Development of Innovative Biosensor Makes Progress

The prototype biosensor is inexpensive to build and is currently operated and powered by a laptop computer. The next-generation will feature integrated electronics and display.

Traditional testing methods for microorganisms in the food industry are relatively costly and time-consuming. The most common testing methods require obtaining an environmental or product sample (such as a swab or a rinse), then incubating that sample in media until enough microorganisms exist to be readily detected on culture plates or using conventional immunoassays, a process that can take several days. Other methods, such as amplification using polymerase chain reaction (PCR), may provide faster detection but require complicated procedures, expensive reagents and equipment, and extensively trained personnel. Nearly all of these methods require samples to be manually collected on the plant floor and taken to a laboratory for subsequent analysis.

Food processors, and the poultry industry in particular, have long been interested in faster, less expensive, and more dynamic testing methods. Recent consumer and government emphasis on food safety has elevated the urgency of this need for better microbial testing methods. In addition, testing will allow statistical process control optimization of intervention procedures leading to decreased processing costs. The challenges of such testing for food microbiology include the complexity of the sample matrix, the low numbers of organisms present, the effect of processing that may result in cell injury but not death, and the inherent inhomogeneity of sampling protocols. Furthermore, because 95 percent of analyzed samples are negative for detectable levels of pathogens, on-line process screening with subsequent confirmatory testing is cost-effective and efficient.

The goal of Georgia Tech’s biosensor project, lead by Senior Research Scientist David Gottfried, is to develop sensitive screening technologies that will allow rapid and affordable detection of pathogenic microorganisms. The research team is using optical waveguides and immunoassay methods to detect low concentrations of bacteria in a liquid sample. This technology has proven detection sensitivity greater than that of enzyme-linked immunoassay (ELISA), is less expensive and complex than PCR techniques, and has a rugged design to allow its use in on-line process control.

Because the device is based on a direct immunoassay protocol, compared to a conventional multi-step sandwich assay, sample preparation is minimal. Samples are injected with a small pump into a fluidics cell covering the waveguide surface. In a recently developed and laboratory-tested prototype portable sensor, a laser diode generates a beam of light, which is directed through the waveguide, under a layer of antibodies, and to the detector. If microorganisms with the correct antigens are present in the sample, they will bind to the antibodies on the waveguide surface, changing its index of refraction, and altering the optical properties of the light beam. The magnitude of this change provides information on the concentration of microorganisms in the sample. The researchers have been evaluating and refining the chemistry of the antibody coupling, the prototype’s opto-mechanical system, and the real-time analysis algorithm.

In initial work with Salmonella, the research team detected intact bacteria at concentrations from 104-107 cells/ml within 30 minutes. No antibody labeling or extended incubation steps are required. Selectivity using a monoclonal antibody was confirmed, and quantitative reproducibility of the assay was measured using independently prepared sensor chips and found to be within 10 percent over several orders of magnitude in concentration. Dried waveguides were prepared and evaluated for long-term storage ability. Finally, samples of poultry chiller water from a processing plant were spiked with Salmonella for detection within a more realistic matrix, indicating the applicability of the biosensor for in-plant operation.

Although there has been a 24 percent decrease since 1996, Campylobacter is now the most common cause of foodborne bacterial infection, affecting more than 2 million people in the United States each year (95 percent of cases arising from C. jejuni). The disease is usually mild, self-limiting, and rarely fatal. One of the most identified risk factors for sporadic cases comes from undercooked or mishandled poultry products; however, raw vegetables and bottled water are also associated with infection.

Food safety and quality control specialists are hampered by the difficulties due to microaerobic culturing conditions and inefficient enumeration of Campylobacter from food samples. Typical methods (in 1995 there were nearly 50 different culture media with no standardization) require 4-5 days and cannot guarantee the recovery of injured cells.

During the previous project year, Georgia Tech Research Scientist Jie Xu developed a direct biosensor immunoassay for Campylobacter to complement the studies of Salmonella. In laboratory experiments, sensitivity of 1,000 cells/ml (a factor of 100x better than conventional ELISA methods) was achieved using commercially available polyclonal and monoclonal antibodies.

The Georgia Tech researchers eventually hope to be able to test for multiple microbes simultaneously using a multi-assay optical waveguide. The design and fabrication of the optical waveguide chips as well as the optical and electronic components to accommodate up to four independent, simultaneous assays have begun. The researchers also plan to refine the device’s fluidics for increased cell capture efficiency, and will design a system for on-line screening and data collection to conduct field trials on a processing line.