Development of FT-IR spectroscopy methods for detection of select pathogens from foods

Yashodhar Burgula, Purdue University

Abstract

Testing of foods to ensure safety and quality is the cornerstone of a successful food quality assurance program. Current rapid tests for pathogens include techniques such as ELISA, PCR, and immunoassays. Despite developments in these techniques there continues to be an increase in the incidence of foodborne illnesses and outbreaks with Salmonella spp. and E. coli O157:H7 responsible for 30.6% and 2.9% of the foodborne illness deaths in the US respectively and costing $3.1 billion in annual medical expenses, productivity losses, and costs for premature deaths annually. Thus, there is still a need for rapid pathogen detection. Fourier-transform infrared spectroscopy (FT-IR) in combination with multivariate statistical analysis is able to identify, distinguish, and classify microorganisms. However, barriers to widespread use of FT-IR include: sample preparation procedures, preparation time, sensitivity, interference from water absorbance, and analytical tools used. The overall goal of this research was the development of sample preparation and FT-IR spectroscopy in combination with chemometrics for identifying, differentiating, and quantifying Salmonella Typhimurium and E. coli O157:H7 from culture media and select foods. In the development phase, sample preparation methods for FT-IR analysis (filtration and Dynabeads ®) were evaluated for their sensitivity and specificity. Detection limits were 5.6 log CFU/mL (Dynabeads®) and 6.0 log CFU/mL (filtration) for E. coli O157:H7 requiring 6 h incubation of 300-500 CFU/mL. FT-IR analyses were able to differentiate the effects of processing treatments (heat, UV, salt, alcohol, and antibiotics), and between live and dead cells of E. coli O157:H7 and E. coli K12. In the calibration phase, E. coli O157:H7 and Salmonella Typhimurium were isolated from foods (fruit juices, ground beef, chicken rinse water, and Cheddar cheese) and limits of detection were established based on amide II peak (1589--1493 cm-1 ) area and partial-least squares (PLS) method (dilution series: 10 8--104 CFU/mL). In the validation phase, artificially contaminated samples of the select foods containing low levels of pathogens were detected when bacterial concentration increased to detection limits after 6 h of incubation at 37°C by the PLS model. A portable sensor (WilksIR) was evaluated and was less sensitive than the FT-IR main bench. However the portability and minimal equipment cost for the WilksIR could make it attractive for some applications.

Degree

Ph.D.

Advisors

Mauer, Purdue University.

Subject Area

Food science

COinS