Abstract
Purpose :
Correlated firing among neurons has been proposed as a key aspect of the neural code in sensory systems. In visual system, correlated firing has been reported in retinal ganglion cells (RGCs), the output neurons of the retina. Although correlated firing among RGCs is observed in most mammalian species, understanding of correlated firing differences across species has been determined yet. Therefore, here, we compared the correlated firing in mouse and monkey retina with normal vision and tried to understand its functional role on visual processing.
Methods :
C57BL/6J mice (6 retinas) at postnatal week 8 and higher and marmoset (callithrix jacchus) monkeys (4 retinas) at postnatal year 2 and higher were used. RGC spikes were recorded in isolated whole-mount retina using 8×8 multi-electrode array without and under white noise visual stimulus. Random binary checkerboard image of 200 μm square size was presented to retina at 10 Hz for 15 min. For the analysis of correlated firing between RGCs, cross-correlation analysis and burst analysis were performed.
Results :
In monkey retina, the spontaneous firing of RGCs without any stimulus showed 1) correlated firing within 2 ms and 2) the correlation index of RGC pair decreases with the increment of distance between RGCs. 3) With burst analysis, the spike count within a burst is larger (4.21±0.01), while the inter-spike interval and burst duration are smaller than those in mouse retina (35.56±0.29; 114.20±1.56) (p < 0.001). 4) The burst correlation index (Δt =100 ms) also decreases with the increment of distance between RGCs. In mouse retina, 1) the spontaneous firing of RGCs showed no correlated firing. 2) With the presence of visual stimulus, nearby ON–ON and OFF–OFF pairs of RGCs showed correlated firing within 100 ms, while nearby ON–OFF pairs of RGCs showed no correlated firing. 3) With mGluR6 agonist, DL-AP4, the correlated firing between ON-ON pairs of RGCs has disappeared.
Conclusions :
The correlated firing in spontaneous state clearly shows difference between monkey retina and mouse retina (in terms of correlation index, burst), which suggests that two species have different retinal encoding strategy. Since we have not performed RGC response recording with visual stimulus in monkey retina yet, in future study we would like to focus on visual information processing of monkey retina, which eventually leads us to better understand the visual encoding of human.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.