We have shown that optical coherence imaging can be used to detect changes in the intrinsic optical properties of single neurons that accompany action potentials. OCT was used to observe scattering changes from neurons in the abdominal ganglion of Aplysia californica during stimulation of a connective nerve. OCM, which offers higher transverse resolution, was used to show a correlation between membrane voltage and scattering intensity from single cultured Aplysia bag cell neurons during action potentials. The changes in scattering follow a similar time scale as the membrane voltage. Past research, as discussed in the introduction, has demonstrated the feasibility of optical coherence imaging techniques for detecting functional activity in various neural tissue types including crustacean nerve fibers, Aplysia ganglion, and mammalian cortex and retina. In this study, we show that these techniques have potential use for optical detection of functional activity in single neurons. In addition, this ability could serve future studies designed to better understand the origins of intrinsic scattering signals from large populations of neurons.
We demonstrated the capability of OCT to image a population of neurons and then selected a particular neuron for investigation of functional activity with M-mode imaging. M-mode imaging of a single neuron in the population showed that there are scattering changes present without electrical stimulation. We speculate that these fluctuations are caused in part by intrinsic biological processes which may include localized spontaneous electrical activity or trafficking of vesicles in the cytoplasm. Upon electrical stimulation of the ganglion, compound action potentials in the ganglion were detected and an increase in scattering from the neuron was observed. These changes were slow with respect to the measured compound action potential, which was less than 15 ms in duration (). The timing of the observed scattering changes indicates that swelling of the neuron is a likely mechanism. Slow optical responses in squid giant axons have been found to be caused by an influx or efflux of ions resulting in swelling or shrinking of the periaxonal space [4
]. On the contrary, scattering changes which follow the time course of the electrical activity and are linearly proportional to the membrane voltage are thought to be caused by the realignment of charged molecules in the membrane of neurons [6
]. Measurement of functional activity based on swelling due to flow of ions [17
] and hemodynamic effects [11
] with OCT has been previously demonstrated. In this study we have shown that OCT can detect such scattering changes from individual neurons.
While OCT allowed scattering changes to be localized to the neurons used in this study, the typical transverse resolutions in OCT (10–20 µm) are likely insufficient to identify smaller mammalian neurons that are often studied in neuroscience. OCM is a variation of OCT that uses higher numerical aperture optics to focus light to a smaller spot size. OCM can have a transverse resolution less than 1 micron at the expense of a short depth-of-field. We demonstrated OCM on single cultured Aplysia
bag cell neurons to show high resolution imaging capabilities as well as to show a direct correlation between membrane voltage and optical scattering intensity during action potentials. Increases in scattering intensity from neurons were observed to followed the time course of the induced action potentials and membrane depolarization. Intrinsic optical signals that are linearly dependent on membrane voltage have been observed in previous studies [5
]. One mechanism of these optical signals is believed to be a realignment of charged membrane proteins in response to a voltage change [6
]. With the correlation between the membrane voltage and the scattering changes observed in this study, it is likely that the realignment of membrane proteins is the dominant mechanism.
A delay of roughly 70 ms was observed between the change in membrane voltage and the change in scattering intensity (). Ongoing research is investigating possible sources of this delay. Possible contributions include the separation between the site of the electrode tip placement and where optical recordings were made, slow conduction velocity of the action potential due to poor condition of the neuron, or the time required for the neuron to dynamically change the optical scattering properties of the membrane. Conduction velocities of roughly 5 mm/s have been reported in Aplysia
bag cell neurons [22
]. Assuming a propagation distance of roughly 50 microns this would result in a delay of 10 ms. A slower conduction velocity could be explained by poor health of the neuron or iatrogenic injury by the stimulating/recording electrode. Conduction velocities of Aplysia
neurons can also vary depending on the state of the animal from which the neurons are harvested [23
]. Despite the delay, the changes in scattering that were observed have the same duration and are proportional to the induced electrical activity.
A major challenge for using OCT and OCM for detecting neural activity is that the intrinsic optical signals are small and are accompanied by signals from many sources of noise. Fluctuations in scattering intensity were observed from cultured bag cell neurons in their resting state. This effect was less evident in the OCT imaging of the intact ganglion, possibly due to spatially averaging of the fluctuations as a result of the lower transverse resolution. The fluctuations in scattering in the cultured bag cells decreased dramatically when they were treated with isotonic potassium chloride (KCl), leading us to believe they are associated with the functional activity of the neurons. Adding KCl to the culture media eliminates the electrochemical gradient that exists across the membranes, effectively killing the neurons. Possible physiological mechanisms of fluctuations in scattering include trafficking of vesicles in the cytoplasm or cytoplasmic streaming. The mechanisms from non-excited neurons are being investigated, and may result in a means for non-invasively determining other forms of functional neuron activity or processes. Since these fluctuations are not correlated to the induced electrical activity, a low amount of signal averaging was needed to identify the changes associated with action potentials. This averaging, however, is consistent with other electrophysiological studies.
Future improvements in detection sensitivity could be achieved using variations of OCT and OCM. Polarization-sensitive OCT (PS-OCT) is an imaging technique that enables depth-resolved mapping of the birefringence of biological tissue [24
]. Changes in the birefringence of nerves due to electrical activity have been shown to be an order of magnitude larger than scattering intensity changes [5
]. PS-OCT requires relatively minor changes to a typical OCT system, thus it could potentially provide significant improvements in sensitivity with little additional instrumentation. Phase-sensitive interferometric techniques have been used to measure nerve displacements less than 1 nm [12
]. OCT is capable of phase-resolved imaging [25
] which could also potentially be used to improve detection sensitivity of scattering changes from individual neurons or sparse populations of neurons. Combinations of phase-resolved and polarization-sensitive OCT are possible [26
] and may ultimately result in the highest detection sensitivity. Other possible improvements include faster imaging speed. In this study, optical scattering changes from a single position of the beam were recorded. However, line rates of several hundreds of kilohertz have been reported for OCT [27
], meaning that 2D cross-sectional images could potentially be acquired at frame rates fast enough to detect most electrical activity in neural tissue. Additional investigations are needed to determine how effectively these intrinsic optical signals can be recorded from individual neurons from other species. For example, single mammalian hippocampal neurons are smaller (< 10 microns in diameter), have faster action potentials (~1 ms) with faster propagation velocities (> 10 m/s), and will likely generate intrinsic optical scattering signals that are lower in magnitude.