Comparing patch-clamp data with the extracellular activity of neurons under electro-stimulation
The aim of this research collaboration was to assess intracellular versus extracellular response properties of retinal ganglion cells under electrical stimulation. Since patch-clamp data are not contaminated by the activity of neighboring cells and because generally, they have a higher signal-to-noise ratio, we can use them as a ground-truth reference to evaluate the validity of results achieved by extracellular recordings. Therefore, for this study similar to extracellular experiments we recorded either visual and electrical induced activity of retinal ganglion cells.
Following the full-field flash light stimulation, we applied both subthreshold electrical noise and single pulses with different levels of amplitude. From noise stimulation data, we estimated the electrical receptive of each cell and with pulse stimulation data we were able to estimate the electrical tuning curve of each cell to find the stimulation threshold
Figure 1.
Electrical tuning curve of a sample OFF cell. The activation threshold is about -1000 mv.
Click here to see the high quality image
Figure 1 shows the spike raster of an OFF cell in response to pulse stimuli with different amplitudes. For this cell, the threshold for activation is about 1000 mv. For this experiment, we also applied noise stimuli with different levels of strength. We started with noise with an average of -400 mv and then increased it to -800 and -1000. The figure below shows sorted spike raster and linear filter estimated using direct (data confined in 10ms after stimulation onset) and indirect (data after excluding first 10 ms) response of the cell while noise was stimulated by noise with the average of -800 mv. Direct response, reflect single pulse activation, whereas indirect plot shows the network mediated activity of the cell. From this plot, we can infer that -800 mv is not high enough to generate an integrative response (see figure 2).
Figure 2.
Electrical stimulation of the retina with Gaussian noise (avg -800 mv). Due to low amplitude the indirect response does not generate electrical input filter.
Click here to see the high quality image
By increasing the average of amplitudes to -1000 mv (the threshold of activation with single pulses) we observed an integrative electrical input filter with a downward deflection in the indirect plot (figure 3), which is similar to the input filters of OFF cells estimated from extracellular recordings (see here).
Figure 3. Electrical stimulation of the retina with Gaussian noise (avg -1000 mv) can generate an integrative response in OFF cell.
Click here to see the high quality image
We observed similar patterns in the electrical input filters of OFF RGC types for intracellular and extracellular recordings.
A similar analysis was performed for other cells. Likewise, we observed an electrical input filter with downward deflection for another OFF cell stimulated by electrical Gaussian noise with -800 and -1000 mv. This neuron seemed to have a lower activation threshold and had integrative filters in response to both stimuli. Due to limited recordings, we did not find any ON cell having an electrical input filter with significant deflection. Further experiments can provide more details about the electrical characteristics of retinal pathways(figure 4).
Figure 4. Electrical stimulation of the retina with Gaussian noise (avg -800 and-1000 mv) can generate an integrative response in and OFF cell with lower
activation threshold.
Click here to see the high quality image
Supervisors: Dr. Daniel Rathbun, Prof. Eberhard Zrenner and Dr.Zohreh Hosseinzadeh.
Collaborators: Dr. Mohit shivdasani, Madhuvanthi Muralidharan and Dr.Tianruo Guo.
