Imaging Exocytosis in Retinal Bipolar Cells with TIRF Microscopy

Total internal reflectance fluorescence (TIRF) microscopy is a technique that allows the study of events happening at the cell membrane, by selective imaging of fluorescent molecules that are closest to a high refractive index substance such as glass1. In this article, we apply this technique to image exocytosis of synaptic vesicles in retinal bipolar cells isolated from the goldfish retina. These neurons are very suitable for this kind of study due to their large axon terminals. By simultaneously patch clamping the bipolar cells, it is possible to investigate the relationship between pre-synaptic voltage and synaptic release2,3. Synaptic vesicles inside the bipolar cell terminals are loaded with a fluorescent dye (FM 1-43®) by co-puffing the dye and a ringer solution containing a high K+ concentration onto the synaptic terminals. This depolarizes the cells and stimulates endocytosis and consequent dye uptake into the glutamatergic vesicles. After washing the excess dye away for around 30 minutes, cells are ready for being patch clamped and imaged simultaneously with a 488 nm laser. The patch pipette solution contains a rhodamine-based peptide that binds selectively to the synaptic ribbon protein RIBEYE4, thereby labeling ribbons specifically when terminals are imaged with a 561 nm laser. This allows the precise localization of active zones and the separation of synaptic from extra-synaptic events.


Prepare solutions listed in
; The pH of ringers' (external) solutions should be adjusted to 7.4 with NaOH and the pH of the internal solution should be adjusted to 7.2 with CsOH. Protect internal solution from light with aluminum foil and keep it at 4°C until use; 2. Dark-adapt a goldfish for at least 30 minutes prior to dissection; 3. While the animal dark-adapts, prepare 5 mL of the hyaluronidase (type V hyaluronidase,1100 units/mL in low Ca 2+ ringer's; Sigma, St. Louis, MO) and 10 mL of the L-cysteine (0.5 mg/mL in low Ca 2+ ringer's) solutions and weigh the papain (lyophilized powder, 40 units/mL; Sigma, St. Louis, MO) for 5 mL of the digestion solution; 4. Euthanize the goldfish by quick decapitation with surgical scissors and destroy the brain and spinal cord with a #11 scalpel blade; 5. Remove the eyes by destroying the extra-ocular muscles with the help of a #7 curved Dumont forceps and cutting the optic nerve with iris scissors; 6. Place one eye bulb on a piece of filter paper and puncture the scleral limbus with the tip of a #11 scalpel blade; 7. Introduce the blade of a pair of vannas scissors inside the punctured whole and cut the whole anterior segment away; 8. Place a small piece of filter paper on top of the remaining optic cup and exert some pressure in order to have the paper soak with vitreous humor; 9. Lift the filter paper with the retina attached to it and cut the optic nerve with a par of vannas scissors; 10. Place the filter paper containing the retina in a 35 mm plastic culture dish with hyaluronidase solution and peel off the retina from the filter paper with the help of #7 Dumont tweezers; 11. Cut the retina into 4-6 pieces with half of an industrial carbon steel single-edged blade and let it sit in the hyaluronidase solution for 20 minutes; 12. While waiting for hyaluronidase to take effect, add 5 mL of the L-cysteine solution to the papain and let it sit until the liquid becomes transparent (approximately 5-10 minutes); 13. Wash the pieces of retina 3x in low Ca 2+ Ringer's and let them sit in the papain solution for 30-35 minutes; 14. Wash the pieces of retina 3x in low Ca 2+ Ringer's and store them until use at 4°C in a 35 mm plastic culture dish containing low Ca 2+ Ringer's; 15. To dissociate the cells, put a piece of retina in a microcentrifuge tube containing 500 mL of low Ca 2+ Ringer's and slowly triturate the retina by pipetting it up and down with a glass dissociation pipette, carefully not to produce any air bubbles. Dissociation pipettes are fabricated by heating up the tip of a glass Pasteur pipette with a Bunsen burner and slightly bending it with the help of anatomical forceps; 16. Plate the isolated cells by adding a drop of the retinal suspension to a home-made recording chamber previously filled with 2 mL of low Ca 2+ ringer's. The chamber consists of the bottom half of a 35-mm plastic culture dish with a circular whole in the middle and a circular coverslip of 1.78 refractive index glass (PlanOptik, Germany) glued to the bottom with a silicon elastomer (Sylgard 184; Dow Corning, Midland, MI).

Part 2: Bipolar Cell Loading and Wash Out
TIRF imaging of synaptic vesicles is best carried out using an objective-type TIRFM microscope with a very high NA objective and a sensitive camera. For our experiments, we choose to use a 1.65 NA objective (Apo x100 O HR, N.A. 1.65, Olympus, Japan) with an EMCCD (Cascade 512B, Roper Scientific, Tucson, AZ). The use of the very high NA objective necessitates the use of high refractive glass coverslips and immersion fluid (di-iodomethane with sulfur). Under our conditions, excitation light is limited to an exponentially decaying field with a length constant of approximately 50 nm.
2. Place the recording chamber carefully on top of the microscope objective and carefully mount ground electrode and superfusion exit pipe to chamber; 3. Let chamber sit on microscope for 10-20 minutes to allow cells to sink and adhere to the bottom; 4. In the meantime, prepare 5 mL of a 1mM trolox® ((±)-6-Hydroxy-2,5,7,8-tetramethylchromane-2-carboxylic acid; Sigma, St. Louis, MO) solution in high K + ringer's. Sonicate until dissolved; 5. Prepare 15 mL of ADVASEP-7 washing solution: 1 mM ADVASEP-7 (Sigma, St. Louis, MO) in low Ca 2+ ringer's. Note that ADVASEP-7 use is optional and can be omitted if desired; 6. Purge the superfusion lines and add ADVASEP-7 washing solution, low Ca 2+ ringer's and control ringer's to the superfusion system; 7. Pull loading pipettes from thin-walled borosilicate glass (Kwik-Fil® TW150-3; WPI, Sarasota, FL). Puffer pipette resistances are in the 1.5-2.5 MΩ range; 8. Prepare the FM1-43® (N-(3-triethylammoniumpropyl)-4-(4-(dibutylamino)styryl)pyridinium dibromide, "special packaging"; Invitrogen, Carlsbad, CA) solutions. First, make a 1 mM stock by adding 160 μL distilled water to one vial (100 mg) of FM1-43®. This stock can be kept at 4°C for up to one week. Then, add 5 μL of FM1-43® to 1 mL high K + ringer's + 1mM trolox®. Protect solution from light with aluminum foil and keep it at 4°C until use; 9. Turn the microscope bright field light on and search for intact bipolar cells. Slightly tap the microscope to make sure that the neurons are firmly attached to the bottom of the chamber; 10. Position the superfusion pen close to the cell of interest and continuously perfuse the preparation with low Ca 2+ ringer's; 11. Turn off the room lights and add a red long pass filter (i.e. RG630; Schott, Germany) to the optical path to minimize excitation of the FM dye; 12. Fill a loading pipette with 10 μL of the FM dye solution, mount the pipette in the micromanipulator and lower the pipette onto the preparation without overpressure until it is at the same focal plane as the bipolar cell you want to load. Make sure you have at least two electrode holders: the one used for the FM dye cannot be used for patch clamping, or else it may contaminate the intracellular solution; 13. Position the puffer opening at a distance of around 10 μm from the axon terminal, turn the superfusion system off and puff the dye solution for 10 seconds by turning the pipette overpressure on; 14. Turn the overpressure off and, without moving the pipette, wait for 30 seconds; 15. Turn the superfusion on and bathe the chamber in ADVASEP-7 solution for 5 minutes. In the meantime, remove the puffer pipette from the bath; 16. After 5 minutes, switch to low Ca 2+ ringer's and perfuse the chamber for 25-30 minutes to allow removal of excess dye.

Figure 1:
The experimental setup. A 488 nm laser (blue) is focused to the periphery of the back focal plane of the objective and suffers total internal reflection when it reaches the glass-aqueous medium interface. The electromagnetic field generated by the reflected beam excites the fluorophore loaded into the synaptic vesicles closest to the bottom of the glass chamber, which then fluoresce (green). The fluorescent light is then guided to the observer's eye (depicted) or to a CCD camera. The membrane potential of the imaged cells is controlled simultaneously by patch-clamping them. This approach allows the study of the relationship between incoming signals (the membrane voltage) and the neuronal output (exocytosis).

Discussion
The advantages of objective-type TIRF microscopy are that 1) it provides excellent optical sectioning by restricting excitation light to a narrow region within the focal plane of the objective, thereby minimizing out-of-focus light; 2) since light drops exponentially with distance, movement in a vertical direction can be monitored as a change in fluorescence intensity; 3) efficient light collection through the high numerical aperture objective 1,5 .
The main drawback of the technique is that it is limited to imaging events happening within 100& nm of the cell surface, which is roughly equivalent to an ultrathin section in electron microscopy. Therefore, visualization of these events depends critically on the cells being firmly adhered to the glass, on the presence of synaptic ribbons close to the patch of membrane adhered to the glass and on the successful loading of vesicles. Our protocol enables the loading of only 1-2% of the total number of vesicles within the bipolar cell synaptic terminal 2,6 . With that said, it is clear that there are much more events happening at the cell surface than the ones we are able to image.