Abstract
A prototype submersible immunosensor with autonomous sampling characteristics has been designed and fabricated in conjunction with Sapidyne Instruments Inc. The watertight instrument is battery-powered and internally controlled; the internal controller can interface with an external computer for modification of the experimental parameters and review of results. An environmental sample is collected from the external space via a motor driven syringe such that displacement of the motor arm corresponds to a specific intake volume. Assay reagents, buffer and fluorescently labelled antibody, stored in bags within the sensor, are drawn into the syringe after the environmental sample and mixed. The final solution containing the environmental analyte and labelled antibody then passes over a flow/observation cell containing rigid 98-micron beads coated with the analyte of interest. The sensor continuously monitors the fluorescence across the flow cell and the difference in signal from the beginning to the end of the run can be converted to an estimate of analyte concentration. After optimisation steps that included selection of the fluorophore and bead support, the sensor could mix preloaded reagents and autonomously develop a standard curve for two different analytes: caffeine, a marker for untreated sewage, and hexavalent uranium, which contaminates the groundwater in the vicinity of uranium mining and processing sites. The coefficient of variation was near 15% for all concentrations examined. The minimum levels of detection for caffeine and hexavalent uranium in this assay system were 60 and 241 pM, respectively. Spike and recovery assays showed that the sensor was able to accurately predict the concentration of both analytes within the linear region of the calibration curve. Analysis of real environmental samples contaminated with uranium showed good agreement between the sensor and a standard analytical method, thus demonstrating the suitability and versatility of the submersible immunosensor as a field instrument.
Acknowledgements
This research was supported by Grant #N00014-06-1-1136 from the United States Office of Naval Research to the Tulane/Xavier Center for Bioenvironmental Research (D.A.B., PI of sensor subproject). Additional support was provided by the Tulane Phase II Katrina Fund. The authors thank Sara A. Morris of S.M. Stoller Corporation, Grand Junction, CO for providing the KPA analysis of Rifle groundwater samples and Terrance Lackie of Sapidyne Instruments for his assistance in sensor development.