Development of on-Site Applicable Immunosensor Combined with Light Microscopic Imaging System for the Detection of Salmonella in Poultry

The common sources of Salmonella outbreak are poultry and poultry products, meat and meat products, and eggs.  Despite extensive efforts to reduce Salmonella outbreaks, it is almost impossible to eliminate all the possible chances of Salmonella contamination.  Thus, the best prevention strategy is to monitor and detect Salmonella using rapid, sensitive, specific, and simple detection methods.  The objective of this study was to demonstrate the feasibility of the on-site applicable immunosensor combined with a light microscopic imaging system (LMIS) for the detection of Salmonella in chicken.

Since the selection of antibodies as a bio-recognition element is the most critical factor for the performance of the biosensor, the anti-Salmonella polyclonal antibodies (pAbs) were produced and purified.  The reactivity and specificity of purified anti-Salmonella pAbs were compared with those of commercial anti-Salmonella pAbs using an indirect ELISA.  The gold sensor platforms were modified with dithiobis-succinimidyl propionate (DSP) to enhance the immobilization of antibodies.  The schematic diagram for general preparation of immune sensor combined with LMIS is presented in Figure 1.  After the optimum concentration of antibody was determined, the immobilization of the antibody was confirmed using an atomic force microscope (AFM).  The detection limit of the developed immunosensor was determined by applying the sensor to detect Salmonella in chicken skin. 

Our results showed that there was no significant difference in the reactivity between the purified anti-Salmonella pAbs and commercial anti-Salmonella pAbs over the entire range of concentrations used.  The purified anti-Salmonella pAbs exhibited not only excellent specificity against 11 non-Salmonella strains but also excellent reactivity towards 20 other non-Salmonella serovars.  The optimal concentration of antibody for immobilization on the sensor was determined to be 100 μg/mL.  The AFM images confirmed the binding of the antibody on the sensor by providing irregular and globular surface features.  The detected number of Salmonella on the immunosensor was proportional to the concentrations of inoculated Salmonella numbers when the sensor was applied to chicken.  The number of detected Salmonella significantly increased throughout the inoculated concentration ranges of 10⁴–10⁸ cfu/chicken with a 100% positive rate (p < 0.05).  The detection limit and time of the detection method were determined to be 10³ cfu/sensor and within 1 h, respectively.  With the advantage of direct observation and enumeration of Salmonella on the immunosensor, the developed immunosensor combined with LMIS could be used to detect Salmonella in chicken as a specific, sensitive, simple, and practical detection method.