Purchase this article with an account.
Katja Reinhard, Thomas A Muench, ; Insights into visual processing of human retina in-vitro. Invest. Ophthalmol. Vis. Sci. 2014;55(13):2380.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Retinal information processing has been characterized in many animal models. However, except for a short communication by Weinstein et al. in 1971, no physiological in-vitro data from human retina has been published. It is often alleged that sophisticated image processing might be absent in human retina. Further, with respect to novel treatment options against visual impairment, we believe that it is important to gain more knowledge about the targeted tissue - the human retina. We thus aimed at studying human retina function on cell and system levels in-vitro.
Pieces of human retina - donated by patients who had to undergo a medically indicated enucleation - were placed ganglion cell side-down on a 60-electrode multi-electrode array (MEA), and stimulated with various light stimuli. In addition, we recently implemented a high-density MEA with 11011 electrodes (Frey et al. 2009), and performed a first successful measurement with human retina.
During stimulation, we recorded spiking activity from retinal ganglion cells (RGCs), and in 8 out of 15 donated retinas we could measure abundant light responses. The recorded cells showed diverse properties: ON-, OFF- and ON-OFF-responses, different response latencies and transiencies, and different speed tuning. Interestingly, human RGCs are tuned to higher speeds and higher temporal frequencies compared to mouse RGCs.
The observed differences in speed tuning might be explained by the higher angular velocities to which human RGCs are exposed in the bigger human eye. Further, differences in temporal frequency tuning suggest species specific kinetic differences of synaptic processing within retinal circuitry. Taken together, we show that it is possible to record light responses from human retina in-vitro, and we provide a glimpse into the diversity of its information processing. In the future, the use of high-density MEAs will allow further comprehensive characterization of human RGC types.
This PDF is available to Subscribers Only