Abstract
Surface-enhanced Raman spectroscopic (SERS) approaches are emerging for bacteria analysis whereby bacteria cells can be measured based on their biochemical composition (label-free) or with the aid of a chemical label to enhance the SERS signal. Combining a microscope, SERS microscopy is capable of imaging bacteria populations en masse based on specific spectrophotometric peaks. Here, we compared the label-free and label-based approaches to study Escherichia coli O157:H7 that was utilized as a model bacterium for SERS imaging analyses. Gold (Au) nanoparticles were utilized to enhance Raman scattering during this study and 3-mercaptophenylboronic acid was utilized as a model chemical label for comparison against label-free conditions. The result shows that SERS images of bacteria cells yielded measurable differences in precision, depending on the application of chemical labels. Chemical labels enabled SERS imaging of whole bacteria populations with single-cell precision. Bacteria coated with labels were also easier to bring into focus using high-magnification optical microscopy, without the need for immersion oil. Label-free analyses of single-cells were lower in geographic precision but provided opportunities to study the natural biochemistry of bacteria cells with strong accuracy. SERS analyses of label-free bacteria cell components were conclusively improved in vitro on a time-dependent basis. This concept can serve as an important benchmark when biochemically profiling or characterizing bacteria cells based on SERS. Electron micrographs proved that chemical labels can be utilized to increase nanoparticle contact with bacteria cells and reduce free nanoparticles that contribute to background noise in SERS spectra. We also demonstrate the use of both 3-mercaptophenylboronic acid and propidium iodide to discriminate live and dead bacteria through the simultaneous collection of data from these two chemical labels. Label-free approaches to SERS bacteria analyses are better suited for biochemical characterization and label-based approaches are better suited when accounting for individual cells among a population.