Figure 1.
Layout of the 60-electrode pMEA.
The electrodes are arranged in an 8×8 array with 200 µm electrode distance. Perforations of various size are visible in-between electrodes (source: 60pMEA200/30iR data sheet by Multi Channel Systems).
Figure 2.
Our MEA setup consists of two perfusion loops. Solution is supplied to the MEA chamber from the top through the upper perfusion (A) and excessive solution is removed by the suction (B). The necessary negative pressure is supplied by the additional perfusion, consisting of the lower perfusion (C) and a vacuum (D). Details are given in the following text and figures.
Figure 3.
Experimental procedure Step 1: Filling of MEA chamber.
Step 1a) Placing the MEA chamber on the baseplate. Step 1b) Preparation of perfusion and vacuum. Step 1c) Filling the MEA. Detailed description is given in the text.
Figure 4.
Experimental procedure Step 3: Fixation on filter paper.
Step 3a) Preparation of filter paper. Step 3b) Fixation of retina on filter paper. Details are given in the text.
Figure 5.
Experimental procedure Step 4: Transfer of retina to MEA chamber and setup.
Step 4a) Placing the retina on the electrodes. Step 4a) iii: Top: Good MEA preparation. All electrodes are clearly visible; the retina looks homogeneous, flat, and without tears or holes. The retina and filter paper are nicely centered over the middle of the electrode array. Bottom: Bad MEA preparation with air bubble (blue arrow) and holes due to excessive negative pressure (gray arrow). Further, the filter paper is shifted towards the upper left corner. Orange arrow: optic nerve head. Step 4b) Transfer of MEA amplifier to setup. Step 4c) Installation of upper perfusion loop. Details are given in the text.
Figure 6.
Experimental procedure Steps 5 and 6: Recording data (Spike recordings).
A) Snapshot of a 500 Hz high-pass filtered MC_Rack display. Spiking activity with good signal-to-noise is visible on many electrodes. B) Snapshot of MC_Rack display after overflow. Noise with amplitudes of 200 to over 1000 µV due to wet electronics is visible on most electrodes. C) Snapshot of MC_Rack display several hours after strong overflow. Slow noise on many electrodes is visible either if the electronics is not fully dry yet or when the electronics has been irreversibly harmed. D) Snapshot with slow fluctuations and spike-like noise peaks (red asterisks). See text (Step 5 and 6, troubleshooting) for details.
Figure 7.
Additional steps for in vitro ERG recordings.
A) Additions to Step 1: The AgCl reference is positioned over the MEA by a reference electrode holder and is attached to pin 15 (REF) by a wire ferrule insulated by shrink-on tubing (asterisk). B) Additions to Step 5: Schematic of the reference electrode and its holder as shown in A. Note the optical shield needed to avoid photoelectric artifacts resulting from light hitting the reference electrode.
Figure 8.
Experimental procedure Steps 5 and 6: Recording data (in vitro ERG recordings).
A1) Snapshot of the Longterm Data Display (raw data) from MC_Rack. Note that on most electrodes the ganglion cell spikes mask the in vitro ERG responses (e.g. the electrode marked in orange). Only the highest contrast flash elicits a response that is visible on most electrodes (red asterisks), while on some electrodes without ganglion cell spikes the in vitro ERG responses are clearly visible (electrode marked in blue). Reference electrode 15 (REF) is on the left. A2) Zoomed view of the electrode marked in blue from panel A1 showing the responses to flash stimuli of different contrast (highest two contrasts marked with red asterisks). The low-pass filtered data around the time highlighted by the box is shown in B1+B2. B1) Data Display with 200 Hz low-pass filter applied. There is a clear response on almost all electrodes. Not all spikes get filtered out by the low-pass filter. Note the different time scale than in A1. B2) Zoomed view of the electrode marked in blue from panel B1 that shows a very clear low frequency in vitro ERG response without contamination by ganglion cell spikes.
Figure 9.
A) Responses of one ganglion cell to a step in contrast over 6 hours. A two second light decrement step has been shown >120 times over a period of 6 hours. Each dot in the raster plot represents one spike produced by the ganglion cell. The ganglion cell stably responded to the stimulus during the whole recording time. Changes in latency and number of spikes are due to different mean brightness levels used during the experiment. B) Receptive field of one ganglion cell calculated from checkerboard stimuli. 15×15 checkers out of 40×40 shown here. The stimulus has been repeated approximately every 90 minutes. Time above each receptive field map: presentation time of the checkerboard stimulus (0 min = beginning of experiment). The receptive field location and shape was stable during the whole 8 hours, indicating that the retina did not move significantly.