You are here : Home > CREAB Team > Exploration of alternative methods for bacteria detection in blood

Vincent Templier

Exploration of alternative methods for bacteria detection in blood

Published on 4 November 2016
Thesis presented November 04, 2016

Abstract:
The presence of bacteria in the blood, a normally sterile environment, can cause dramatic consequences for an organism. In order to diagnose this infection, called bacteremia, the identification of the microorganism present in blood must be performed. Furthermore, proper diagnosis enables the administration of a suitable antibiotic therapy. Blood complexity as well as the low bacterial load, usually lower than 1 CFU.mL-1, make the diagnosis of this infection quite challenging. Indeed, most identification methods begin only after the blood culture turns positive due to their insufficient sensitivity. For this they require incubation of a large blood sample volume (20 – 30 mL) in specific culture media that allows bacterial growth above their detection limit. Therefore, its increases considerably the time of diagnosis, which usually takes between 2 and 48 hours and sometimes even more time after blood culture positivity depending on the method and the microorganism present in blood. A reduction of the time required for identification would have a positive impact for both the patient and the healthcare systems by reducing selective pressure on resistant bacteria and hospitalization costs by giving proper treatment faster. In this work, the evaluation of a new strategy based on the identification of bacteria during their multiplication in the blood culture is presented. This method is based on Surface Plasmon Resonance imaging (SPRi) which enables real time and label-free measurements of interactions occurring between bacteria and specific probes. Alternative ligands like aptamers, innate immune proteins and vancomycin have been tested. Following this study antibodies have been chosen as the major specific probes in this work. Nonetheless, the presence of the staphylococcal protein A leads to false-positive results in all immunoglobulin G (IgG). Enzymatic cleavage to remove the constant fragment of antibody where protein A interacts and the use of chicken antibodies (IgY) for which protein A has no affinity have been evaluated. Both methods allow to get rid of protein A interactions in pure culture media. But the presence of human serum in the media results in the total loss of signal. Our results show that interactions between blood components and staphylococcal proteins exposed at the bacterial surface, including the interactions between protein A and circulating antibodies, are responsible for this phenomenon. Solutions to alleviate this inhibition are discussed and tested. Detection experiments of another bacterial model, Salmonella enterica Serovar Enteritidis in blood culture media are presented. The crucial role played by the anticoagulant Sodium Polyanethole Sulfonate in non-specific interactions on antibodies is demonstrated. These interactions leading to a total loss of specificity for some antibodies are influenced by the isoelectric point (pI) of the probes which interact with this anionic compound and then attract blood components. After the partial resolution of this issue, we show the feasibility of detecting less than one bacteria per blood milliliter in a total volume of 32 milliliters, conditions close to real blood culture.

Keywords:
Surface Plasmon Resonance imaging (SPRi), S. aureus and S. enteritidis, Bacteremia, Blood, Rapid identification, Probes

On-line thesis.