Detection of ESBL-Producing Bacteria in the Neonatal Unit at Muhimbili National Hospital Using PNA-FISH and Smartphone-Based Fluorescence Microscopy

Abstract

Sub-Saharan Africa has the highest neonatal mortality rate, accounting for 43% of the global neonatal deaths. This is due to bacterial infections caused by bacteria such as Escherichia coli (EC) and Klebsiella pneumoniae (KP). The global threat of antimicrobial resistance complicates the treatment of bacterial infections, particularly those caused by Extended-spectrum beta-lactamase (ESBL)-producing bacteria. ESBL-producing bacteria destroy most commonly prescribed antibiotics for the treatment of bacterial infections. This project aimed to develop and implement a fluorescence in situ hybridization (FISH)-based assay to identify ESBL-producing bacteria in neonates (0-28 days) at Muhimbili National Hospital (MNH), Tanzania. The two target bacteria was EC and KP which were detected with a fluorescently labeled peptide nucleic acid (PNA) probe. Starting with a preparatory study at Chalmers University of Technology (CTH), initial tests were performed with the PNA probe for detection of EC and KP. Additionally, tests were conducted to evaluate whether the PNA-FISH assay could be performed faster, at lower temperatures, and with fewer chemicals. In the second phase at MNH, the PNA-FISH assay was established to detect for ESBL-producing EC in all admitted and discharged neonates over four days. Results analysis was done using a simple fluorescence microscope commonly found in labs in low and middle-income countries. The preparatory study at CTH concluded that the EC probe showed good results, with high fluorescence signal, for detection of EC, while the KP probe showed a low fluorescence signal for detection of KP. This could be due to low degree of hybridization. For EC, the hybridization temperature should not exceed 50°C, with optimal results even at 22°C for a laboratory strain of EC. Reducing the hybridization time from 1 hour to 15 minutes for the EC PNA-probe did not compromise with the quality of results from what could be concluded visually. Reducing the number of chemicals used in the PNA-FISH assay did not impact the results either. However, similar tests must be performed on fecal samples before definite conclusions can be drawn. The field studies at MNH showed that the PNA-FISH results of all patient samples matched the expected outcomes from the project’s verification step. The project’s findings did not only demonstrate the potential for using PNA-FISH as a diagnostic tool, but also stressed the importance of developing simple tools to detectantibiotic resistant bacteria in low-income countries, where it is most needed.