Development of novel molecular methods for the detection of C. difficile infections

BACKGROUND: C. difficile-associated infection (CDI), particularly in hospital patients has led to an increase in mortality and morbidity rate in US, UK and Europe. Virulence is mainly dependent on the expression of two key C. difficile-specific proteins, toxin A (TcdA) and toxin B (TcdB). Current CDI diagnostic is by ELISA or polymerase chain reaction (PCR); the former is limited in terms of sensitivity the latter in terms of clinical relevance, as detection of bacterial DNA is not informative about viability or whether the bacteria express toxins. Hence the development of this project, which aims to combine the clinically relevant information provided by an antibody-based test with the sensitivity of a PCR assay by using the proximity ligation assay (PLA) for detection of C. difficile TcdA and TcdB. PLA detects proteins via their interaction with pairs of antibodies coupled to noncomplementary DNA oligonucleotides. The binding of both antibodies to their target protein brings the oligonucleotides into proximity, allowing them to be bridged by a third oligonucleotide with complementarity to the other two. This facilitates their ligation and the detection of the resulting amplicon by real-time quantitative PCR (qPCR) acts as a surrogate marker for the protein of interest. Hence PLA has potential as a clinically relevant diagnostic tool for the detection of pathogens where nucleic acid based tests are inconclusive proof of infection. METHODS: We prepared monoclonal and polyclonal PLA probes targeting purified C. difficile toxins A (TcdA) and B (TcdB) and also targeting TcdA and TcdB spiked in canine faeces. Further evaluation of the assay was also done targeting TcdA and TcdB in clinical faeces and swab samples. Hydrolysis probe-based qPCR as well as digital PCR (dPCR) assays were used to detect antibody/antigen interactions. RESULTS: The performance of the PLA assays was antibody-dependent but both TcdA and TcdB assays were 10X more sensitive than comparable ELISAs in either single or duplex format when detecting purified toxins and spiked canine faeces shows sensitivity similar to ELISA performed in our lab. But the assay did not show sufficient sensitivity when evaluating the clinical faeces and swab samples. Both PLAs could be performed using single monoclonal antibodies coupled to different oligonucleotides. Finally, we used digital PCR to demonstrate accurate and reliable quantification of TcdA by digital PLA (dPLA). CONCLUSIONS: PLA has potential as new diagnostic applications for the detection of C. difficile. Further optimization of an assay is required to develop the assay for the detection 3 of TcdA and TcdB in clinical samples. Once this assay is developed into a diagnostic kit for C. difficile TcdA and TcdB, PLA can be used for further development of an assay for other pathogenic organisms where nucleic acid based tests do not indicate viability or expression of toxins, resulting in more targeted clinical decision-making, helping reduce the mortality rate for high-risk individuals. Importantly, since it is not always necessary to use two different antibodies, the pool of potential antibodies useful for PLA diagnostic assays is vastly enhanced. Finally, in the future, the combined testing of DNA and protein targets from the same sample on the same analytical platform (i.e. qPCR) may further improve the sensitivity and specificity of disease diagnosis leading to improved clinical outcomes, patient satisfaction and reduced associated costs.