Development of Disposable Sensors for Rapid Multianalyte Detection: Acetylcholinesterase and Microbial Biosensors Entwicklung von Einmalsensoren zur schnellen Multianalytdetektion: Acteylcholinesterase- und mikrobielle Biosensoren

II The objective of this work was to develop new biosensors for multianalyte detection. Target analytes were cholinesterase inhibiting insecticides and chlorinated aromatic hydrocarbons. The task was fulfilled by a combination of biological receptors bearing different selectivities for the desired analytes and data evaluation by feed-forward artificial neural networks (ANN). The basic sensor design consisted of a fourelectrode thick film electrode which was fabricated by screen printing. As biological receptors for insecticide detection, variants of acetylcholinesterase (AChE) and for chlorinated aromatics Ralstonia eutropha JMP 134 cells were selected. AChE-multisensors: Three types of sensitive amperometric biosensors for the rapid discrimination of paraoxon, carbofuran, and malaoxon in binary mixtures were developed. The analysis based on AChE inhibition and ANN data evaluation. Each wild type AChE-sensor employed four types of native and recombinant AChEs (electric eel, bovine erythrocytes, rat brain, Drosophila melanogaster). The sensors registered both analytes in a detection range of 0.2 20 μg/l. Thus, paraoxon and carbofuran in mixtures of 0 20 μg/l for each analyte could be analysed with prediction errors of 0.9 μg/l for paraoxon and 1.4 μg/l for carbofuran within less than 60 min. Improved AChE-sensors could be obtained with engineered variants of Drosophila melanogaster AChE. They consisted of wild-type Drosophila AChE and mutants Y408F, F368L, F368H, or F368W. Both multisensors were used for analysis of paraoxon and carbofuran mixtures in a concentration range of 0 5 μg/l within 40 min. The two analytes were best determined with prediction errors of 0.4 μg/l for paraoxon and 0.5 μg/l for carbofuran. In addition, multisensor II was also investigated for analyte discrimination in real water samples. Finally, the properties of the system was confirmed by simultaneous detection of binary organophosphate mixtures. Malaoxon and paraoxon in composite solutions of 0 5 μg/l were discriminated with prediction errors of 0.9 μg/l and 1.6 μg/l respectively. Microbial multi-sensors: The starting point was the design of a novel miniature oxygen multi electrode. The disposable, Clark-type electrode was manufactured by a combination of screen printing, spin and drop coating and lamination techniques. Sensitivity, dynamic and stability characteristics proved essentially satisfactory. The sensor principle for simultaneous detection of halogenated aromatics based on the correlation of cell respiration and analyte concentration. Different selectivities of the receptor component were adjusted by selective induction of desired enzyme activities through cultivation of R.eutropha JMP 134 on 2,4-dichlorophenoxyacetic acid, 4-chloro-2-methylphenoxyacetic acid, 2-methylphenoxyacetic acid and phenol as carbon sources. The sensors were capable for a simultaneous detection of 2,4dichlorophenol and phenol in mixtures of 0 40 μM with mean errors of 4.5 and 3.6 μM (0.62 and 0.34 mg/l), respectively. Mixtures of 2,4-dichlorophenol and 4chlorophenol in the same concentration range were analysed with errors of 5.1 and 2.6 μM (0.33 and 0.7 mg/l), respectively. The multisensors developed in this work are the first examples of disposable enzyme and microbial sensors in combination with data evaluation by artificial neural networks. They resemble advantageous tools for multi-analyte detection. Future considerations will focus on widening the analyte spectrum, further reduce of the assay time, and improve sensitivity and robustness of the sensors to evolve the preliminary prototypes to field systems for on-site environmental monitoring.