Rapid and Reliable Identification of Staphylococcus aureus Capsular Serotypes by Means of Artificial Neural Network-Assisted Fourier Transform Infrared Spectroscopy

Abstract

Staphylococcus aureus capsular polysaccharides (CP) are important virulence factors and represent putative targets for vaccine development. Therefore, the purpose of this study was to develop a high-throughput method to identify and discriminate the clinically important S. aureus capsular serotypes 5, 8, and NT (nontypeable). A comprehensive set of clinical isolates derived from different origins and control strains, representative for each serotype, were used to establish a CP typing system based on Fourier transform infrared (FTIR) spectroscopy and chemometric techniques. By combining FTIR spectroscopy with artificial neuronal network (ANN) analysis, a system was successfully established, allowing a rapid identification and discrimination of all three serotypes. The overall accuracy of the ANN-assisted FTIR spectroscopy CP typing system was 96.7% for the internal validation and 98.2% for the external validation. One isolate in the internal validation and one isolate in the external validation failed in the classification procedure, but none of the isolates was incorrectly classified. The present study demonstrates that ANN-assisted FTIR spectroscopy allows a rapid and reliable discrimination of S. aureus capsular serotypes. It is suitable for diagnostic as well as large-scale epidemiologic surveillance of S. aureus capsule expression and provides useful information with respect to chronicity of infection.