There has been a steep rise in the emergence of antibiotic-resistant
bacteria in the past few years. A timely diagnosis can help in initiating
appropriate antibiotic therapy. However, conventional techniques for
diagnosing antibiotic resistance are time-consuming and labor-intensive.
Therefore, we investigated the potential of Raman spectroscopy as
a rapid surveillance technology for tracking the emergence of antibiotic
resistance. In this study, we used Raman spectroscopy to differentiate
clinical isolates of antibiotic-resistant and -sensitive bacteria
of Escherichia coli, Acinetobacter baumannii, and Enterobacter species. The spectra were collected with or without exposure to
various antibiotics (ciprofloxacin, gentamicin, meropenem, and nitrofurantoin),
each having a distinct mechanism of action. Ciprofloxacin- and meropenem-treated
sensitive strains showed a decrease in the intensity of Raman bands
associated with DNA (667, 724, 785, 1378, 1480, and 1575 cm–1) and proteins (640 and 1662 cm–1), coupled with
an increase in the intensity of lipid bands (891, 960, and 1445 cm–1). Gentamicin- and nitrofurantoin-treated sensitive
strains showed an increase in the intensity of nucleic acid bands
(668, 724, 780, 810, 1378, 1480, and 1575 cm–1)
while a decrease in the intensity of protein bands (640, 1003, 1606,
and 1662 cm–1) and the lipid band (1445 cm–1). The Raman spectral changes observed in the antibiotic-resistant
strains were opposite to that of antibiotic-sensitive strains. The
Raman spectral data correlated well with the antimicrobial susceptibility
test results. The Raman spectral dataset was used for partial least-squares
(PLS) analysis to validate the biomarkers obtained from the univariate
analysis. Overall, this study showcases the potential of Raman spectroscopy
for detecting antibiotic-resistant and -sensitive bacteria.