Biomedical Engineering Reference
In-Depth Information
14. CARS microscopy enables the detection of a resonant CARS
signal that is generated by a specifi cally excited chemical bond.
For the detection of neutral lipids, the symmetric CH
2
molecu-
lar vibration of fatty acids is excited at ~2,845 cm
−1
. However,
the signal from the molecules of interest may be impaired by
the nonresonant signal of a solvent or—in the case the live
yeast cell imaging—by the aqueous growth media [
25
,
41
].
Fortunately, the nonresonant background produced by yeast
media (e.g., rich medium or synthetic medium) does not sig-
nifi cantly interfere with the detection of LD in F-CARS mode.
Thus, yeast cells can be imaged directly in the particular growth
media. For critical experiments and for imaging of very small
LD, cells are typically washed 3× with distilled water, which
further reduces the nonresonant background. For detection of
CARS signals both in E-CARS or F-CARS modes, the cells are
mounted on standard microscope slides or on agar slides (
see
Subheading
3.4
). No background signal of the agar that is
used for cell immobilization was observed in E-CARS or in
F-CARS mode detection of CH
2
molecular vibrations excited
at 2,845 cm
−1
.
15. For CARS imaging of yeast LD, we apply high-numerical-aperture
objectives such as a 63× NA1.4 oil-immersion objective or a
63× NA1.2 water-immersion objective. For 2D imaging of
yeast cells with optimized sampling rate, a line frequency of
400 Hz (512×512 pixels) and bidirectional scanning mode
are applied. Optionally, 8× or 16× line averaging is performed.
Slower scan speed may result in light-driven artifi cial move-
ment of LD or may lead to cell destruction. For 3D imaging of
yeast cells, an increased line frequency of 700 Hz (512 × 512
pixels) and bidirectional scanning mode are applied. With this
setup no visible cell damage was observed. Using this setup
also small LD (<200 nm) can be imaged multidimensionally
[
36
]. However, in 4D experiments with CARS over extended
periods of time, cell growth may become signifi cantly impaired,
most likely caused by cell stress. Yeast LD can also be imaged
at video-rate using a resonant scanner [
23
]; however, imaging
of smaller LD (<200 nm) is limited.
16. GFP fl uorescence is rapidly bleached upon illumination with a
laser intensity that is used for CARS microscopy. Thus, CARS
signals and GFP fl uorescence are acquired sequentially.
17. Fluorescent dyes for labeling LD also stain phospholipids and
membrane proliferations, which may occur in specifi c mutants.
In such cases, the differentiation between the “true” neutral
lipid signal and the “artifi cial” membrane signal is diffi cult. In
such cases the resonant CARS signal at 2,845 cm
−1
can be
used as a control for neutral lipids in fl uorescently labeled
specimens. It should be noted, however, that the fl uorescent
