A lectin-coupled porous silicon-based biosensor: label-free optical detection of bacteria in a real-time mode

Abstract

AbstractAccuracy and speed of detection, along with technical and instrumental simplicity, are indispensable for the bacterial detection methods. Porous silicon (PSi) has unique optical and chemical properties which makes it a good candidate for biosensing applications. On the other hand, lectins have specific carbohydrate-binding properties and are inexpensive compared to popular antibodies. We propose a lectin-conjugated PSi-based biosensor for label-free and real-time detection of Escherichia coli (E. coli) and Staphylococcus aureus (S. aureus) by reflectometric interference Fourier transform spectroscopy (RIFTS). We modified meso-PSiO2 (10–40 nm pore diameter) with three lectins of ConA (Concanavalin A), WGA (Wheat Germ Agglutinin), and UEA (Ulex europaeus agglutinin) with various carbohydrate specificities, as bioreceptor. The results showed that ConA and WGA have the highest binding affinity for E. coli and S. aureus respectively and hence can effectively detect them. This was confirmed by 6.8% and 7.8% decrease in peak amplitude of fast Fourier transform (FFT) spectra (at 105 cells mL−1 concentration). A limit of detection (LOD) of about 103 cells mL−1 and a linear response range of 103 to 105 cells mL−1 were observed for both ConA-E. coli and WGA-S. aureus interaction platforms that are comparable to the other reports in the literature. Dissimilar response patterns among lectins can be attributed to the different bacterial cell wall structures. Further assessments were carried out by applying the biosensor for the detection of Klebsiella aerogenes and Bacillus subtilis bacteria. The overall obtained results reinforced the conjecture that the WGA and ConA have a stronger interaction with Gram-positive and Gram-negative bacteria, respectively. Therefore, it seems that specific lectins can be suggested for bacterial Gram-typing or even serotyping. These observations were confirmed by the principal component analysis (PCA) model.