A Group-Specific Microbiological Test for the Detection of Tetracycline Residues in Raw Milk Jussi Kurittu,* Stefan Lo ¨nnberg, Marko Virta, and Matti Karp Department of Biotechnology, University of Turku, Tykisto ¨katu 6A, FIN-20520 Turku, Finland The potentiality of using a luminescent Escherichia coli strain for the specific detection of tetracycline residues in raw bovine milk was investigated. The sensor cells contain a reporter plasmid carrying the bacterial luciferase operon of Photorhabdus luminescens under the control of the tetracycline responsive control region from transposon Tn10. Incubation of the cells with the sample containing tetracyclines increases the light emission of the sensor cells. The most sensitive tetracycline detection was achieved in 120 min and by using CDTA as a chelating agent in the assay. Heat-treatment of milk before the assay decreased the variations in background luminescence signals and in tetracycline-induced luminescence between different milk samples. The detection limits for tetracycline, oxytetracycline, chlortetracycline, doxycycline, methacycline, demeclocycline, and minocycline were between 2 and 35 ng/mL. Nontetracycline antibiotics did not significantly interfere with the detection of tetracyclines. Keywords: Tetracycline, luminescence, luxCDABE, luciferase, antibiotic residue INTRODUCTION Tetracycline antibiotics are widely used in animal husbandry for the treatment of bacterial infections and in some countries at subtherapeutic levels as feed additives to enhance growth of food-producing animals. Use of tetracyclines has been intensive mainly due to their activity against a wide range of pathogens, low price, and relatively low toxicity (Chopra et al., 1992; Standiford, 1995; Roberts, 1996). However, use of tetracyclines and other antibacterials in the farming industry has certain drawbacks: anti- biotic residues in food products, such as in milk, may provoke allergic symptoms in humans, or may lead to monetary losses in the dairy industry, e.g., by inhibiting starter cultures in food technological processes (Hee- schen, 1993; Ma ¨ yra ¨ -Ma ¨ kinen, 1995). Furthermore, resi- dues may promote the development and distribution of bacterial resistance to antibiotics (Roberts, 1996; Mc- Manus, 1997). Because of such disadvantages, food products need to be tested for residues before use. Current methods for the detection of tetracycline residues include microbiological inhibition tests, immu- noassays and chemical-physical methods. All the meth- ods available have their characteristic disadvantages: microbiological tests are relatively slow and nonspecific for tetracyclines, whereas immunoassays are usually quite expensive. Drawbacks of chemical-physical meth- ods based on high-priced instrumentation, such as HPLC and MS, include complexity, low amount of samples analyzed per time unit, and requirement of trained personnel. In view of these disadvantages and difficulties, there is a need for new simple, fast, inex- pensive, and sensitive assay methods for the detection tetracycline residues. An emerging concept to detect different organic and inorganic compounds is to use genetically modified bacterial cells, which respond specifically and rapidly to a certain analyte or group of analytes (Lewis et al., 1998). These sensor cells contain a reporter plasmid, in which a compound-specific regulatory unit is coupled to a reporter gene, such as luciferase or -galactosidase. In this kind of sensing system, the expression of the reporter gene is activated when the cells are exposed to the specific analyte. Gene expression can be sensi- tively detected, e.g., with electrochemical or luminescent methods depending on the reporter system utilized. Specific sensor strains have been constructed e.g. for heavy metals, organic pollutants and antibiotics (Chopra et al., 1990; King et al., 1990; Virta et al., 1995; Ramanathan et al., 1997; Tauriainen et al., 1998). We have previously constructed an Escherichia coli sensor strain for the detection of tetracyclines (Korpela et al., 1998). The sensor plasmid of the strain contains bacte- rial luciferase operon from P. luminescens under the control of the regulatory unit of tetracycline resistance determinant from transposon Tn10 (Hillen and Berens, 1994). The luciferase operon consists of five genes, luxC, D, A, B, and E, which are necessary and sufficient for in vivo light production of the E. coli cells (Meighen, 1991). In the absense of tetracyclines, the expression of bacterial luciferase genes in the plasmid is repressed, whereas the presence of tetracyclines in the sample induces luciferase gene expression, which can be seen as an increase in the light emission of the sensor cells. Although several specific sensor bacteria for a wide range of analytes have been developed, no extensive application of these sensors for real environmental or food samples has been described. The principal objective of this study was to investigate whether the previously constructed tetracycline sensor strain, E. coli K12- (pTetLux1), could be used to detect tetracycline residues in raw bovine milk, which is commonly known to be a rather problematic sample matrix: it contains large amounts of organic (e.g., proteins and carbohydrates) and inorganic compounds (e.g., calcium), as well as * To whom correspondence should be addressed. Phone: +358-2-333 8059. Fax: +358-2-333 8050. E-mail: juskur@utu.fi. 3372 J. Agric. Food Chem. 2000, 48, 3372-3377 10.1021/jf9911794 CCC: $19.00 © 2000 American Chemical Society Published on Web 07/11/2000