Optical recording of fast neuron activities

Summary form only given: Spatiotemporal imaging of neuronal activity in the brain plays a critical role in studies of brain function and disorders. However, thorough monitoring of neuronal activity in the brain has been far from practical due to the limitations of conventional technologies. Since brain activity is fundamentally the ensemble of excitation of neuronal cells, it exhibits extremely high spatiotemporal heterogeneity on the micrometer and millisecond scale. Further, a number of functional studies are accompanied by behavioral control and/or observation. An ideal functional imaging technology, therefore, would involve noninvasive, label-free and three-dimensional dynamic imaging of cortical neuronal activity with μm and ms resolution. Optical methods look like one of the most feasible approaches because fast optical signals of neuronal activity have several advantages over conventional methods, including the advantages that they require no probe and thus can be recorded noninvasively in 3D. It is logical to start from validating the measurement of fast optical signals in the fundamental unit of the neural system (neurons) and then pursue toward the measurement in the most complex neural system (human brain). As several scientists had reported the measurement of fast optical signals in large neurons isolated from low-level animals, we performed the next step in this study optical measurement of neuronal activity in ex vivo brain tissue.We built a high-speed confocal NIR spectrometer to monitor the transmittance of live brain slices during functional activation. A relative change in the optical transmittance was recorded from a measurement area with a diameter of ~100 μm, at 1 ms time resolution, and at 256 wavelengths ranging from 780 to 1315 nm. Voltage stimulations were applied for functional activation, and local field potentials (LFPs) were simultaneously measured during optical recording. As results, we observed transient optical responses that are associated with neuronal activity. The optical signals were concurrent and correlated with LFPs, and they disappeared following tetrodotoxin (TTX) application. In addition, the amplitude of fast optical signals was statistically significantly larger in longer wavelengths (~1200 nm). The measurement of ex vivo fast optical signals was statistically confirmed across five brain slices. Interestingly, the time course of optical signals was quite different from that of LFPs. As this difference implies that optical signals may originate not directly from the membrane potential but from other neurophysiological events accompanying excitation, we tested the hypothesis that optical signals are attributed to transient cellular volume changes. With no previous dynamic equation for neuronal volume dynamics during excitation on the millisecond time scale, we developed a novel mathematical neuron model. Use of this model elucidated the difference between the temporal dynamics of optical and electrical signals. Findings of this study will be used in pursuing the next step of our research 3D in vivo imaging of neuronal activity with μm and ms resolution.