Synthesis and characterization of oxytetracycline imprinted magnetic polymer for application in food

Magnetic imprinted polymer was prepared by polymerization of methacrylate and ethyleneglycoldimethacrylate in the presence of oxytetracycline on the surface of iron magnetite. Selectivity of prepared polymer was calculated from ratio of partition coefficient of oxytetracycline for imprinted and non- imprinted polymer in water, acetonitrile, methanol and at different pH in aqueous buffer. pH of solvent exhibited pronounced effect on selectivity. Selectivity at pH 7.0, 6.0 and 5.0 was 36.0, 2.25 and 1.61 fold higher than at pH 4.0. Imprinted polymer was not selective for oxytetracycline in methanol. However, selectivity in water and acetonitrile was 19.42 and 2.86, respectively. Oxytetracycline did bind to imprinted polymer in water or aqueous buffer (pH 7.0) and could be eluted with methanol. Prepared polymer extracted 75–80 % oxytetracycline from water, honey and egg white.


Introduction
The widespread use of antibiotics in dairy cattle, poultry and bee keeping leads to presence of antibiotic residues in food chain. To prevent any harmful health effects on consumers, FAO, WHO and European Union have established the maximum residual limits for veterinary drugs (Council Regulation 2377/90/EEC). Oxytetracycline (OTC) is a broad spectrum antibiotic against Gram-positive and Gram-negative bacteria. It is used for the treatment of bacterial brood disease and is the most commonly applied tetracycline group of antibiotics to food producing animals. OTC also finds application in human therapy, apiculture and fruit crop production (Niazi et al. 2008). The presence of OTC residue in foods may constitute a variety of public health hazards including toxicological, microbiological, immunological and pharmacological hazards (Navrátilová et al. 2009). Frequent use of antibiotic impacts on emergence of antimicrobial resistance creates disorder in intestinal flora and may lead to possible occurrence of allergic reactions (Honkanen-Buzalski and Reybroeck 1997;Roberts 1997;Cerniglia and Kotarski 2005). Food analysis requires extensive sample preparation prior to its instrumental analysis (Ridgway et al. 2007). Classical sample preparation technique, which involves repeated sample extraction with solvent under shaking, followed by centrifugation, results in low extraction yield (Yan et al. 2007). Some high extraction yielding techniques such as pressurized liquid extraction (Haglund et al. 2007;Soler et al. 2007), supercritical fluid extraction (Jiménez-Carmona and Luque de Castro 1998; Rodil et al. 2005) microwave-assisted extraction (Cheng et al. 2007;You et al. 2007) and ultrasonic-assisted extraction (Ruiz-Jiménez et al. 2004;Rezić et al. 2005)] have also been applied for food analysis. However, prepared sample is not suitable for direct instrument analysis and requires one or more prior purification steps (Heems et al. 1998;Buldini et al. 2002;Paleologos and Kontominas 2004;Hamscher et al. 2005;Kuhnle et al. 2007). Extraction protocol specific to analyte can lead to clean sample preparation. Molecular imprinting is a technique which creates stable synthetic polymers possessing tailor-made selective recognition sites, obtained using high amount of crosslinker in the presence of template molecule (Xie et al. 2003). Once polymerization is over, removal of template from polymer leaves specific cavities which can rebind template selectively on the basis of shape, size and functionality. Despite the advantages of molecular imprinting polymer such as stability at extreme of pH and temperature, ease of preparation, low cost and reusability, it requires column or pre-packed column and negative or positive pressure generating pump or centrifuge.
The magnetic polymer is prepared by encapsulating inorganic magnetic particles in organic polymers (Yao et al. 2008). On integration of magnetic separation technology with imprinted polymer, the resulting magnetic polymer will have magnetic properties and selective for template molecule. Magnetic separation technology makes separation of analyte easier and faster and also provides large surface to volume ratio. The present study reports synthesis and application of OTC-imprinted magnetic polymer.

Food samples
Honey was obtained from Dabar India, Ltd., and eggs were obtained from local market. These food samples were spiked with OTC stock solution (100 lg/mL) prepared in water.

Preparation of oleic acid coated Fe 3 O 4 magnetite and EGDMA mix
The Fe 3 O 4 magnetite was prepared by co-precipitation method as described by Chen and Li (2013). 1.98 g of FeCl 2 Á4H 2 O and 5.41 g of FeCl 3 Á6H 2 O were dissolved in 100 mL of deaerated water. The mixture was stirred vigorously and purged with nitrogen gas while temperature was raised to 80°C. Then, 40 mL of sodium hydroxide solution (2 N) was added dropwise. After 1 h, magnetic precipitates were separated with help of external magnet and washed several times with deaerated water. After washing, precipitates were dried at 60°C. 1 g Fe 3 O 4 magnetite (freshly prepared) and 1.0 mL oleic acid were mixed. Then, 3.77 mL EGDMA was added and mixed.

Preassembly solution of monomer and OTC
In separate beaker, 1.0 mmol OTC and 344 lL MAA were added to 10 mL deaerated water to allow interaction between MAA and OTC. The contents were mixed for 10 min.

Pre-polymerization mix
Iron magnetite-EGDMA mix and preassembly solution were mixed and sonicated for 30 min.
Polymerization reaction 0.4 g PVP was dissolved in 100 mL 80 % ethanol at 60°C and purged with nitrogen gas. Then, pre polymerization mix and 3 mL AIBN were added. Reaction was allowed to proceed at 60°C for 24 h under continuous stirring at 300 rpm. Prepared polymer was separated with help of external magnet and washed several times with methanol: acetic acid (8:2 v/v) mixture, till washings were free from OTC. The polymer was further washed three times with deaerated water and dried at 60°C.
Magnetic non-imprinted polymer (MNIP) was prepared similarly to that of magnetic molecular imprinted polymer (MMIP), except OTC was omitted from pre-assembly solution.
Characterization MMIP was characterized by scanning electron microscope (SEM; Make: Carl Zeiss, Germany) and Fourier transform infrared spectrometry (FT-IR Sumazdu, IR affinity, Japan). The size of prepared iron nano particle and MMIP was determined with Zetasizer, (Malvern, USA).

Selectivity of imprinted polymer
Selectivity was calculated from ratios of partition coefficient of OTC in imprinted and non-imprinted polymer (Cai and Gupta 2004). Polymers (20 mg) were incubated with 2 mL OTC (40 lg) prepared in different solvent for 24 h at 30°C. Unbound OTC and other studied antibiotics were assayed in supernatant by measuring absorbance at 325 nm. Bound antibiotic was calculated by subtracting unbound antibiotic from total antibiotic added.

Evaluation of MMIP performance in food matrix
Two grams of honey or 2 mL of egg white or 2 mL of water was spiked with 200 lg OTC and diluted with 20 mL water. Diluted sample or water was mixed with 40 mg imprinted or non-imprinted polymer. Then, polymer was ten times washed with 3 mL water and supernatant from each washing was collected. The bound OTC was eluted six times with 3 mL methanol.
The imprinted polymer was preconditioned with water before use. Preconditioning of polymer was achieved by vortexing polymer for 10 min. Supernatant was then removed.

Results and discussion
Magnetic properties of imprinted polymer coated over the iron magnetite surface lead to easy separation of polymer after binding to target molecules. Therefore, such polymers have added advantage over non-magnetic imprinted polymers. Iron magnetite and polymers are core and coat of magnetic imprinted polymer, respectively. Polymer yield from 4 mmol MAA and 20 mmol EGDMA in the presence of 1 mmol OTC was 4.2 g. Imprinted polymer was brown in color while non-imprinted polymer was black. OTC imprinted particles were of 220 nm in size and appear porous (Fig. 1). Polymer exhibited peak at 540 cm -1 (characteristic of Fe-O bond stretching), 1370 cm -1 (C-H bond vibration), 1450 cm -1 (C-H bond stretching) and 1724 cm -1 (carbonyl bond stretching) in FTIR (Chen et al. 2010;Hong et al. 2010) (Fig. 2). There was no difference in FTIR spectra of OTC-imprinted and non-imprinted polymer.

Effect of solvents on binding of oxytetracycline
The binding efficiency of OTC imprinted polymer was dependent on solvent (Table 1). Binding efficiency of magnetic imprinted polymer was highest in water (82 %), followed by acetonitrile (30 %) and methanol (11 %). This shows that binding in polar solvents was considerably high in comparison to non-polar solvent and interaction between OTC and polymer appears largely hydrophobic in nature. Binding in acetonitrile shows that there is also possibility of involvement of hydrogen bond for interaction between OTC and magnetic imprinted polymer. Selectivity value reflects preferential binding of target to imprinted polymer over non-imprinted polymer. Solvent can influence noncovalent interaction between OTC and polymer and thereby influence selectivity. Selectivity of magnetic molecular imprinted polymer was solvent-dependent and was highest in water followed by acetonitrile and abolished in methanol. The results suggest involvement of hydrophobic interaction and hydrogen bond.

Effect of pH on binding of oxytetracycline to magnetic imprinted polymer
Binding efficiency of imprinted and non-imprinted polymers for OTC was dependent on pH (Table 1). Striking difference was observed in binding of OTC to non-imprinted polymer at different pH. As pH decreased, binding of OTC to non-imprinted polymer increased. Binding of OTC to imprinted polymer was highest at pH 7.0. The difference in binding of OTC to both imprinted and nonimprinted polymer at different pH resulted from dependence of selectivity on pH. The selectivity was highest at pH 7.0, whereas at pH 4.0 there was no selectivity. Nonimprinted polymer also binds to OTC in pH-dependent way. Dissociation of carboxylic group (pKa 5.5) in poly methacrylic acid (Zhang and Nicholas 2000) and of three hydroxyl groups (pKa values being 3.32, 7.78 and 9.58) present in OTC (Sassman and Lee 2005;O'Connor and Aga 2007) is dependent on pH. When both carboxylic and hydroxyl groups exist in dissociated form, hydrogen bond formation between polymer and OTC will be discouraged. At pH 7.0, hydrogen bond formation will be least favorable amongst the studied pH ranges (7.0, 6.0, 5.0 and 4.0). The significant binding in the pH range of 4-6 resulted from hydrogen bond between surface carboxylic group of polymer and OTC. Imprinted polymer still binds with OTC

Effect of time duration on binding of oxytetracycline to imprinted polymer
The effect of time duration on binding of oxytetracycline to imprinted polymer was studied. Polymer was immersed in OTC solution prepared in 20 mM phosphate buffer (pH 7.0) at 25°C up to 40 h (Fig. 3). Binding of OTC to imprinted polymer increases with increase in time duration up to 24 h and then remains constant. This shows that maximum binding sites were occupied by OTC in 24 h.

Cross-reactivity of oxytetracycline imprinted polymer
Forty microgram each of antibiotics was incubated with 20 mg polymer in 2 mL, 20 mM phosphate buffer (pH 7.0) for 24 h at 25°C. Bound antibiotic was calculated by subtracting unbound antibiotic from total antibiotic. As against 90 % binding of OTC to imprinted polymer, binding of metacycline, chlortetracycline, tetracycline, cefquinome and gentamycine was 10, 15, 20, 20 and 12 %, respectively (Fig. 4). It may be also noted that only 10-20 % of antibiotic did bind to non-imprinted polymer. Prepared MMIP was more selective for OTC as other studied antibiotics showed low cross-reactivity

Conclusion
In this paper, non covalent molecular imprinting protocol was used to prepare OTC-imprinted magnetic polymer. The selectivity of polymer for OTC was dependent on solvents used for binding. The prepared magnetic imprinted polymer extracted 75-80 % OTC from food matrix. Magnetic imprinted polymer provides an easy method for separation of polymer from food matrix or water. selective extraction of PCBs and dioxins from fish and fish oil using in-cell carbon fractionation pressurized liquid extraction.