Massachusetts Institute of Technology, United States
Introduction: Raman spectroscopy has been demonstrated for its applicability for microbial identification, including various bacterial species at strain-level specificity and single-cell resolution. Most Raman-based identification required additional sample preparation protocols and/or special substrates, apart from standard liquid and solid agar culture, to remove/minimize the influence of background Raman signal from agar and plastic petri-dishes and focus on those arising from cells. However, these approaches cannot be easily implemented in current practices at clinical and microbiology labs in combination with other standard genetic sequencing-based methods, mainly due to additional steps required for Raman measurement. We address this issue by demonstrating bacterial identification directly from unopened agar plates in the same preparation as bacterial culture and storage. We tested with Escherichia coli (E. coli) of two variants, with and without green fluorescent protein (GFP) tagged on the same strain grown on LB agar plates. Compared to the spectra focused directly on E. coli colonies placed in an upright orientation with the lid removed, Raman spectra collected through the plate (approximately 0.5 mm) and solid agar (2-4 mm thickness) showed comparable signal-to-noise ratio (SNR) while preserving characteristic peaks of E. coli. Our approach can recover Raman signal of E. coli colonies, positioned 3-5 mm below the surface preserving spectral quality, and distinguish GFP-tagged cells. Also, Raman spectra were collected through a 10× objective lens, facilitating high-throughput measurement that allows multi-millimeters of working distance and relatively low demands for precise focusing onto the sample, which can be compatible with portable, handheld Raman devices.
Materials and
Methods: E. coli strain Seattle 1946 cells purchased from ATCC (#25922 and #25922GFP for non-GFP and GFP tagged cells, respectively) were used in this study. Both types were streaked onto quartz (∅ 1 mm)-bottomed petri dishes (∅ 35 mm) filled with solidified LB agar media from frozen stock and incubated at 37°C without shaking for 24 hours. The Raman spectra of E. coli colonies prepared on solid agar plates were collected from two different orientations: 1) plate facing upside down with the lid closed and sealed followed by; 2) plate placed upright with open lid. We used a commercial Raman microscope (WITec Alpha300) equipped with a 785 nm laser. Colonies were illuminated for 100 s and single accumulation at 50 mW. Raman spectra were acquired using a 10× objective and recorded with a spectrometer connected via a multimode optical fiber and electron multiplying CCD. Collected Raman spectra were subject to pre-processing, including cosmic ray removal and denoising followed by baseline correction and normalization.
Results, Conclusions, and Discussions: We could identify Raman peaks characteristic to E. coli. at 1002, 1447 and 1662 cm-1 from both approaches, which are attributable to phenylalanine, CH2 bending of protein and lipids, and Amide I, respectively (Figure 1A). The collected Raman spectra are subjected to principal component analysis (PCA), which showed clear separation between the presence of GFP within E. coli regardless of the orientation, indicating that our approach not only provides uncompromised spectral quality, but also can objectively identify E. coli colonies of two different variants (Figure 1B). Our proposed approach will open up a new avenue for direct bacterial identification using Raman spectroscopy in a non-destructive manner that can be seamlessly integrated in clinical and lab settings by eliminating concerns for potential contamination, sample damage, and extra preparation requirements. Also, we could prevent issues with agar dehydration, which would otherwise necessitate frequent re-focusing on the colony surface in-between each measurement. We proposed a direct Raman measurement approach for bacterial identification in the same condition as its growth, prepared on a solid agar plate and placed upside-down orientation with the lid covered. This new method enables to distinguish E. coli colonies with respect to their presence of GFP tagging while discharging concerns for potential sample contamination, agar dehydration, and frequent re-focusing demands.
Acknowledgements (Optional): J.H.K. was supported by the MIT Postdoctoral Fellowship for Engineering Excellence.
.jpg "Jeong Hee Kim, PhD (she/her/hers) photo")
