2.2. Retinal Dissociation and Cell Identification
Salamander Ringer (85 mM NaCl, 1.5 mM KCl, 25 mM NaHCO
3, 0.5 mM CaCl
2, 0.5 mM NaH
2PO
4, 24 mM glucose, 0.03 mM phenol red, 1.0 mM sodium pyruvate) and salamander media (108 mM NaCl, 2.5 mM KCl, 2 mM HEPES, 1 mM NaHCO
3, 1.8 mM CaCl
2, 0.5 mM NaH
2PO
4, 1 mM NaHCO
3, 24 mM glucose, 0.5 mM MgCl
2, 1 mM Na pyruvate, 7% medium 199, 1× minimum essential (MEM) vitamin mix, 0.1× MEM essential amino acids, 0.1× MEM nonessential amino acids, 2 mM glutamine, 2 μg/mL bovine insulin, 1 μg/mL transferrin, 5 mM taurine, 0.8 μg/mL thyroxin, 10 μg/mL gentamicin, and 1 mg/mL bovine serum albumin (pH 7.7)) were prepared as previously described [
12]. Salamander retinas were isolated as previously described, with some modification [
12]. Briefly, salamanders were enucleated, and the anterior portion of the eyes was removed. Each eyecup was then bisected, using a razor blade. Each half retina was then removed from each half eyecup, using curved forceps, washed 2 times with salamander Ringer, and kept in salamander Ringer at 10 °C, until dissociation. Dissociation was performed immediately prior to optical tweezing, by placing each retinal piece into a 1.5 mL papain solution consisting of 14 U/mL papain (Worthington, Freehold, NJ, USA) in salamander Ringer’s solution, for 40 min, at room temperature. Each retinal piece was then washed twice with salamander Ringer and finally gently triturated in 1 mL of ice-cold salamander Ringer. Then, 50 µL of cell suspension was immediately used for optical tweezing.
Cells were identified by using their morphology, as previously described [
13,
14] (see Figure 4). For retinal neurons, many of the five cell types can be reliably identified by their distinguishing axonal and dendritic extensions. All salamander cells are large. Photoreceptor cell somas measure about 10–15 µm in diameter. In addition, photoreceptors have an extension of the soma called the inner segment, which contains a collection of mitochondria called an ellipsoid. Rod cells have a triangular ellipsoid, whereas cone photoreceptors contain a spherical ellipsoid. Rod cells may also have a long rod-like outer segment, the photosensitive part of the sensory cell, and an axon terminal extending from a short axon-like process; however, during dissociation, these structures can become separated from the cell body of the rod photoreceptor. In contrast, cone photoreceptors have an axon terminal directly attached to the cell body and a short cone outer segment, which is usually not retained during dissociation. Bipolar cells can be identified based on the presence of a Landolt club, a thick primary dendrite; moreover, their cell somas are slightly smaller, about 8–10 µm. Horizontal, amacrine, and ganglion cells have distinctive dendritic trees with multiple branches emerging from the cell soma. However, it was uncommon for the branching neuritic processes to be adequately retained after retinal dissociation, and therefore these cells were grouped into a “multipolar” category. On occasion, however, ganglion cells were recognizable because of their large cell bodies, ~20 µm, prominent primary dendrite, and thin single axon.
2.3. MEA Coating and Preparation
Multiple Electrode Arrays (MEAs) were purchased from Multi Channel Systems MCS GmbH (Reutlingen, Germany). Arrays chosen for these experiments had an electrode diameter of 10 µm and an interelectrode distance of 30 µm (Multi Channel Systems MCS GmbH, Cat #60ThinMEA30/10iR-ITO-(w/o)). The titanium nitride (TiN) electrode heads with conductive strips of transparent indium tinoxide (ITO), along with silicon nitride (SiN) insulation isolating each electrode, were fabricated on top of #1 coverslip glass (0.13–0.17 mm thick). The total thickness of the MEA arrays is 180 µm. It was necessary to use thin MEAs due to the optics of the objective used for optical tweezing (a 40x oil immersion plan neofluor objective with a numerical aperture of 1.3 and a working distance of 210 µm), as well as to focus the laser on a single cell. No physical alteration of these MEAs was required. These arrays were cleaned overnight in a 1% w/v Tergazyme (Sigma Aldrich, Cat# Z273287, St Louis, MO) solution, then washed three times with DI H2O, and finally sterilized under UV, in a biosafety hood, for 1 h.
Poly (2-hydroxyethyl methacrylate) (p-HEMA) (Sigma Cat#3932) solution was prepared as previously described, by dissolving 20 mg/mL of p-HEMA in 95% ethanol (EtOH), overnight, rocking at RT [
9,
10]. A p-HEMA coating was necessary to prevent adhesion of retinal cells to some areas of the MEA so that the optical tweezers could pick up and move the retinal cells. This solution was applied to specific areas of the surface of the MEA by placing the MEA in a 35 mm dish, such that it was inclined at a 60° angle. Then, 100 µL of p-HEMA solution was carefully dripped onto the surface of the MEA, being careful to not allow the p-HEMA solution to cover the central electrode region (
Figure 1). The MEAs were then laid flat into 94 mm dishes, covered, and allowed to dry for 1 h in a biosafety hood. The MEA was then rotated 90°, and the p-HEMA coating was repeated so that ¾ of the total surface area of the MEA was coated, but not the electrodes in the center (
Figure 1). If p-HEMA did accidentally drip onto the electrode area, the MEA was quickly sprayed with 70% EtOH, washed 3 times with sterile DI H2O, and allowed to dry. The MEA was then recoated with p-HEMA, using the steps previously described.
Polydimethyl siloxane (PDMS) rings were made to hold liquid on the MEA. PDMS (Dow Corning Corporation Cat#3097358-1004) was made as previously described [
15]. Briefly, elastomer base was vigorously mixed with the curing agent in a 10:1 ratio by weight. The solution was then placed in a desiccator, under vacuum, for 30 min, to remove air bubbles. The PDMS polymer was poured into a 94 mm culture dish, placed under vacuum for another 30 min, and finally cured in a 70 °C oven for at least 2 h. A ring with a 1” outer and ¾” inner diameter was punched from the PDMS slab, cleaned using Scotch Tape, and sterilized by submerging in 70% EtOH. Subsequently, the PDMS ring was washed with sterile DI H
2O twice and allowed to dry under UV light overnight. The PDMS ring was then placed on the MEA around the central electrodes. Vaseline was applied around the outside of the PDMS ring, in order to ensure there would be no leakage of media during culture.
The ¼ of the MEA not covered with p-HEMA was coated with a substrate that provides adhesion for salamander neurons, a monoclonal antibody called Sal-1 [
16]. First, 75 µL of goat-anti-mouse IgG (0.1 mg/mL in salamander Ringer) was placed inside the ring for 3 h; the goat-anti-mouse IgG solution was then aspirated off of the MEA, washed twice with 75 µL of salamander Ringer, and finally coated with 75 µL Sal-1, overnight, at 10 °C. Care was taken during the coating and washing steps to ensure no liquid spilled onto the p-HEMA surface. Immediately prior to optical tweezing, the Sal-1 was aspirated and the MEA washed twice with salamander Ringer before filling the culture area with 1 mL of ice-cold salamander media. The MEA was placed into a sterile 100 mM Petri dish.
To reuse MEAs, PDMS rings were removed and thrown away. Vaseline was removed, using cotton swabs soaked in isopropyl alcohol. The MEA was then briefly dipped in 70% EtOH, washed 3 times in DI H2O, and finally soaked in 1% Tergazyme solution, overnight, before being extensively washed with DI H2O and stored in DI H2O.
2.4. Optical Tweezing
Optical tweezing was performed in a similar manner to that previously described [
8,
9,
10]. Briefly, an 800 mw 980 nm continuous wave diode laser (Cell Robotics Inc., Albuquerque, NM, USA) was mounted onto an Axiovert 100 inverted light microscope (Carl Zeiss Inc., Oberkochen, Germany) with a television port. Laser light was directed through a custom-drilled hole in the bottom of the microscope, allowing for an optical path through the objective to the cells. The microscope was equipped with a motorized stage (Cell Robotics Inc.), and both the laser and stage were controlled by a custom software package based on Labview (National Instruments, Austin, TX, USA) from Bloomy Controls (Windsor, CT, USA). Laser intensity was controlled through an NI DAQ UMI7764 controller (National Instruments), and stage control and motion control feedback were connected to the NI DAQ through a motor control unit MCDC 3006S (Faulhaber, Schönaich, Germany). For optical tweezing, a 40× oil immersion plan neofluor objective with a numerical aperture of 1.3 was used with brightfield optics.
A prepared MEA was placed upon the stage, and the central electrode array area was set as the home reference point. Then, 50 µL of cell suspension was pipetted onto the area covered by p-HEMA, using a 1000 µL tip. The MEA was covered with a sterile 35 mM Petri dish cover, to maintain sterility. The p-HEMA area was then searched for retinal cells. Cells chosen for tweezing were those that appeared “unattached” to the p-HEMA surface, determined by gently tapping the microscope stage and observing the reaction of the cells. Identifiable cells were imaged, and their initial location was recorded. A cell was then picked up by the laser beam, by focusing slightly above the cell, setting the optical tweezers to 40–100% power (depending upon the difficulty in picking up the cell), and gradually focusing upward from the surface of the MEA. The power was then lowered to 25%, and the cell, thus “optically trapped”, was moved toward the central electrodes by moving the stage. At the electrodes, the power was lowered again to 10%, and the cell slowly maneuvered over an electrode. After maneuvering, laser power was lowered to 6%, and the focus was brought closer to the MEA electrodes (see
Supplementary Materials Video S1). Finally, the laser was turned off, and the cell was allowed to gently settle onto the electrode. As shown in
Supplementary Materials Video S1, care was taken to minimize movement over the ITO tracks and to prevent focusing the laser directly upon the MEA electrodes, as these actions could damage the electrodes and insulation. When directly focused on the electrodes or tracks, the laser beam, at high power, was absorbed and caused significant damage to both. The final location of each cell was logged and compared to the starting location. Success of movement of cells was defined as being able to pick up a cell, move it to an electrode, and place it onto an electrode. One-third or more of the electrode must also have been covered by the cell. After cells were plated and moved, they were maintained at 10 °C. Cells were allowed to grow for 1 week and imaged daily, using brightfield optics on an Axiovert 135 (Carl Zeiss Inc., Oberkochen, Germany).
2.5. Ca2+ Imaging and Stimulation
Calcium levels in optically tweezed retinal cells were measured, using Fura-2 AM (Life Technologies, Cat# F1221, Carlsbad, CA, USA). Then, 5 µL of 1 µg/µL Fura-2 in DMSO was added to 1 µL of 20% pluronic acid (Molecular Probes, P-6867, Eugene, OR, USA) and briefly vortexed. A total of 1 mL of salamander Ringer was then added and vortexed for 5 min. After rinsing the MEA with cells in salamander Ringer for 5 min, 1 mL of Fura-2 containing solution was added. Cultures were incubated at RT for 40 min, while gently rocking. Fura-2 solution was then carefully aspirated off, and the MEA was washed with 1 mL of salamander Ringer, for 30 min, at RT, while rocking. Fura-2 was excited at 340 and 380 nm, using a Polychrome V (Till Photonics, Bavaria, Germany) monochromator coupled to an inverted Leica DM IRB microscope. Image acquisition was accomplished with an Imago CCD camera (Till Photonics) controlled by an Image Control Unit (Till Photonics). Cells were exposed for 50 ms to each wavelength and imaged every second. Signals were averaged over the entire cell not covered by the electrode.
Stimulation was accomplished with an MEA1060 preamplifier (Multi Channel Systems MCS GmbH) with an attached STG4002 stimulator (Multi Channel Systems MCS GmbH). Parameters for stimulation were set by using the MC Stimulus II software provided by Multi Channel Systems MCS GmbH. We used two different types of stimulation: rectangular and sinusoidal. Rectangular stimulations consisted of 100 µs negative stimulation, followed by 60 µs of off, followed by 100 µs of positive stimulation. The sinusoidal stimulation frequency was set to 5 Hz. For each type of stimulation, we used three amplitudes, −100, −200, and −400 mV. In addition, we used a −1 µA current stimulation for sinusoidal stimulation. All electrodes not covered by cells were connected to ground. Prior to stimulation, cells were imaged for 50 s, to establish a baseline calcium reading. After stimulation, cells were allowed to recover, to return to baseline, before testing other stimulation parameters. Cells which did not display any increase over their baseline response at −400 mV rectangular or −1 µA sinusoidal stimulation were not analyzed. The number of unstimulated rod photoreceptors was not quantified, as it was difficult to determine if the lack of stimulation was due to an MEA defect, slight movement of the cells off of the electrode, incomplete penetration of Fura-2 AM into the cell, or lack of response by the cell to extracellular stimulation. Significance was achieved at p < 0.05, as determined by Student’s t-test.
Stimulation was accomplished with an MEA1060 preamplifier (Multi Channel Systems MCS GmbH) with an attached STG4002 stimulator (Multi Channel Systems MCS GmbH). Parameters for stimulation were set by using the MC Stimulus II software provided by Multi Channel Systems MCS GmbH. We used two different types of stimulation: rectangular and sinusoidal. Rectangular stimulations consisted of 100 µs negative stimulation, followed by 60 µs of off, followed by 100 µs of positive stimulation. The sinusoidal stimulation frequency was set to 5 Hz. For each type of stimulation, we used three amplitudes, −100, −200, and −400 mV. In addition, we used a −1 µA current stimulation for sinusoidal stimulation. All electrodes not covered by cells were connected to ground. Prior to stimulation, cells were imaged for 50 s, to establish a baseline calcium reading. After stimulation, cells were allowed to recover, to return to baseline, before testing other stimulation parameters. Cells which did not display any increase over their baseline response at −400 mV rectangular or −1 µA sinusoidal stimulation were not analyzed. The number of unstimulated rod photoreceptors was not quantified, as it was difficult to determine if the lack of stimulation was due to an MEA defect, slight movement of the cells off of the electrode, incomplete penetration of Fura-2 AM into the cell, or lack of response by the cell to extracellular stimulation. Significance was achieved at p < 0.05, as determined by Student’s t-test.