Translator Disclaimer
1 July 2008 Raman microscopy for dynamic molecular imaging of living cells
Author Affiliations +
Abstract
We demonstrate dynamic imaging of molecular distribution in unstained living cells using Raman scattering. By combining slit-scanning detection and optimizing the excitation wavelength, we imaged the dynamic molecular distributions of cytochrome c, protein beta sheets, and lipids in unstained HeLa cells with a temporal resolution of 3 minutes. We found that 532-nm excitation can be used to generate strong Raman scattering signals and to suppress autofluorescence that typically obscures Raman signals. With this technique, we reveal time-resolved distributions of cytochrome c and other biomolecules in living cells in the process of cytokinesis without the need for fluorescent labels or markers.

1.

Introduction

Researchers have long sought to improve optical methods of extracting information from living samples by fluorescence, light scattering, or other noninvasive techniques. Raman scattering can be used to optically investigate the chemical properties of samples due to its ability to detect capability of detecting molecular vibration frequencies that characterize molecular species, structures, and environmental conditions. In combination with optical imaging, Raman scattering can be applied to the direct sensing of biological molecules without requiring preprocessing or fluorescence staining of samples.1, 2, 3, 4, 5, 6 Although Raman scattering is a powerful tool for analyzing biomolecules, it has rarely been attempted as a contrast mechanism for imaging living specimens due to the extremely low scattering efficiency. Since typical Raman scattering signals are weak (scattering cross-section 1030cm2 ) compared to fluorescence yields (absorption cross-section 1016cm2 ), the measurement of Raman spectra usually requires long exposure times, making observations of living specimens difficult.

Here, we extend Raman scattering spectroscopy to enable dynamic imaging of molecular distributions in living cells with high temporal and spatial resolution. We combined a slit-scanning detection technique and optimized the excitation wavelength to image molecular distributions of cytochrome c, protein beta sheets, and lipids in unstained living HeLa cells. By investigating the wavelength dependencies of Raman signal yields and background signals, we found that 532-nm -wavelength excitation can be used to generate Raman scattering signals strong enough for imaging and for suppressing the background signals that result from autofluorescence. Additionally, cytochrome c exhibits resonant Raman scattering by 532nm excitation, and we found that in situ imaging of cytochrome c in living cells can finally be performed using our technique. These results show that the activities of molecules, including exogenous molecules such as in pharmaceutical drugs, can be directly monitored in living cells by Raman scattering to identify cellular functions that conventional fluorescence techniques are incapable of revealing.

2.

Raman Spectra of Living HeLa Cells

In order to select the excitation wavelength, we first compared Raman spectra of cultured HeLa cells obtained with different excitation wavelengths. We used the standard laser lines: 488nm from a semiconductor laser, 514.5nm from an Ar ion laser, 532nm from a frequency-doubled Nd:YVO4 laser, and 633nm from a HeNe laser. Wavelength choice is a tradeoff; shorter wavelengths provide excellent Raman scattering efficiency and spatial resolution, both of which are inversely proportional to wavelength, while longer wavelengths produce significantly less background autofluorescence signals than shorter wavelengths. We used a 1.2 numerical aperture (NA) objective lens, both for illuminating the cells and collecting the Raman signal, using a spectrophotometer (320PI, Acton) and a cooled CCD camera (PIXIS 400BR, Princeton Instruments).

Figure 1 shows the effect of the excitation wavelength on Raman spectra obtained from cytosol regions in living HeLa cells. For each excitation wavelength, we obtained Raman spectra from 36 different positions in the cytosol of a single cell and averaged them to produce each spectrum in Fig. 1, separated by wavelength with no background removal applied. The laser intensity at the focus was 4mWμm2 , and the 36 spectra were obtained in parallel over an exposure time of 20seconds . In Fig. 1, the Raman spectra obtained with the 488-, 514.5-, and 532-nm excitation wavelengths exhibited much stronger scattering signals than the spectra obtained with the 633-nm wavelength, which was expected due to the wavelength dependence of the scattering efficiency. The measured spectra contain peaks that are known to occur in biological samples, such as the ring breathing of phenylalanine ( 1000cm1 ), CH2 deformation ( 1451cm1 ) and CH2 stretching mode ( 2850cm1 , 2885cm1 ), CH3 stretching mode ( 2935cm1 ), and Amide-I vibrational mode of peptide bonds ( 1660cm1 ) (Refs. 7, 8).

Fig. 1

Raman spectra obtained from the cytosol of a living HeLa cell. Raman spectra obtained from 36 different points in the cytosol of one cell were averaged for each wavelength, showing the tradeoff between Raman peak signal strength and background autofluorescence. The cells were irradiated with laser light of 488-, 514.5-, 532-, and 632.8-nm wavelengths.

044027_1_014804jbo1.jpg

In addition to these typical Raman shifts, strong Raman peaks appear at 753, 1127, 1314, and 1583cm1 in the spectra obtained by the 514.5- and 532-nm excitation wavelengths. These peaks can be assigned to vibration modes of cytochrome c.9 Since cytochrome c contains a heme protein that absorbs light at 510to550nm , strong resonant Raman scattering is observed when irradiated with this wavelength range. The Raman peak at 753cm1 , which shows the pyrrole breathing mode ν15 in cytochrome c, was previously measured in vitro by 532-nm irradiation,9 and can be clearly observed in situ in this result. By observing the peaks at 753, 1127, 1314, and 1583cm1 , this technique can be used to detect cytochrome c in living cells by resonant Raman scattering. We also measured Raman spectra from the nuclear regions of living HeLa cells (not shown). The Raman spectra from the nuclei were similar, and no substantial spectral dependence on the excitation wavelength was observed, which shows that resonant Raman signals were not a significant contribution to the total Raman emission from the nuclear regions.

We also investigated the contribution of autofluorescence to the background signal in the Raman spectra, which is of particular interest since spectroscopic sensitivity is dramatically reduced by any background contributions due to autofluorescence. The flavin coenzymes FAD and FMN are known to be sources of autofluorescence in the detected wavelength range.10, 11 Lipofucin is another possible source of autofluorescence; however, it absorbs light predominantly in the UV region and is not significantly excited by the wavelengths used in this experiment.12 We measured the average fluorescence intensity of FAD at the regions between 600 and 1800cm1 for excitation wavelengths of 488, 514.5, and 532nm . We found that 532-nm excitation produced an autofluorescence signal approximately 12 times lower than 514.5-nm excitation, and 167 times lower than 488-nm excitation. The autofluorescence background signal for 532-nm excitation light was markedly decreased compared to 514.5-nm excitation, and the Raman scattering signals were of comparable strength, which shows that 532-nm excitation is superior for imaging living cell samples.

3.

Slit-Scanning Confocal Raman Microscopy

We used a home-made Raman microscope with 532-nm excitation and slit-scanning excitation and detection.13, 14 The slit-scanning technique allowed us to detect Raman spectra from different positions in parallel, and as a result, greatly improved the image acquisition rate. The sample was irradiated by a line-shaped focus, and Raman scattering signals from the illuminated line were imaged at the entrance slit of a spectrophotometer. Line illumination is also useful in reducing photodamage of the sample because the light intensity at the focal plane is much lower than that of single-focus scanning at the same exposure. Additionally, the slit of the spectrophotometer eliminates Raman scattering from out-of-focus planes, providing spatial resolution in three dimensions and improving of image contrast.15 To produce the line-shaped laser light, we used a cyrindrical lens and imaged the illumination line at the sample by a 1.2-NA water immersion objective lens.

4.

Raman Scattering Images of an Unstained Living Hela Cell

Using our slit-scanning Raman microscope, we obtained a hyperspectral image of living HeLa cells in the range of Raman shifts between 500cm1 and 3000cm1 . The cell was observed in a HEPES-buffered Tyrode’s solution composed of (in mM) NaCl, 150; glucose, 10; HEPES, 10; KCl, 4.0; MgCl2 , 1.0; CaCl2 , 1.0; and NaOH, 4.0. Then the cell was irradiated with a light intensity of 3.3mWμm2 at the focal plane. Singular value decomposition (SVD) was used for noise reduction, and we chose seven loading vectors that significantly contribute to the image contrast for the image reconstruction.4 Following noise reduction, we subtracted the fluorescence background signal from the Raman spectra at each pixel in the image by a modified polyfit fluorescence removal technique.16

The distribution of cytochrome c is reconstructed in Fig. 2a from the intensity distribution of the Raman peak at 753cm1 . Since cytochrome c is used for electron transfer in oxidative phosphorylation in mitochondria, the image has a contrast similar to the distribution of mitochondria. Figure 2b shows the Raman signal distribution at 1686cm1 given by the Amide-I vibration mode of peptide bonds in protein beta sheets.7 We chose this wave number because the C-C stretching vibration modes (1650to1670cm1) in hydrocarbon chains of lipid molecules overlap the shorter Raman shifts of the Amide-I band. Since beta sheets are commonly seen in proteins, Fig. 2b is strongly correlated with protein distribution in the cell, and consequently, there is a slightly higher protein concentration at the nucleus. Figure 2c shows the Raman signal distribution at 2852cm1 where the signal due to the CH2 stretching vibration is strongly detected from the hydrocarbon chain of lipid molecules. Although proteins and other biological molecules also contain CH2 , the image contrast is provided mainly by the lipid vesicles that are rich in lipid molecules.17 By combining these images via the different color channels of a single image, we obtained the distributions of protein beta sheets, cytochrome c, and lipid vesicles shown in Fig. 2d.

Fig. 2

Raman scattering images of unstained, unlabelled living HeLa cells reconstructed using the distribution of Raman signals at (a) 753cm1 , (b) 1686cm1 , and (c) 2852cm1 , showing the distribution of cytochrome c, protein beta sheet, and lipid molecules, respectively. Image (d) was constructed by merging images (a) through (c) with color channels. The sample was irradiated with a light intensity of 3.3mWμm2 at the focal plane in 78 lines of exposure. The exposure time of each line was 5sec , and the images consist of 78×281pixels .

044027_1_014804jbo2.jpg

5.

Raman Observation of Dynamic Distributions of Biomolecules

Time-resolved, dynamic molecular distributions are shown in Video 1 , where Raman images of a living HeLa cell were taken during cytokinesis. The image acquisition time for each Raman image was 185seconds , with an interval between images of 115seconds . The images in Video 1 were obtained with 48 line exposures of 1second each, for a total exposure time of 48seconds . The difference between the image acquisition time and the exposure time is due to the data transfer time from the CCD camera to the data storage computer. For Video 1, we also applied noise reduction by the use of SVD, and the images were reconstructed using five loading vectors. The contrast due to Raman scattering in Video 1 indicates that proteins exist in relatively higher concentrations at the chromosomes than in other parts of in the cells, which allows us to trace the progress of cytokinesis by the temporal variation of protein distribution. We also observe that highly concentrated cytochrome c appeared near the cleavage furrow, presumably to provide sufficient energy to the contractile ring that divides the cell into two. In addition, the movement of lipid vesicles associated with cellular dynamics during mitosis can be discerned. During extended observation, photodamage of the cell is a possibility; however, our results show that any photodamage which may have occurred was not significant enough to stop cytokinesis from proceeding.

Video 1

Time-resolved imaging of cytochrome c, protein, and lipid molecule distributions in a label-free HeLa cell during cytokinesis. The images were taken at 5-min intervals with a frame rate of 185sec /image. The progress of cytokinesis is recognized by the change in the distribution of proteins, and a high concentration of cytochrome c is observed at positions near the contractile ring. The sample was irradiated with light intensity of 3.5mWμm2 at the focal plane, and the images consist of 161×48pixel (MOV, 562KB). Video 1.

044027_1_014804jbov1.jpg
10.1117/1.2952192.1

6.

Conclusions

Using the Raman microscopy method described here, we demonstrated label-free observation of biological molecules in living cells using Raman scattering for a contrast mechanism. Label-free imaging provides us with opportunities to observe biological activities without the disturbances of labeling procedures and agents that usually degrade the viability of samples. It frees us from the photobleaching problems inherent in fluorescence staining techniques and allows us to obtain distributions of chemicals in samples that are impossible to stain or in locations where staining is undesirable. We also showed that the resonant Raman scattering of cytochrome c can be distinctly observed using a 532-nm wavelength for excitation, allowing label-free observation of cytochrome c distributed in living cells. Cytochrome c is a protein well known for its important role in the production of ATP in mitochondria, and the distribution of cytochrome c is thought to change drastically during apoptosis.18, 19 The Raman imaging technique demonstrated in this paper will allow in situ studies of the role of cytochrome c in apoptosis as well as its function in other cellular activities.

References

1. 

G. J. Puppels, F. F. M. de Mul, C. Otto, J. Greve, M. Robert-Nicoud, D. J. Arndt-Jovin, and T. M. Jovin, “Studying single living cells and chromosomes by confocal Raman microspectroscopy,” Nature (London), 347 301 –303 (1990). https://doi.org/10.1038/347301a0 0028-0836 Google Scholar

2. 

G. J. Puppels, M. Grond, and J. Grave, “Direct imaging Raman microscope based on tunable wavelength excitation and narrow-band emission detection,” Appl. Spectrosc., 47 1256 –1267 (1993). 0003-7028 Google Scholar

3. 

N. Uzunbajakava and C. Otto, “Combined Raman and continuous-wave-excited two-photon fluorescence cell imaging,” Opt. Lett., 28 2073 –2075 (2003). 0146-9592 Google Scholar

4. 

H.-J. Manen, Y. M. Kraan, D. Roos, and C. Otto, “Intracellular chemical imaging of heme-containing exzymes involved in innate immunity using resonance Raman microscopy,” J. Phys. Chem. B, 108 18762 –18771 (2004). 1089-5647 Google Scholar

5. 

Y.-S. Huang, T. Karashima, M. Yamamoto, and H. Hamaguchi, “Molecular-level investigation of the structure, transformation, and bioactivity of single living fission yeast cells by time- and space-resolved Raman spectroscopy,” Biochemistry, 44 10009 –10019 (2005). https://doi.org/10.1021/bi050179w 0006-2960 Google Scholar

6. 

J.-X. Cheng and X. S. Xie, “Coherent anti-Stokes Raman scattering microscopy: instrumentation, theory, and applications,” J. Phys. Chem. B, 108 827 –840 (2004). https://doi.org/10.1021/jp035693v 1089-5647 Google Scholar

7. 

A. T. Tu, Raman Spectroscopy in Biology: Principles and Applications, John Wiley & Sons Inc., New York (1982). Google Scholar

8. 

I. Notingher and L. Hench, “Raman microspectroscopy: a noninvasive tool for studies of individual living cells in vitro,” Expet. Rev. Med. Dev., 3 215 –234 (2006). Google Scholar

9. 

T. G. Spiro and T. C. Strekas, “Resonance Raman spectra of hemoglobin and cytochrome c: inverse polarization and vibronic scattering,” Proc. Natl. Acad. Sci. U.S.A., 69 2622 –2626 (1972). https://doi.org/10.1073/pnas.69.9.2622 0027-8424 Google Scholar

10. 

D. L. Heintzelman, R. Lotan, and R. R. Richars-Kortum, “Characterization of the autofluorescence of polymorphonuclear leukocytes, mononuclear leukocytes and cervical epithelial cancer cells for improved spectroscopic discrimination of inflammation from dysplasia,” Photochem. Photobiol., 71 327 –332 (2000). https://doi.org/10.1562/0031-8655(2000)071<0327:COTAOP>2.0.CO;2 0031-8655 Google Scholar

11. 

R. C. Benson, R. A. Meyer, M. E. Zaruba, and G. M. Mckhann, “Cellular autofluorescence—is it due to flavins,” J. Histochem. Cytochem., 27 44 –48 (1979). 0022-1554 Google Scholar

12. 

F. Schutt, B. Ueberle, M. Schnolzer, F. G. Holz, and J. Kopitz, “Proteome analysis of lipofuscin in human retinal pigment epithelial cells,” FEBS Lett., 528 217 –221 (2002). https://doi.org/10.1016/S0014-5793(02)03312-4 0014-5793 Google Scholar

13. 

D. K. Veirs, J. W. Ager III, E. T. Loucks, and G. M. Rosenblatt, “Mapping materials properties with Raman spectroscopy utilizing a 2D detector,” Appl. Opt., 29 4969 –4980 (1990). 0003-6935 Google Scholar

14. 

K. Hamada, K. Fujita, M. Kobayashi, and S. Kawata, “Observation of cell dynamics by laser scanning Raman microscope,” Proc. SPIE, 6443 64430Z (2007). 0277-786X Google Scholar

15. 

S. Kawata, R. Arimoto, and O. Nakamura, “Three-dimensional optical-transfer-function analysis for a laser-scan fluorescence microscope with an axtended detector,” J. Opt. Soc. Am. A, 8 171 –175 (1991). 0740-3232 Google Scholar

16. 

C. A. Lieber and A. Mahadevan-Jansen, “Automated method for subtraction of fluorescence from biological Raman spectra,” Appl. Spectrosc., 57 1363 –1367 (2003). https://doi.org/10.1366/000370203322554518 0003-7028 Google Scholar

17. 

J.-X. Cheng, A. Volkmer, L. D. Book, and X. S. Xie, “Multiplex coherent anti-Stokes Raman scattering microspectroscopy and study of lipid vesicles,” J. Phys. Chem. B, 106 8493 –8498 (2002). https://doi.org/10.1021/jp025771z 1089-5647 Google Scholar

18. 

J. Yang, X. Liu, K. Bhalla, C. N. Kim, A. M. Ibrado, J. Cai, T. Peng, D. P. Jones, and X. Wang, “Prevention of apoptosis by Bcl-2: release of cytochrome c form mitochondria blocked,” Science, 275 1129 –1132 (1997). 0036-8075 Google Scholar

19. 

R. M. Kluck, E. Bossy-Wetzel, D. R. Green, and D. D. Newmeyer, “The release of cytochrome c from mitochondria: a primary site for Bcl-2 regulation of apoptosis,” Science, 275 1132 –1136 (1997). 0036-8075 Google Scholar
©(2008) Society of Photo-Optical Instrumentation Engineers (SPIE)
Keisaku Hamada, Katsumasa Fujita, Nicholas Isaac Smith, Minoru Kobayashi, Yasushi Inouye, and Satoshi Kawata "Raman microscopy for dynamic molecular imaging of living cells," Journal of Biomedical Optics 13(4), 044027 (1 July 2008). https://doi.org/10.1117/1.2952192
Published: 1 July 2008
JOURNAL ARTICLE
4 PAGES


SHARE
Advertisement
Advertisement
Back to Top