Page 113 - Vibrational Spectroscopic Imaging for Biomedical Applications

Sample Pr eparation of Cells and T issue 89

normal amide A envelope. Although subjective, Miljkovic et al. 59
demonstrated that it was possible to obtain a protein intensity image
in which the HeLa cell displayed expected high-protein intensity,
centered at its nucleus.
In contrast to using cells in suspension as in Miljkovic’s et al.
57
59
Study, cells used by Moss et al. were cultured directly onto
CaF plates. This plate was placed into a liquid cell consisting of a
2
15-μm Teflon spacer, providing a pathlength of 11 to 12 μm and
maintained at 35°C. A constant flow of cell culture medium was
passed through the cell at a rate of 230 μL/h. As in Miljkovic’s et al.
59
Study, a background spectrum of growth medium was collected
in a cell-free region of the sample and ratioed to the cell spectrum.
There was high reproducibility between SR-FTIR spectra obtained
from 10 individual fibroblast cells when a spectrum of each cell was
acquired every 24 minutes for 2 hours. Although intrasampling differ-
ences were observed between cells, these were very much smaller than
the standard deviation of repeated measurements for each cell. Moss et
57
al. provides further support for the low-spectral variance observed
for live cell FTIR spectra, when collected within the first few hours of
transfer to the sample analysis chamber.
59
In agreement with the study by Miljkovic et al. a distorted amide
57
I to amide II intensity ratio was observed by Moss et al. However,
since the spectrum of background water is different to that of water
bound to macromolecules, it was suggested that it is not possible to
accurately eliminate this background absorbance. The authors also
suggest that if the goal of the experiment is to obtain a spectrum from
the same position of the exact same cell, before and after administra-
tion of a stimulus, then the difference between the spectra can be
resolved even in the presence of background water. Additionally, it
was found in this study that nonconfluent cells could migrate out of
57
the measuring SR beam. Moss et al. suggest that this could be mini-
mized by placing the cell into a well.
Raman Studies
The spatial resolution of Raman spectroscopy is inherently higher
than that of FTIR due to the shorter wavelength of the excitation
radiation (the diffraction limit is generally given as ~λ/2). An image
obtained by Raman microspectroscopy requires raster scanning a
focused laser beam across the cell. Using this mode of data collec-
tion, an increase in spatial resolution, which is a function of step
size and beam diameter, also increases the time for chemical map-
ping. Although in FTIR studies it has been shown that for up to 3 hours,
spectral changes are not observed at the whole-cell level (see section
“FTIR Studies”), previously reported Raman maps of living cells
60
have required ≥3 hours collection times. At the subcellular level one
might expect that biochemical changes could occur within this period
for a given sampling point. However, it has been shown that localization