer instructions. Briefly, tosyl-modified magnetic beads (300 μL; 100 mg mL−1) were washed twice with coating buffer (1 mL), avoiding foaming. Magnetic beads were then re-suspended in coating buffer (350 μL) and the antigen SA2-OVA was added (150 μL; 1 mg mL−1). The beads were incubated 48 h at room temperature (RT) with slow orbital rotation to minimize bead sedimentation. Next, the modified magnetic beads were washed twice with PBS-BSA, (1 mL) for 5 min at 4 °C and then with Tris-BSA (1 mL), for 24 h at RT, with slow orbital agitation. Finally, the magnetic beads were re-suspended in PBS-BSA, (500 μL) to reach a concentration of 15 mg mL−1 and stored at 4 °C ready to use. The coupling efficiency was evaluated with the Bradford test (Bradford, 1976), analyzing the protein concentration in the supernatant before and after the conjugation.

2.2. General methods and instrumentation

The pH and the conductivity of all buffers and solutions were measured with a pH-meter pH 540 GLP and a conductimeter LF 340, respectively (WTW, Weilheim, Germany). Polystyrene microtiter plates were purchased from Nunc (Maxisorp, Roskilde, Denmark). Washing steps were performed on an SLY96 PW microplate washer (SLT Lab instruments GmbH, Salzburg, Austria). A Heidolph Titramax 1000 vibrating platform shaker (Brinkmann Instruments, Westbury, NY) was used to shake the microplates at 900 rpm. Absorbances were read on a Spectramax Plus (Molecular Devices, Sunnyvale, CA). The competitive curves were analyzed with a four-parameter logistic equation using the software SoftmaxPro v4.7 (Molecular Devices) and GraphPad Prism4 (GraphPad Software Inc., San Diego, CA). The preparation of the immunoreagents used in the ELISA assays was previously reported (Adrian, Font et al., 2009). The preparation of the immunoreagents for the magneto-ELISA is described below. For the magneto-iELISA washing steps, Dynal MPC-S (Dynal Biotech ASA, Oslo, Norway) or 96-well plate magnetic separation racks (CD1001, Cortex Biochem, San Leandro, CA, USA) were used.

2.3. Buffers and solutions

Phosphate-buffered saline (PBST) is 0.01 M phosphate buffer, 0.8% saline solution, pH 7.5 with 0.05% Tween 20. PBS 2xT contains 0.1% Tween 20. For the immunoassay experiments with honey samples, PB (0.1 M phosphate buffer, pH 7.5) was used. HCl and NaOH are 2 N. NaCl is 2 M. Coating buffer is 0.05 M carbonate–bicarbonate buffer, pH 9.6. Citrate buffer is a 0.04 M solution of sodium citrate, pH 5.5. The substrate solution contains 0.01% of 3,3′,5,5′-tetramethylbenzidine (TMB) and 0.004% H2O2 in citrate buffer. During biofunctionalization of the magnetic particles, PBS-BSA (PBS containing 0.1% (w/v) BSA) and Tris-BSA (0.2 M Tris, pH 8.5, 0.1% (w/v) BSA) was employed. All buffers and solutions were prepared using milliQ water (17.8 MΩ cm at 25 °C)

2.4. Microplate-iELISA (As155/SA2-OVA)

Microtiter plates were coated with SA2-OVA (0.75 μg mL−1 in coating buffer, 100 μL/well) overnight at 4 °C covered with adhesive plate sealers. The next day, plates were washed four times with PBST (300 μL/well), and then sulfapyridine (SPY) standards (50 μM–0.64 nM and zero in PBS) or the samples were added to the microplate (50 μL/well), followed by the antiserum As155 (1/8000 in PBS 2xT, 50 μL/well) and incubated for 30 min at RT. The plates were washed as before, and a solution of anti-IgG-HRP (1/6000 in PBST) was added to the wells (100 μL/well) and incubated for 30 min at RT. The plates were washed again, and the substrate solution was added (100 μL/well). Color development was stopped after 30 min at RT with 4 N H2SO4 (50 μL/well), and the absorbances were read at 450 nm. The standard curves were fitted to a four-parameter equation according to the following formula: y = (A − B}/[1 + (x/C)D] + B where A is the maximum absorbance, B is the minimum absorbance, C is the concentration producing 50% of the maximal absorbance, and D is the slope at the inflection point of the sigmoid curve.

2.5. Magneto-iELISA

The assay was performed using non-protein binding microplates and the washing steps were performed with the aid of a magnetic separation rack. Optimal concentration of immunoreagents was selected by two dimensional checkerboard titration experiments, where the avidity of the different dilutions of the antisera (As155, 1/250–1/16000 in PBS 2xT; 50 μL/well) for different concentrations of SA2-OVA-MP (2–0.015 mg mL−1 in PBST; 50 μL/well) was assessed following a similar procedure as for the microplate-iELISA. For the competitive experiments, a SA2-OVA-MP suspension (0.25 mg/mL in PBST; 50 μL/well) was first added to microtiter plate, followed by the standard solutions of SPY (10 μM–0.64 nM and zero in PBS; 25 μL/well) and the antiserum As155 (1/2000 in PBS 2xT, 25 μL/well). Then, the mixture was incubated for 30 min at RT under gentle shaking. After washing, the assay proceeded as described before.

2.6. Honey samples

Blank honey samples were certified for the absence of sulfonamide antibiotic residues by the Nestle Research Center (Lausanne, Switzerland) using their internal normalized protocols. Briefly, the honey samples were hydrolyzed by dissolving the samples in trichloroacetic acid and heating the solutions at 64 °C for 1 h. After cooling, the solution was extracted twice with acetonitrile-dichloromethane and the combined organic extracts were evaporated to dryness. The dry extracts were then dissolved and derivatized with fluram. After filtration, the derivatized solutions were analyzed by HPLC with fluorimetric detection. Under these conditions, twelve sulfonamides (sulfanilamide; SAM, sulfadiazine; SDZ, sulfapyridine; SPY, sulfathiazole; STZ, sulfamerazine; SMR, sulfamethazine; SMZ, sulfamethoxipyridazine; SMP, sulfachloropyridazine; SCP, sulfadoxine; SDX, sulfamethoxazole; SMX, sulfadimethoxine; SDM, sulfaquinoxaline; SQX) can be detected at limit of detection of 5 μg kg−1.
Hydrolyzed honey samples for immunochemical analysis were prepared by adding a solution of 2 N HCl (200 μL) to the honey samples (200 mg) placed on Eppendorf tubes. The mixture was stirred and sonicated to produce a homogeneous clear solution. After heating for 1 h at 45 °C (Method A) or 5 min in boiling water (Method B), the resultant solutions were neutralized with 2 N NaOH (200 μL) and buffered with 100 mM PB (400 μL). Finally, the samples were filled up to 2 mL with Milli-Q H2O to obtain a 1/10 dilution of the honey samples in 20 mM PB (100 mg hydrolyzed honey mL−1).
Non-hydrolyzed honey samples ((Method C) were prepared by dissolving the honey (200 mg) placed in Eppendorf tubes, with a 2 M NaCl aqueous solution (200 μL), diluted as before with 100 mM PBS (400 μL), and filled up with Milli-Q H2O to obtain a 2 mL buffered 1/10 diluted solution (100 mg honey mL−1).
Fortified honey samples were prepared by spiking six homogenized (agitated for 15 min at 40 °C) blank honey samples (3 g) with solutions (15, 30 and 60 μL) of SPY (5 mg L−1 in Milli-Q H2O) and sulfathiazole (STZ, 5 mg L−1 in Milli-Q H2O) each. This provided honey samples fortified with SPY and STZ at three different concentrations (100, 50 and 25 μg kg−1). The mixtures were agitated overnight at RT with slow orbital rotation. Fortified honey solutions were prepared by spiking hydrolyzed and non-hydrolyzed blank honey solutions, prepared as described above, with SPY and STZ (at 10, 5 and 2.5 μg L−1, corresponding to 100, 50 and 25 μg kg−1).

2.7. Matrix effect studies

Hydrolyzed and non-hydrolyzed blank honey samples were used to prepare SPY standard curves that were measured with the microplate and the magneto-iELISAs. The sigmoid curves obtained were compared to that prepared in buffer. Analyses were done in duplicates and in different days (n = 4).

2.8. Studies on the sulfonamide-sugar conjugation and the efficiency of the hydrolysis step

Sulfapyridine and sulfathiazole fortified honey samples (100, 50 and 25 μg kg−1), prepared as described above, were measured before and after the hydrolysis step, and the results were compared with those obtained by measuring the fortified hydrolyzed and non-hydrolyzed honey solutions (10, 5 and 2.5 μg L−1). Analyses were done in duplicate in three different days (n = 3). Recovery values were calculated by comparison of the concentration of the fortified samples spiked before and after treatment.

2.9. Accuracy studies

Blind samples of buffer and hydrolyzed and non-hydrolyzed honey samples were prepared by spiked them with SPY at different concentrations (102–1.2 μg kg−1, and zero). The samples were treated following the procedures described above (Method A–C) and measured with the microplate and the magneto-iELISAs in triplicates. Correlation was evaluated by analyzing the linear regression between the spiked and the measured concentration values. For the microplate-based ELISA the samples were spiked at 98.4, 49.2, 24.6, 12.3, 6.1, 2.5 and 1.2 μg kg−1, for the magneto-ELISA, the samples were spiked at 102, 51, 25.5, 12.7 and 6.4 μg kg−1.

2.10. Selectivity studies in honey

Stock solutions of ten sulfonamides: sulfadiazine, sulfamethazine, sulfamethoxazole, sulfamerazine, sulfisomidine, sulfadimethoxine, sulfachloropyridazine, sulfamethoxipyridazine, sulfathiazole and sulfaquinoxaline were prepared (10 mM in DMSO) and stored at 4 °C. Standard curves of each of them (0.64 nM to 10 μM, and zero) were prepared in hydrolyzed honey samples (Method B) and measured with the magneto-iELISA. The cross-reactivity (CR) values were calculated according to the equation [IC50(SPY)/IC50(cross-reactant) × 100].

3. Results and discussion

3.1. Immunochemical assay performance in honey samples

Few years ago, we reported the development of sulfonamide immunoreagents addressed to provide wide selectivity to the immunochemical analytical procedures developed. This was accomplished by designing a hapten that maximized recognition of the common aniline group of this antibiotic family (Adrian, Font et al., 2009). In this work, we address the implementation of a high-throughput immunochemical screening method to the analysis of these residues in honey samples. However, since it has been reported that, in honey, sulfonamide antibiotics may be covalently bound to sugars through their free N4-amino moiety of the p-aminobenzene sulfonamide structure, yielding stable conjugates ( Sheth et al., 1990), the need of a hydrolytic step prior the immunochemical analysis was quite likely.
To probe this fact, fortified honey samples with and without hydrolysis were analyzed by the SA microplate based-iELISA, but initially we assessed the performance of this assay in honey samples, using SPY as reference standard. Fig. 1 shows the calibration curves recorded in buffer and in hydrolyzed and non-hydrolyzed honey samples solutions prepared as described in the experimental section (100 mg mL−1 in the assay buffer). At that point, hydrolysis of the honey samples was performed as described in the literature, treating the sample with 2 N HCl at 45 °C for 1 h (Method A) ( Schawaiger & Schuch, 2000). As it can be observed, although compared to buffer, the SPY standard curves prepared in honey samples showed lower maximum absorbance, the IC50 values and LODs did not vary significantly (see Table 1 for parameters of these assays). The detectability accomplished in honey was much below of the action limits established by some EU countries such as Belgium (20 μg kg−1), United Kingdom or Switzerland (both at 50 μg kg−1) for this type of antibiotics ( Bernal et al., 2009 and Reybroeck et al., 2012). Thus, the LOD values achieved ranged between 0.6 and 1.1 μg kg−1, independently on whether the sample had been hydrolyzed or not. Moreover, the detectability at the middle point of the calibration curve was also very good, reaching IC50 values around 20 μg kg−1. These results pointed to the possibility to measure SA residues in these matrices, by using the blank extract to prepare the standards.
Calibration curves obtained by the microplate-based iELISA (graph A) and the ...
Fig. 1.
Calibration curves obtained by the microplate-based iELISA (graph A) and the magneto-iELISA (graph B) in buffer, hydrolyzed and non-hydrolyzed samples. Hydrolytic method A (2 N HCl, 1 h 45 °C); Hydrolytic method B (2 N HCl, 5 min 100 °C). Each concentration point was measured in triplicates. The concentration in the x-axes do refer to the concentration in the honey solution, in order to know the concentration in honey the value has to be multiplied by 10 and expressed in μg kg−1.
Figure options
Table 1. Features of the microplate-iELISA and magneto-iELISAa (SPY has been used as standard).
MethodMicroplate-iELISA
Magneto-iELISA
SampleBufferHoney (1/10)b
BufferHoney (1/10)b
Hydrolysis method–Method AcMethod BdMethod Ce–Method BdMethod Ce
Amax1.08 ± 0.180.64 ± 0.240.67 ± 0.340.74 ± 0.261.26 ± 0.071.13 ± 0.101.15 ± 0.07
Amin0.17 ± 0.040.09 ± 0.050.11 ± 0.060.09 ± 0.040.26 ± 0.010.15 ± 0.050.17 ± 0.04
IC50 (μg L−1)2.24 ± 0.991.89 ± 0.511.99 ± 0.571.96 ± 0.453.92 ± 1.782.65 ± 0.684.64 ± 2.24
IC50 (μg kg−1)b–18.9 ± 5.119.9 ± 5.719.6 ± 4.5–26.5 ± 6.846.4 ± 22.4
LOD (μg L−1)f0.09 ± 0.040.06 ± 0.040.11 ± 0.070.085 ± 0.0470.39 ± 0.190.14 ± 0.110.24 ± 0.05
LOD (μg kg−1)b–0.61 ± 0.41.1 ± 0.70.85 ± 0.47–1.4 ± 1.12.4 ± 0.5
Working range (μg L−1)0.28 ± 0.110.24 ± 0.110.32 ± 0.170.24 ± 0.150.89 ± 0.370.39 ± 0.270.68 ± 0.10
22.05 ± 7.9814.14 ± 3.8513.85 ± 2.6114.20 ± 5.0519.99 ± 10.0519.94 ± 4.9631.82 ± 12.42
Working range (μg kg−1)b–2.4 ± 1.13.2 ± 1.72.4 ± 1.5–3.9 ± 2.76.8 ± 1.0
–141.4 ± 38.5138.5 ± 26.1142.0 ± 50.5–199.4 ± 49.6318.2 ± 124.2
Slope−0.70 ± 0.08−0.66 ± 0.11−0.73 ± 0.14−0.77 ± 0.06−0.75 ± 0.07−0.69 ± 0.17−0.69 ± 0.09
R20.993 ± 0.0040.994 ± 0.0020.98 ± 0.0060.991 ± 0.0020.98 ± 0.0090.98 ± 0.0040.97 ± 0.013
N4444333
a
Values correspond to the average and standard deviation of each parameter of at least four assays performed on different days.
b
Honey samples were prepared as described in the experimental section.
c
Samples were hydrolyzed with 2 N HCl at 45 °C for 60 min.
d
Samples were hydrolyzed with 2 N HCl at boiling water for 5 min.
e
Non hydrolyzed samples diluted 10 times prior the analysis.
f
The limit of detection (LOD) has been defined as the concentration giving a signal which is 90% of the maximum absorbance.
Table options

3.2. Sulfonamide-sugar conjugation in honey samples

With these initial results and in order to assess the ability of our immunochemical assay to detect sulfonamide-sugar conjugated residues, honey samples fortified at three different concentration levels (100, 50 and 25 μg kg−1) with SPY and STZ, were splinted in two parts, and measured 24 h later with and without hydrolysis using the corresponding calibration curves. SPY and STZ were chosen as sulfonamide representative congeners having in their chemical structure two very different heterocycles. The results shown in Table 2, demonstrate the efficiency of the hydrolysis procedure since high recovery values were obtained for the hydrolyzed honey samples, independently of the nature of the heterocycle in the sulfonamide (70–93%, for SPY, 88–104% for STZ). In contrast, the measured immunoreactivity of the non-hydrolyzed samples (Method C) was extremely low, indicating that conjugation to the sugar occurred, probably in less than 24 h, on an almost quantitative manner.
Table 2. Results of the recovery studies, analyzing free sulfonamide obtained from acid hydrolysis reaction of sulfonamide conjugated to glucosea.
SA spiked (μg kg−1)Method A (1 h 45 °C) (%)Method B (5′ 100 °C) (%)Method C (no hydrolysis) (%)
SPY10084.2 ± 9.493.0 ± 5.51.9 ± 0.8
5093.2 ± 7.8106.1 ± 12.33.5 ± 2.7
2570.2 ± 10.099.7 ± 16.53.9 ± 1.9

STZ10088.8 ± 9.396.0 ± 2.92.8 ± 0.7
5096.2 ± 17.3107.5 ± 20.71.3 ± 0.3
25104.2 ± 10.891.7 ± 5.41.0 ± 1.3
a
Results are expressed in % of the measured values in the fortified samples in respect to values measured for the fortified sample solutions (see experimental section). Recovery studies were performed with the microplate-iELISA, in duplicate in three different days (standard deviation, n = 3). The proposed hydrolytic reaction is extracted from Sheth et al. (1990).
Table options

3.3. Optimization of the sulfonamide-sugar hydrolysis method

Further investigations were addressed to reduce the time of the sample preparation method, by increasing the temperature of the hydrolysis step. The results revealed that very good recoveries could also be accomplished by performing the hydrolysis during just 5 min in a boiling water bath (Method B, 93–106%, for SPY, 91–107% for STZ, see Table 2). Moreover, as it can be observed in Fig. 1 (graph A), there was no difference on the effect of both hydrolytic methods (A and B) on the immunoassay performance.

3.4. Magneto-ELISA

On another set of experiments, we aimed at removing the residual matrix interference that produced a decrease in the maximum absorbance. With this purpose, we evaluated the possibility to using magnetic beads to washout the potential interferences causing this effect. Biofunctionalized magnetic particles have found to be excellent tools in bioanalytical chemistry because of the possibility to use them as selective extraction tools, or as a strategy to separate the bioreagents from the assay media. Moreover, reaction kinetics may be favored with respect to other bioreagent-linked strategies, since the reaction takes place in solution. With this purpose, the SA2-OVA bioconjugate, used as coating antigen, was covalently coupled to tosyl-functionalized magnetic beads through its free lysine amino groups, with a very good efficiency (95%, n = 2). Checkerboard titration experiments were used to select the appropriate concentrations of the SA2-OVA-MP (magnetic particles biofunctionalized with SA2-OVA), the class-specific antibody As155 and the secondary antibody antiIgG-HRP. Using the selected conditions, a SA indirect magneto ELISA was established with very good features. Although the detectability recorded in buffer was slightly lower than that of the microtiter plate ELISA (LOD of 0.39 μg L−1vs. 0.09 μg L−1, see Table 1 and Fig. 1, graph B), it was still very good, taking into consideration the reference action points established in some European countries.
The potential non-specific interferences caused by the honey matrix in the magneto-iELISA was determined building SPY standard curves in hydrolyzed (Method B) and non-hydrolyzed (Method C) blank honey samples. As can be observed in Fig. 1 (graph B), the use of magnetic particles resulted in calibration curves with almost the same features as those obtained in buffer. Thus, in contrast to the behavior observed with the microplate-iELISA, the maximum absorbance and the IC50 values recorded in buffer and in honey were very similar, even after the hydrolysis step (see Table 1). For the hydrolyzed sample, the detectability at the middle point of the assay (IC50) was 26 μg kg−1 with a LOD of 1.4 ± 1.1 μg kg−1. These results pointed to the possibility to just using the calibration curve built in buffer as a reference for quantification of the sulfonamide-sugar fraction of the contaminated honey samples. Moreover, selectivity studies performed in hydrolyzed honey samples indicated that the assay kept the same wide recognition pattern (see Table 3) reported for the microplate-based iELISA in buffer, being possible to detect a significant number of sulfonamides below the action limits proposed by some EU countries (between 20 and 50 μg kg−1) ( Bernal et al., 2009 and Reybroeck et al., 2012). Thus, at least nine different sulfonamides below 50 μg kg−1 could be detected with the magneto-iELISA here reported.
Table 3. Selectivity of the magneto-iELISA in hydrolyzed honey samples.
CompoundAcronymFigureMagneto-iELISAa in Honey
% CRIC50 (μg L−1)LOD (μg L−1)bLOD (μg kg−1)c
SulfapyridineSPYFull-size image (2 K)1002.65 ± 0.680.14 ± 0.111.4 ± 1.1
SulfadiazineSDZFull-size image (2 K)1026.631.4214.2
SulfamethazineSMZFull-size image (2 K)1172.520.131.3
SulfamerazineSMRFull-size image (2 K)813.470.191.9
SulfisomidineSIDFull-size image (2 K)1232.410.131.3
SulfadimethoxineSDMFull-size image (2 K)482.554.4144.1
SulfachloropyridazineSCPFull-size image (2 K)2151.478.1081.0
SulfamethoxipyridazineSMPFull-size image (2 K)555.40.292.9
SulfathiazoleSTZFull-size image (2 K)2111.30.070.7
SulfaquinoxalineSQXFull-size image (2 K)2512.80.686.8
The Amax and Amin of the assays ranged between 1.13 ± 0.10 and 0.15 ± 0.045, respectively.
a
Results indicate the cross reactivity (CR%) expressed as [IC50(SPY)/IC50(cross reactant)] × 100.
b
Limit of detection in the honey extract.
c
Limit of detection in honey.
Table options

3.5. Accuracy studies

Accuracy of the immunochemical procedures established for the analysis of the sulfonamide-sugar-conjugated fraction of contaminated honey samples, were performed by analyzing blind spiked samples. The results shown in Fig. 2 (graphs A and B for the microplate and magneto iELISAs, respectively) show that measured values matched very will the spiked concentrations. In all cases, the slopes resulting from the linear regression analysis were close to 1 and the correlation coefficients higher than 0.995. Moreover, two honey samples spiked with SPY and STZ at 25 ppb were hydrolyzed using method A (2 M HCl, 1 h at 45 °C) and subsequently analyzed by HPLC–MS/MS and by the immunochemical method. The recovery values obtained were very similar for both methods. Thus, while for HPLC–MS/MS were 71.60% and 91.60% for the SPY and STZ spiked samples, respectively, the same samples, analyzed with the immunochemical procedure here reported provided recovery values of 70.2% and 104.2%, respectively. At the light of these results, we believe that the procedure here reported may provide quantitative results with an accuracy very similar to the actual chromatographic reference methods.
Results from the accuracy studies performed with the microplate-based iELISA ...
Fig. 2.
Results from the accuracy studies performed with the microplate-based iELISA (graph A) and the magneto-iELISA (graph B). Honey samples were spiked and 24 h analyzed by just diluting the sample (non-hydrolyzed samples) or applying the hydrolytic procedures described in the experimental section. Analyses were performed in triplicates. The dotted line corresponds to a perfect correlation (slope = 1). The correlation was in all cases very good. Microplate-based iELISA: Buffer: y = 1.05x − 0.94, R2 = 0.998; non hydrolyzed honey: y = 1.10x + 0.67; R2 = 0.996; hydrolyzed honey (Method A): y = 1.04x + 0.34; R2 = 0.997; hydrolyzed honey (Method B); y = 1.02 + 0.64; R2 = 0.998. Magneto-iELISA: hydrolyzed honey (Method B): y = 0.93x − 2.52; R2 = 0.991.
Figure options

4. Conclusions

It has been demonstrated that sulfonamide antibiotics rapidly react with components of the honey matrix, probably sugars, as stated in the introduction. This fact indicates the mandatory need to perform hydrolysis prior the analysis. In that respect, an efficient and reliable immunochemical analytical procedure for the analysis of sugar-conjugated sulfonamide antibiotic residues in honey samples has been developed. The method involves hydrolysis of the sugar conjugates in just 5 min (Method B), followed by immunochemical analysis using either a microplate-based iELISA or a magneto-iELISA. The whole immunochemical procedure, including sample treatment, requires only a little bit more than 1.5 h and many samples can be simultaneously screened. The use of magnetic beads has found to reduce the nonspecific matrix interferences which allowed the samples to be quantified using the calibration curve built in PBS. Under these conditions, and using SPY as reference standard, detectability (LOD) achieved in hydrolyzed honey samples is 1.1 ± 0.7 μg kg−1 and of 1.4 ± 0.1 μg kg−1for the microplate and magneto-iELISAs, respectively. Moreover, due to the broad specificity of the immunoreagents used, a significant number of sulfonamide residues can be detected with the immunoanalytical protocols reported in this work.

Acknowledgments

This work has been supported by the European Community (Conffidence project, KBBE2008-211326). CIBER-BBN is an initiative funded by the he Spanish National Plan for Scientific and Technical Research and Innovation 2013–2016, Iniciativa Ingenio 2010, Consolider Program, CIBER Actions and financed by the Instituto de Salud Carlos III with assistance from the European Regional Development Fund. The Nanobiotechnology for Diagnostics group (Nb4D) group (previously named Applied Molecular Receptors group (AMRg)) is a consolidated Grup de Recerca de la Generalitat de Catalunya and has support from the Departament d’Universitats, Recerca i Societat de la Informació la Generalitat de Catalunya (expedient 2014 SGR 1484). We would like to thanks CAbS for its assistance on immunoreagent production.

References

    • Bogdanov, 2005
    • S. Bogdanov
    • Contaminants of bee products
    • Apidologie, 37 (2005), pp. 1–18
    • Commision Regulation, 1990
    • Commision Regulation (ECC) No. 2377/99
    • Official Journal of the European Communications, L224 (1990), p. 1
    • Schawaiger and Schuch, 2000
    • I. Schawaiger, R. Schuch
    • Bound sulfathiazole residues in honey – need of a hydrolysis step for the analytical determination of total sulfathiazole content in honey
    • Deutche Lebensmittel-Rundschau, 96 (3) (2000), pp. 93–98
    • Sensderson et al., 2006
    • J.P. Sensderson, D.J. Naisbitt, B.K. Park
    • Role of bioacitvation in drug-induced hytpersensitivity reactions
    • AAPS Journal, 8 (2006), pp. E55–E64
    • Serra Bonvehí and Gutiérrez, 2008
    • J. Serra Bonvehí, A.L. Gutiérrez
    • Residues of antibiotics and sulfonamides in honeys from Basque Country (NE Spain)
    • Journal of the Science of Food and Agriculture (2008) n/a
    • Sheth et al., 1990
    • H.B. Sheth, V.A. Yaylayan, N.H. Low, M.E. Stiles, P. Sporns
    • Reaction of reducing sugars with sulfathiazole and importance of this reaction to sulfonamide residue analysis using chromatographic, colorimetric, microbiological, or ELISA methods
    • Journal of Agriculture and Food Chemistry, 38 (4) (1990), pp. 1125–1130
    • Full Text via CrossRef
      |
      View Record in Scopus
       | 
      Citing articles (22)
    • Tafintseva et al., 2009
    • I. Tafintseva, A. Zherdev, S. Eremin, B. Dzantiev
    • Enzyme immunoassay for determination of sulfamethoxypyridazine in honey
    • Applied Biochemistry and Microbiology, 46 (2) (2009), pp. 216–220
    • Wan et al., 2005
    • G.-H. Wan, H. Cui, H.-S. Zheng, J. Zhou, L.-J. Liu, X.-F. Yu
    • Determination of tetracyclines residues in honey using high-performance liquid chromatography with potassium permanganate–sodium sulfite–β-cyclodextrin chemiluminescence detection
    • Journal of Chromatography B, 824 (1–2) (2005), pp. 57–64
    • Article
      |
       PDF (289 K)
      |
      View Record in Scopus
       | 
      Citing articles (44)
Corresponding author at: Nanobiotechnology for Diagnostics group (Nb4D), IQAC-CSIC, Jordi Girona, 18-26, 08034-Barcelona, Spain. Tel.: +34 934006100; fax: +34 932045904.
1
http://www.ciber-bbn-es.
2
www.usisensor.be.
3
www.research.nestle.com.
Copyright © 2015 Elsevier Ltd. All rights reserved.
Posted by at