(482) Fourier transform infrared spectroscopy (FT-IR) for strain typing of Burkholderia cepacia complex and/or monitoring of strain carriage in chronically infected patients with cystic fibrosis

Background: Epidemiological surveillance of infections caused by Burkholderia cepacia complex (Bcc) became a cornerstone of infection control in many cystic fibrosis (CF) centers. Based on early detection and molecular typing of Bcc, we ceased the transmission of epidemic strain ST32 at our CF center in the 2000's. With the advance of new typing approaches, we aimed to evaluate whether the Fourier transform infrared spectroscopy (FT-IR) can replace the more expensive and time-consuming reference techniques such as multi-locus sequence typing (MLST). The advantage of FT-IR is low running cost, easy sample preparation and short turnaround time. We tested the discriminatory power of the FT-IR and its potential to i) identify individual strains of Bcc and ii) check whether a chronically infected patient harbors the same Bcc strain or not.

Methods: We examined 37 clinical isolates (each in technical triplicate) representing 11 different strain types (STs) of two Bcc species. Out of them, 21 isolates belonged to 7 patients (3 isolates per patient), covering their period of chronic Bcc infection from 3 to 8 years (median 4 years). Isolates were grown on tryptic soy agar for exact 30 hours at 37°C before sample preparation. 1µL loop of biomass from the confluent growth area was homogenized in 50 µL of 70% ethanol and 50uL of deionized water. 15µL of suspension was spotted in triplicate on a silicon microtiter plate, dried, and measured by IR-Biotyper (Bruker Daltonics, Germany). Using the built-in IR-Biotyper software (Version 3.1), we applied a default splicing method which enables to analyze the spectra of wavenumber ranging from 1300 to 400 cm^-1. The 2D scatter plots using the principal component analysis (PCA) and linear discriminant analysis (LDA) were used for spectra analysis.

Results: Initially, when a single isolate per patient was analyzed, we were able to distinguish all STs from each other using the PCA and LDA scatter plots (i.e., B. cenocepacia: ST32, ST234, ST241, ST628, ST643, ST761, ST834, ST1547 and ST184; B. contaminans ST102 and ST482). However, by including 21 longitudinal isolates from 7 patients (representing ST32 by 12 isolates; ST234, ST834, and ST102 by 3 isolates each) the variability of the individual ST spectra increased and overlapped (see Fig 1A). The longitudinal isolates of the same patient created clearly identifiable clusters (see Fig 1B) that matched their respective ST.

Conclusions: In our opinion, the FT-IR meets the requirements for the regular and repeated molecular typing of the Bcc in chronically infected CF patients as it can instantly recognize whether a patient carries the same strain or it was replaced with another one. However, the ST discrimination and thus, its separation from the others, is unreliable.

Acknowledgements:

References: