An outer membrane protease based approach for detection of E. coli in water and food

• Main Author: Sinsinbar, Gaurav
• Other Authors: Bo Liedberg
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Science::Medicine::Biosensors; Engineering::Materials
• Online Access: https://hdl.handle.net/10356/136589
• Institution: Nanyang Technological University

Bacterial pathogens such as E. coli and Salmonella are major contributors to food and waterborne diarrhoeal illness throughout the world. According to a 2015 WHO report, nearly 0.42 million people worldwide die each year because of the consumption of contaminated food or water. The spread of foodborne infection outbreaks can be restricted, preventing the numbers of deaths caused, by early detection of these bacterial pathogens in contaminated food and water. However, most commercially available bacterial detection kits rely on lab-based approaches, which are time-consuming, labour-intensive, or require dedicated instruments with a trained user. This thesis describes the development of a simple and rapid colourimetric assay for detecting E. coli in water and food. This is the first reported E. coli detection method that relies on activity of an outer membrane protease (OmpT) while employing unlabelled peptides. OmpT, a protease present on the outer membrane of E. coli, acts as a defense mechanism for E. coli against human immune response and cleaves cationic antimicrobial peptides (e.g. LL37) at dibasic site. As reported in the literature, obtaining a fully functional wild type OmpT via in vitro recombinant protein expression is extremely difficult due to its propensity to undergo autoproteolysis. This is because, OmpT carries di-basic sites within its own sequence making it prone to autoproteolysis. However, the findings from the current study suggest that during in vitro protein expression, the properly folded protein does not undergo autoproteolysis but it is only the misfolded protein that is prone to autoproteolysis. Hence, efforts to improve protein refolding can significantly limit in vitro autoproteolysis of wildtype OmpT. This study as well as previous investigations suggest that lipopolysaccharide (LPS) is crucial for protease activity. LPS activates OmpT and is reported to bind to a predicted-binding site composed of five amino acids. However, mutating these five residues did not diminish OmpT activity, suggesting that the predicted LPS binding residues on OmpT are not critical for OmpT activity or that LPS might have more than one binding site on OmpT. OmpT's substrate preference was investigated by designing, synthesizing and screening peptide libraries to identify the most optimal residues at the catalytic site. A high throughput mass spectrometry-based technique, SAMDI-MS, was employed to screen rationally designed peptide arrays composed of tetrapeptide sequences whose P2, P1, P1’ and P2’ positions were varied from a starting peptide sequence, ARRA. The data obtained from screening such a peptide array suggested an absolute requirement of dibasic amino acids at P1 and P1’ position and between lysine and arginine, OmpT preferred arginines at both P1 and P1' positions. This first array was useful in determining the preferred residues at each of the four positions. However, it is recognized that there is often an interdependence of amino acids at each position in the peptide sequence, and therefore arrays having two variable positions would be expected to reveal still more active substrates. So, a larger peptide array was generated by simultaneously varying the amino acids at P2 and P2’ and whose screening led to the identification of a tetrapeptide sequence, FRRV with high catalytic activity for OmpT. Further optimizations were performed by incorporating FRRV into a fragment of the bacteria’s natural substrate, yielding a peptide sequence for which OmpT exhibited a catalytic efficiency of 6.1 x 106 M-1s-1, a 400 fold improvement in catalytic efficiency over the natural sequence. Thus, E. coli expressing OmpT on its surface can efficiently cleave this optimized 15-mer peptide into two shorter fragments of seven and eight residues. By combining the (1) proteolytic activity of OmpT, (2) an optimized peptide substrate specific for OmpT, and (3) an anionic polythiophene (PT) as a reporter molecule, an assay was designed and developed to detect E. coli in water samples. Mixing the unlabelled peptide substrate and PT in an aqueous buffered solution resulted in a significant increase in PT’s fluorescence intensity with a concomitant colour change from orange to yellow that is detectable via naked eye. However, exposing the unlabelled peptide substrate to E. coli (or OmpT) prior to mixing it with PT resulted in no colour change, thus facilitating the detection of E. coli. Circular dichroism data confirmed that this change in the optical properties of PT coincides with a conformational change in the secondary structure of both the PT and the peptide. Unlike the intact peptide, the cleaved peptide fragments were unable to bring about a conformational change in the PT. Besides relying on the proteolytic activity of surface OmpT on E. coli, the assay also relies on the ability of PT to distinguish between the intact and cleaved peptides. Using this assay, a detection limit of as little as 1 CFU/mL of E. coli (K12 and J96 strains) was achieved within 6 and 8 h, respectively. With an assay time that is ~70 % lower when compared to the standard 24 h for most commercially sold home water testing kits that are based on β-galactosidase activity, this OmpT-Pep-PT assay is clearly a significant improvement towards a faster, easier, and more sensitive E. coli detection method.