Abstract
Purpose :
Starburst amacrine cells are critical for the generation of retinal direction selectivity. The first characterization of direction selectivity in starburst cells in the primate retina (Detwiler et al., ARVO 2020) prompted us to initiate an analysis of their synaptic connectome. Electron microscopic (EM) reconstructions of mouse starburst cells showed a profound segregation of inhibitory synaptic input to proximal dendrites and convergent input from multiple bipolar types. Both of these properties may be critical for the generation of direction selectivity. Our purpose was to determine whether the pattern observed in the mouse is also present in the primate.
Methods :
Identified ON starbursts cells recorded in the vitro macaque monkey retina were filled with a fluorescent dye (Alexa 568). A confocal image stack of the dendritic tree was made and the retina fixed to preserve ultrastructure. The filled cell was then localized by epiillumination and Near Infrared Branding (NIRB) was used to etch fiducial marks around the starburst soma. The cell was embedded in plastic and the starburst cell was localized for serial block-face scanning EM. Imaged layers were aligned and the starburst dendritic tree was reconstructed using TraKEM software.
Results :
The reconstructed starburst dendrites generated synaptic outputs to both amacrine and ganglion cell processes from peripheral dendritic varicosities as found previously for both rabbit and mouse retina. We also observed sparse output to bipolar cell axon terminals confirming a previous observation (Yamada et al., JCN 2003). However, the proximal dendrites (~50 µm radius around soma) received only very sparse inhibitory synaptic input (15 synapses in total) similar to that described originally in the rabbit (Famiglietti, JCN 1992) but markedly distinct from the mouse (Ding et al., Nature 2016) where dense inhibitory inputs targeted the very proximal dendrites. Moreover, reconstructions of presynaptic bipolar cells suggest input from a single type, likely the diffuse bipolar type 5 (DB5), again different from the mouse where starburst cells receive input from multiple bipolar cell types (Greene et al., Cell Rep 2016).
Conclusions :
Synaptic input to starburst cells in macaque monkey is fundamentally distinct from that in the mouse. A space-time wiring model based on convergent input from multiple bipolar cell types appears unlikely in the primate retina.
This is a 2020 ARVO Annual Meeting abstract.