Biomedical Engineering Reference
In-Depth Information
enhancement nature of SHG as described earlier, SHG is performed with the dye at wavelengths that
also generate two-photon fluorescence and, therefore, photodamage [8]. As such, one needs to limit the
laser illumination time at a single point to avoid photodamage while collecting enough SHG photons.
In practice, we found that neurons can sustain tens of milliseconds of laser illumination at a single
point in a given time, which is enough to observe whole action potential event. In addition, to improve
signal-to-noise ratio, we collect tens of events in single SHG recordings and average them. Furthermore,
we provide laser pulses at an interval of 500 ms (2 Hz recordings) so that neurons have enough time to
recover in between the laser pulses.
10.5 Applications
Here, we describe some examples of membrane potential measurements using SHG. As described in
the introduction, the unique strength of membrane potential imaging by SHG becomes evident in mea-
suring membrane potential dynamics at small targets in neurons, such as axons, distal dendrites, and
dendritic spines that are hardly accessible to conventional electrophysiological techniques.
10.5.1 SHG imaging of Somatic Voltages
The potential of SHG imaging for membrane potential measurement in cells was realized and investi-
gated from the early 1990s [20,21]. With the development of techniques, it has been applied to mam-
malian neurons to measure the membrane potentials in these cells [10-12], and particularly to perform
optical measurements of somatic voltage (Figure 10.2).
Shown in Figure 10.2 is the actual response of SHG to membrane potential changes taken with the
regular frame scan with alternating membrane potential controlled by the voltage-clamp technique. As
can be seen, mean SHG signals at the soma changes with membrane potential. Further analyses with
different voltage steps reveal that SHG signals collected at the soma respond to membrane potential
changes in a linear fashion within and beyond the physiological voltage fluctuation ranges. SHG chro-
mophores (in this case, FM4-64) can be applied both from inside (intracellular loading) and outside
(extracellular loading) to give rise to the same magnitude of membrane potential sensitivity (~10% per
100 mV), but with the opposite sign. This is because the chromophores fill the plasma membrane from
exactly an opposite direction in these loading schemes and therefore sense the membrane potential
changes in the opposite direction as well. Combined with the intrinsic high spatial resolution of SHG
and high temporal resolution provided by point-scan protocol or other methods, this high magnitude
of the linear response of SHG signals to membrane potential changes allows quantitative imaging of
membrane potential changes in the areas where previous techniques had limited access.
10.5.2 SHG imaging of Dendritic Spines
The imaging of voltage at dendritic spines with SHG constitutes one of the clearest cases where the
advantages of SHG are exploited by neuroscientists (Figure 10.3). The membrane potential dynamics in
dendritic spines had been a mystery as these structures are too small for conventional electrophysiologi-
cal recordings. However, from previous experiments, especially from those using two-photon calcium
imaging, it had been well recognized that these small structures play very important roles in neuronal
physiology [1,22]. As direct measurements are difficult, researchers have utilized numerical simulations
to explore the dynamics of electrical signaling in dendritic spines [23,24]. In addition, attempts have
been made to use fluorescence signals of voltage-sensitive chromophores to measure the membrane
potential changes in dendritic spines [5,25]. However, direct measurement of membrane potential
changes in spines continues to be a big challenge due to background signals originating from intracel-
lular dyes, as mentioned in the introduction. Taking advantage of SHG's ability to measure membrane
