NATIONAL SCIENCE FOUNDATION
TOKYO REGIONAL OFFICE

The National Science Foundation's (NSF) Tokyo Regional Office periodically receives and disseminates reports on research developments in Japan that are related to the Foundation's mission. It also provide occasional reports on developments in other East Asian Countries.

These reports present information for the use of NSF program officers and policy makers; they are not statements of NSF policy.

Special Scientific Report #02-10 (September 30, 2002)

Our ability to understand how the brain works is limited by the spatial and temporal resolution of the techniques we currently have to record neural activity in vivo. Much progress has been made in the past 50 years utilizing electrophysiological methods (single unit recordings, etc.) that have ideal temporal resolution, but no spatial resolution. By recording from 100+ units in a region of the brain and looking for a statistical significance of the neuron’s response to a specific stimulus, we can generalize the region’s functional property. However this method lacks the efficiency required to survey a population of neurons effectively. To determine a neuron’s ideal stimulus (receptive field) and how this changes among neighboring columns of neurons we prefer a method that can record many neurons simultaneously with good spatial and temporal resolution. The next generation of optical imaging is the most promising step in this direction.

Optical imaging allows for simultaneous recordings of a population of neurons. Two approaches are currently being used- intrinsic imaging and voltage sensitive dyes (Grinvald, et. al.). A new technique of using two-photon microscopy utilizing voltage sensitive dyes has achieved the best spatial and temporal resolution currently available for recording neural activity (Denk, et. al.).

Intrinsic imaging is the most basic type of optical imaging, using only changes in the natural absorption of light energy by neural tissue without the use of external probes. Deoxyhemoglobin has a greater absorption of light energy when compared to Oxyhemoglobin, thus an increase in Deoxyhemoglobin will cause a darkening of the tiny capillaries in the neural tissue. Increased neural activity causes increased oxygen consumption, and thus an increase in deoxyhemoglobin. This increase has a short delay (1-2 sec), but is correlated with neural activity (much like fMRI will detect neural activity through measuring oxyhemoglobin/deoxyhemoglobin ratios). Thus intrinsic imaging allows for good spatial resolution, but unfortunately it also has poor temporal resolution. However, by a combination of intrinsic imaging and electrophysiology (Tsunoda K, et. al.) neural activity can be recorded with an improved spatial and temporal resolution.

Another approach is to use a voltage sensitive dye. The dye molecule attaches its hydrophobic end to the cell membrane of the neuron, with its hydrophilic end resting inside the cell. The large dipole of the dye molecule creates a high sensitivity to a change in membrane potential, with a depolarization causing the dye to dislodge itself from the membrane. Displacement from the membrane causes a change in either the fluorescence intensity or color. Thus by exciting the dye molecule with light energy of a particular wavelength, the voltage of the cell can be determined by monitoring the emitted fluorescence (with has a response time of a few microseconds). Although this temporal resolution is a significant improvement over intrinsic imaging, several problems exist. The signal to noise (S/N) ratio is extremely low (lower than intrinsic imaging), with the best dyes only increasing their fluorescence by 25% per 100 mV change in membrane potential. Secondly, phototoxicity and photobleaching are additional issues that minimize the recording time allowable. The most crucial issue, however, is that the majority of the signal from the dye is from the dendrites (because dendrites contribute the majority of the neuron’s surface area). As system neurophysiologists, we are interested in the output (action potentials) rather than the input information (depolarizations of dendrites) of the neuron primarily thus the information that the dye provides is not ideal. However, future dyes may solve this issue by being designed (i.e. genetically encoded) to selectively attach to specific regions of the neuron (cell body, axon, and dendrite). In addition, more sophisticated CCD cameras will improve the S/N ratio and temporal resolution (0.1 ms with current prototypes).

Two-photon microscopy is a relatively new development in the optical imaging of neurons. A dye molecule will fluoresce when it reaches a crucial energy level from the temporary absorption of a single photon (of a specific wavelength). However, if the dye molecule simultaneously absorbs two photons of slightly less than twice the wavelength of the single photon, this will also cause it to fluoresce. The ability for two-photon microscopy to only stimulate fluorescence from neurons receiving a high excitatory photon density provides an excellent spatial resolution of the received signal with minimal bleaching and photo toxicity. The fluorescence is confined to a three dimensional area (focus point of laser beam) and out of focus photons (from scattering) will not contribute to fluorescence. The larger wavelength of light used provides a deeper penetration into the neural tissue (up to 0.5 mm) and a reduction in scattering (eliminating the need of a confocal microscope). Impressive spatial resolution has been demonstrated with this new technique, allowing the imaging of single dendritic spines (Oertner, et. al.). This new technique has also been applied recently to image neural tissue in awake behaving animals (Helmchen, et. al.). However the cost of a 2-photon microscope has been one of the limiting factors to the widespread use of this impressive technique. The majority of the cost is attributed to the Ti-Sapphire pulsed laser required in the setup.

For an in vitro preparation (Tominaga, et. al.), the rat is anesthetized using ether and sacrificed. The brain is removed and put in a cooled ACSF (artificial cerebral spinal fluid) solution for more than five minutes. The hippocampus is then removed from the remaining brain, and is sliced to 400 um thick slices in the vibrotome. Next the slices are stained with a voltage sensitive dye (Di-4-ANEPPS), and placed in a chamber with 95% O₂ and 5% CO₂. Fetal Bovine Serum and ACSF are added to the medium to facilitate staining. After two hours, the slice is removed from the chamber and placed under the microscope. A stimulating electrode is placed in the striatum radiatum. A recording electrode is placed nearby in the stratum radiatum to measure EPSPs. A difference electrode is placed in the ACSF (outside the slice), and is used to reduce the noise picked up by the recording electrode. A grounded electrode is placed in the surrounding 3M KCl. At the start of a trial, the dye is excited for two seconds with 530 nm light, during which the slice is electrical stimulated. The dye fluoresces at 590 nm, which is selected by the dichroic mirror and imaged by the fast CCD camera (0.7 ms frame). This data is then processed, averaged (16 trials), and filtered by the imaging software.

In the first experiment, the effect of inactivating the GABA channels in a hippocampal slice using Bicuculline (20 uM) was examined. In the second experiment, the effect of Bicuculline on the cortex was investigated. In the last experiment, the effect of CHQX (20 uM), an AMPA blocker, was examined. For each experiment, the slice was perfused with the solution for ten minutes, before the slice was electrically stimulated and the recording began.

Bicuculline inhibits the ability of GABA channels to effectively hyperpolarize nearby neurons. Thus, we expect an increased spread of activity in both the cortical (figure 1) and hippocampal slice (figure 2). Cortical activity did not spread in our control (no Bicuculline) while hippocampal activity spread, but to a lesser extent. In our third experiment, we examined the influence of AMPA receptors on the spread of activity in the hippocampus. We saw a decreased spread, as this was due to NMDA activity only (figure 2).

Hippocampus control

4. Discussion

The S/N ratio in these experiments was fairly poor, and the data was averaged over 16 trials to achieve a reasonable level of signal to noise.  For in vivo experiments, additional noise sources (from blood vessels) are present.  Thus a higher S/N ratio (through the use of better dyes) must be achieved to make in vivo preparations more feasible.

In addition, most optical imaging setups are rather bulky, making in vivo optical imaging on freely moving, awake behaving animals extremely difficult.  Several labs have designed compact systems (Helmchen, et. al.) for this purpose.  With these developments, over the next several years as two-photon microscopy and voltage sensitive dye imaging become more practical and ubiquitous, we can hope to have a powerful new tool to probe electrical activity from a neuronal population.

References

Grinvald A., Shoham D., Shmuel A., Glaser D., Vanzetta I., Shtoyermann E., Slovin H., Sterkin A., Wijnbergen C., Hildesheim R. and Arieli A. In-Vivo Optical Imaging of Cortical Architecture and Dynamics Modern Techniques in Neuroscience research. U Windhorst and H. Johansson (eds), Springer Verlag, 2001.

Helmchen F, Fee MS, Tank DW, Denk W. A miniature head-mounted two-photon microscope. high-resolution brain imaging in freely moving animals. Neuron, 2001 Sep 27;31(6):903-12

Oertner TG, Sabatini BL, Nimchinsky EA, Svoboda K. Facilitation at single synapses probed with optical quantal analysis.
Nat Neurosci. 2002 Jul;5(7):657-64.

Tominaga T, Tominaga Y, Yamada H, Matsumoto G, Ichikawa M.Quantification of optical signals with electrophysiological signals in neural activities of Di-4-ANEPPS stained rat hippocampal slices.
J Neurosci Methods. 2000 Oct 15;102(1):11-23.

Tsunoda K, Yamane Y, Nishizaki M, Tanifuji M. Complex objects are represented in macaque inferotemporal cortex by the combination of feature columns. Nat Neurosci. 2001 Aug;4(8):832-8.

Acknowledgements

I would like to thank the Ichikawa lab at RIKEN BSI for providing the laboratory, materials, and assistance to conduct my experiment.  Without their help, my project would have not been possible.

In addition I would like to thank the Tanifuji lab at RIKEN BSI for allowing me to observe several experiments including intrinsic imaging, electrophysiology recording from IT, and voltage sensitive dye in vivo on an anesthetized macaque.  This has provided tremendous insight into the difficulties and advantages of using optical recording methodologies in vivo.

Lastly, I would like to thank my host scientist Thomas Knopfel for providing me with a computer to write this report.

Back to top

Return to Tokyo homepage