The umu-test was developed for the detection of effects of chemical mutagens and carcinogens in environmental samples. It is performed according to ISO 13829 with Salmonella choleraesius subsp. chol. (strain TA1535/pSK1002). By automating the entire test, large numbers of toxicants and environmental samples as well as more treatments and parallels can be tested and, additionally, only low sample volumes are needed. In this work, an automated umu-test has been set up by installing a robotic XYZ-platform and a microplate reader inside a cabin. The use of established technical equipment for the automation in combination with a performance according to ISO standards was the essential aim of the approach. After initial preparation, the test is conducted software-controlled, follows the standard and fulfils the validity criteria of the standard procedure. For the optimization of the automated test umu-tests with one concentration of methyl methanesulfonate (MMS) of 166.7 mg/L were carried out. After optimization of incubation and pipetting conditions in the automated test, dose-response curves of various chemicals and environmental samples were assessed. The results of the automated umu-test have been compared with those of the standard manual test. The aim of the study was to show the applicability of an automated test system for the assessment of the genotoxic effects of various chemicals and environmental samples. During optimization, tests with 166.7 mg/L of MMS in every well of the microplate are carried out. Chemicals with different physical, chemical and toxicological properties are applied in both test systems. Water samples from different waste water treatment plants, and water extracts of contaminated and uncontaminated soils are assessed in the umu-test. The test is performed in parallel manually according to the standard and automatically using the robotic platform. Dose-response relationships and DLI-values are recorded and compared. The umu-test is applied on a RoboSeq 4204 SE pipetting station (MWG AG, Ebersberg, Germany). The robot is equipped with four holders for disposable tips to avoid undesired mixing of liquids while testing. With this system, it is possible to pipette all liquids. Photometric measurements are performed using a microplate reader. The pipetting station and the photometer are placed in an incubation cabin. According to the standards, exposure and growth of the bacteria are performed at 37 degrees C and the enzyme activity is assessed at 28 degrees C. Since both temperatures can't be adjusted simultaneously in the cabin, the test is performed in deviation from the standard at 33 degrees C. The results show that both the testing of non-volatile substances with strong or moderate genotoxic effects and the testing of water and soil samples in the automatic system work very well. Nevertheless, it is still difficult to characterize volatile chemicals automatically. This is illustrated, e.g. by testing 2- +/- -hydroxyquinoline. In this case, the chemical would not be assessed genotoxically after automatic performance of the umu-test. Sealing of the microplates, for example, avoids the loss of volatile substances, but this step of the procedure can not be performed automatically. Discussion. Only very few studies deal with the automation of bioassays. Eisentraeger et al. (2004) showed the suitability of a miniaturized and automated algae test for the testing of large numbers of environmental samples. Genotoxicity with an automated liquid handling are introduced by White et al. (1996) and Janz et al. (1989). A complete automation including liquid handling, incubation and photometric measurement is a new approach and leads to satisfying results. Optimization and suitability of the automated test are demonstrated in this study. Induction rates and growth factors do not differ significantly if the incubation temperature and the pipetting mode are optimized. Due to flexible scripting, the newly developed automated test system can also be used to perform other genotoxicity tests. Most results clearly show that genotoxicity tests can be automated completely allowing rapid testing that can be performed over night, for instance. Nevertheless, the test performance has to be optimized step by step depending on the technical characteristics of the automat. In order to overcome these technical problems, detailed knowledge of the hardware and the software of the automat and of the respective genotoxicity test system are needed. During this process of automation, it is very useful to test one genotoxic substance with an identical concentration in every well. The automated test system based on the RoboSeq 4204 SE pipetting station (MWG AG, Ebersberg, Germany) still has to be optimized with respect to the testing of volatile compounds. There is a need for removable, gas-tight microplate covers. For non-volatile chemicals and environmental samples, it can be used routinely. Nevertheless, the experiences made during this study can only partly be transferred to other robotic platforms and other bioassays, and automation of bioassays still can be a time-consuming matter.
Read full abstract