Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Aug 22;8(1):12604.
doi: 10.1038/s41598-018-30407-8.

Raman Micro-Spectroscopy for Accurate Identification of Primary Human Bronchial Epithelial Cells

Affiliations
Free PMC article

Raman Micro-Spectroscopy for Accurate Identification of Primary Human Bronchial Epithelial Cells

Jakub M Surmacki et al. Sci Rep. .
Free PMC article

Abstract

Live cell Raman micro-spectroscopy is emerging as a promising bioanalytical technique for label-free discrimination of a range of different cell types (e.g. cancer cells and fibroblasts) and behaviors (e.g. apoptosis). The aim of this study was to determine whether confocal Raman micro-spectroscopy shows sufficient sensitivity and specificity for identification of primary human bronchial epithelial cells (HBECs) to be used for live cell biological studies in vitro. We first compared cell preparation substrates and media, considering their influence on lung cell proliferation and Raman spectra, as well as methods for data acquisition, using different wavelengths (488 nm, 785 nm) and scan protocols (line, area). Evaluating these parameters using human lung cancer (A549) and fibroblast (MRC5) cell lines confirmed that line-scan data acquisition at 785 nm using complete cell media on a quartz substrate gave optimal performance. We then applied our protocol to acquisition of data from primary human bronchial epithelial cells (HBEC) derived from three independent sources, revealing an average sensitivity for different cell types of 96.3% and specificity of 95.2%. These results suggest that Raman micro-spectroscopy is suitable for delineating primary HBEC cell cultures, which in future could be used for identifying different lung cell types within co-cultures and studying the process of early carcinogenesis in lung cell culture.

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1
Figure 1
Comparison of area and line scan data acquisition from A549 and MRC5 cells. (A) Area scan Raman and fluorescence imaging data at 488 nm. Clusters were derived using Manhattan analysis (pre-mode: derivative). Cluster analysis reveals the following assignments based on spectra presented in Supplementary Figure 2: Black (cluster 1) = area without the cells (background); Grey (cluster 2) = cell border; Green (cluster 3) = cytoplasm; Blue (cluster 4) = nucleus; Red (cluster 5) = endoplasmic reticulum/mitochondria; Orange (cluster 6) = lipid droplets. For comparison, the lipid distribution at 2888 cm−1 (sum filter: 2838–2938 cm−1) is shown relative to fluorescence Nile Red staining in A549, while the nucleus area represented by 2970 cm−1 (sum filter: 2920–3020 cm−1) is compared to NucBlue. Raman area scans of A549: scale bar is 10 μm (148 × 100 points, 0.1 s per pixel, ~25 min per image); MRC5: scale bar is 9 μm (100 × 110 points, 0.1 s per pixel, ~20 min per image). (B) Comparing average of single Raman spectra along a line passing through the center of the cell (blue) to the full cell area scan (red) from A549 provides very similar results at 10 spectral samples, as shown in the differential spectrum (black). Average spectra were normalized to area under curve for this comparison.
Figure 2
Figure 2
Impact of longitudinal line-scan data acquisition at multiple time points on live cell spectra. (A) Average Raman spectra of 10 points taken along a line across an A549 cell from scan 1 to scan 4 for 488 nm (10 mW, 0.5 s, 30 accumulations, 30 minutes between scans). Raman peak intensity ratios were calculated across all 4 scans at 488 nm (B) and also compared as a function of time (C) for: I(784/1003), pyrimidine bases, DNA/phenylalanine; I(1003/1301), phenylalanine/CH2 twist of lipids; I(1003/1446), phenylalanine/CH deformation; I(1003/1654), phenylalanine/mixed Amide I protein and C = C stretching of lipids; I(1446/1654), CH deformation/mixed Amide I protein and C = C stretching of lipids; and I(2854/2930) CH2 stretching of lipids/ CH3 symmetric stretching of proteins. (D) Average Raman line-scan spectra from scan 1 to 8 for 785 nm (120 mW, 1 s, 30 accumulations, 15 minutes between scans). Raman peak intensity ratios except for I(2854/2930) were calculated across all 8 scans at 785 nm (E) and also compared as a function of time (F). I(2854/2930) was inaccessible at 785 nm due to the installed diffraction grating. Fewer replicate scans were performed at 488 nm to avoid photodamage while enabling comparable time points to 785 nm. Spectra acquired using line-scan method.
Figure 3
Figure 3
Comparison of spectral features of A549 and MRC5 immortalized cell lines. Mean and standard deviation (shaded area) of the non-normalized spectra recorded from 30 independent A549 (A) and MRC5 (B) cells are illustrated. Difference spectra are shown in (C) for direct comparison of the differences in molecular vibrations between the two cell lines. (D) Scatter plot of the score values of each single Raman spectrum for the second and third principal components (PCs) from the A549 (black triangles) and MRC5 cells (green circles). (E) Loadings plot of PC1, PC2 and PC3. Data acquired at 785 nm with 1 sec exposure and 30 accumulations per single spectrum. Total 300 spectra from 30 cells using the line-scan method for each cell line were analysed. Preprocessing PCA mode: normalization to area.
Figure 4
Figure 4
Comparison of spectral features of primary human bronchial epithelial cell (HBEC) lines. Mean and standard deviation (shaded area) of the non-normalized spectra recorded from 30 independent ATCC (A), LONZA (B) and PAP243 (C) cells are illustrated. Difference spectra are shown for direct comparison of the differences in molecular vibrations between ATCC and LONZA (D), ATCC and PAP243 (E) and LONZA and PAP243 (F). Data acquired at 785 nm with 1 sec exposure and 30 accumulations per single spectrum. Total 300 spectra from 30 cells using the line-scan method for each cell line were analysed.
Figure 5
Figure 5
Principal components analysis of primary HBECs. (A) Scatter plot of the score values of each single Raman spectrum for the second and third principal components from the ATCC (orange squares), LONZA (blue circles), and PAP243 (purple triangles). (B) Loadings plot of PC1, PC2 and PC3. Data acquired at 785 nm with 1 sec exposure and 30 accumulations per single spectrum. Total 300 spectra from 30 cells using the line-scan method for each cell line were analysed. Preprocessing PCA mode: normalization to area.
Figure 6
Figure 6
Partial least squares discriminant analysis (PLS-DA) of all Raman data MCR5, A549 and HBEC (ATCC, LONZA and PAP243). (A) Plot of calibration (Cal) and cross validation (CV) classification error average against the number of latent variables (LVs). (B) Plot of the loadings of LV1-LV4 against wavenumber. (C) Receiver operating characteristic (ROC) curves for the classification. (D) Scores plots on latent variables (LV1-LV4). Symbols: ATCC (orange squares); LONZA (blue circles); and PAP243 (purple triangles); A549 (black triangles); MRC5 (green circles). Preprocessing PLS-DA mode: vector normalization, 2nd derivative, smoothing and mean center. The spectra were split into calibration and validation sets by removing every fourth spectrum to form the validation set. Spectra acquired using line-scan method.
Figure 7
Figure 7
Variable Importance in Projection (VIP) scores for PLS-DA model of all Raman data. (A) VIP scores, shown offset for direct comparison in (B). Raman bands attributed to lipids (718, 1264, 1301, 1440 and 1658 cm−1), proteins (641, 1003, 1166/1174, 1239, 1580, 1658, 1674 cm−1), nucleic acids (784, 828, 1316, 1458 cm−1) and carbohydrates (881, 944, 1043, 1085 cm−1) can be observed to be the most important spectroscopic signatures for discrimination of the cell lines.

Similar articles

See all similar articles

Cited by 3 articles

References

    1. Bohndiek SE, et al. A small animal Raman instrument for rapid, wide-area, spectroscopic imaging. Proc. Natl. Acad. Sci. USA. 2013;110:12408–13. doi: 10.1073/pnas.1301379110. - DOI - PMC - PubMed
    1. Smith R, Wright KL, Ashton L. Raman spectroscopy: an evolving technique for live cell studies. Analyst. 2016;141:3590–3600. doi: 10.1039/C6AN00152A. - DOI - PubMed
    1. Schie I, Huser T. Methods and applications of Raman microspectroscopy to single-cell analysis. Appl. Spectrosc. 2013;67:813–28. doi: 10.1366/12-06971. - DOI - PubMed
    1. Charwat V, et al. Potential and limitations of microscopy and Raman spectroscopy for live-cell analysis of 3D cell cultures. J. Biotechnol. 2015;205:70–81. doi: 10.1016/j.jbiotec.2015.02.007. - DOI - PubMed
    1. Clemens G, Hands JR, Dorling KM, Baker MJ. Vibrational spectroscopic methods for cytology and cellular research. Analyst. 2014;139:4411–4444. doi: 10.1039/C4AN00636D. - DOI - PubMed

Publication types

Feedback