Detection and characterisation of bacteria causing lung infection in people with Cystic Fibrosis (CF) by surface-enhanced Raman spectroscopy (SERS)

Rapid and accurate identification of pathogens in CF could ensure prompt treatment with the most appropriate antibiotic; potentially improving outcomes and shortening hospital stays. As traditional culture methods for detecting bacteria are time-consuming there is a growing interest in SERS as a novel culture-free technique that produces a whole-organism spectroscopic fingerprint at high speed. Bacterial isolates including Pseudomonas aeruginosa (n=32), Staphylococcus aureus (n=5), Streptococcus pneumoniae (n=5) were incubated for 6 h at 37 °C/180 r.p.m., with a starting optical density (OD) of 0.15. After adjustment of the OD to 0.3, bacterial cells were harvested by centrifugation at 9000 rcf for 3 min and washed three times with dH2O. Bacterial pellets were mixed with citrate reduced silver colloid (CRSC) and dried. Spectra were recorded (4×10 s at 785 nm) and analysed within GRAMS/Al using Principal Component Analysis (PCA). Spectra of P. aeruginosa isolates (n=32) were separated into two distinct groups; the spectra of one group (n=12) was dominated by the pigment pyocyanin with vibrational brands present at 1350, 1492, 1598 and 1615 cm-1. The other group (n=20) had characteristic vibrational bands at 661, 735 and 800 cm-1 which correspond to guanine, adenine and uracil, respectively. S. aureus has a main characteristic band at 735 cm-1 and S. pneumoniae has a characteristic band present at 480 cm-1. Bacterial species clustered separately when analysed by PCA. Reproducible and distinguishable SERS spectra of bacterial isolates were obtained, and it was possible to differentiate between different bacterial species using PCA. These results suggest SERS has the potential to rapidly detect bacteria.