Abstract
Purpose: :
Many amacrine and ganglion cells (GCs) communicate via gap junctions. One function of electrical coupling is to generate correlated activity between neighboring cells. Here, we infected neurons with Channelrhodopsin 2 (ChR2) to generate light-activated currents in the inner retina and to differentiate their effect on GC correlated activity from that derived from photoreceptors.
Methods: :
ChR2 was expressed in C57BL/6J mouse retinas by infection with AAV2. Spontaneous and light-evoked GC spiking was recorded with a multi-electrode array. A cocktail of neurotransmitter antagonists was applied to block chemical synapses and 18β-glycyrrhetinic acid (18-GA) was used to block electrical synapses. Cross-correlation functions (CCFs) were computed to analyze correlated activity of GC pairs.
Results: :
GC pairs were divided into three groups: (1) both GCs expressed ChR2; (2) neither GC expressed ChR2; and (3) only one GC expressed ChR2. For GC pairs that both expressed ChR2, blue light produced correlated activity regardless of the inter-somatic distance. As expected, these correlations were due to independent expression of ChR2 and survived blockade of synaptic transmission. In contrast, light-evoked correlated activity derived from photoreceptors was seen for GC pairs separated up to 360 µm and was abolished by the cocktail of antagonists. Interestingly, the correlations derived from photoreceptors were reflected by relatively broad CCFs >100 ms, whereas correlations derived in the inner retina, isolated by blockade of synaptic transmission, were more transient. Many light-evoked CCFs showed oscillatory correlations (~150 Hz). These oscillations were reduced or abolished when blue light was paired with synaptic blockade. Surprisingly, we found that the correlated spontaneous spikes of coupled GC pairs were often increased by 30-50% when chemical synaptic transmission was blocked. Finally, under synaptic blockade, we found that ChR2-induced activity could generate spikes in coupled non-ChR2-expressing GC neighbors with an efficiency rate of nearly 60%. As expected, these correlations were abolished by addition of 18-GA.
Conclusions: :
Our results indicate a greater temporal precision for correlated spiking derived from inner retinal circuits than those from outer retina. Surprisingly, we found that spike correlations driven by direct gap junctional coupling between neighboring GCs were reduced by chemical synaptic transmission, suggesting that inhibitory pathways reduce spike coherence between neighboring GCs. By activating ChR2, we found that direct coupling between GCs can propagate spikes with very high efficiency.
Keywords: gap junctions/coupling • ganglion cells • retinal connections, networks, circuitry