Dishes of cells were transferred to an inverted microscope (Nikon TMS) and superfused by gravity with a modified bicarbonate-buffered Krebs’ solution
26 °C). Voltage-clamp experiments were carried out using single patch pipettes to achieve the whole-cell or perforated patch recording configuration (Hamill et aL, 1981; Horn and Marty, 1988; Rae et aL, 1991). Patch pipettes were pulled from thin-walled, filamented glass (Clark Electromedical) using either a Flaming/ Brown programmable horizontal puller (Sutter) or a List vertical puller. To minimize pipette/ bath
capacitance, pipettes were coated near the tip with Sylgard; pipette tips were then fire- polished (using a Narishige microforge) to enhance giga-seal formation, giving
resistances of 4-7 M Q (for whole-cell experiments) or 2-5 M H (for perforated patch experiments). Patch pipettes were back-filled with one of the intracellular solutions detailed below. For perforated patch recordings using the pore-forming antibiotic agents nystatin or amphotericin B, patch pipettes were first front-filled with antibiotic- free intracellular solution (by dipping for 20-60 seconds), which allowed time for seal formation before permeabilization began. Patch pipettes were then fitted onto a Ag wire coated with AgCl, connected to an Axoclamp 2A amplifier (Axon instruments).
Pipettes were positioned above cells in the extracellular solution. The voltage difference across the pipette was set to zero in bridge mode; the reference electrode was a KCl/ agar bridge connected to a pool of 3 M KCl which was in turn connected to ground. This was checked again at the end of the experiment to monitor any drift in the voltage detected across the pipette: the voltage offset at the end of the experiment was subtracted from the cell membrane potentials recorded. While in bridge mode, the pipette voltage in response to a current injection (0.1 nA, 20 ms duration every 100 ms) was continuously monitored using an oscilloscope (Tektronix). The voltage (V) drop across the tip of the pipette during current (I) passing was compensated prior to seal formation using the bridge balance; this provided an estimate of pipette resistance (R) from Ohm’s Law (V = I.R). The pipette was gently touched on to the cell using a Narishige hydraulic manipulator. Contact was monitored (using the oscilloscope) by a voltage drop in response to the injection of current, due to the increase in resistance. A high resistance seal (of at least 1 GO) was formed between the pipette and the
membrane by the application of gentle suction. Seal formation was monitored by the increasing voltage deflection in response to the current injection. For whole-cell recordings, the whole-cell configuration was achieved by applying a short increase in
suction to break the cell membrane under the patch pipette, and monitored by a sudden decrease in the voltage deflection, to achieve access resistances of 10-15 Mfl. For perforated patch recordings, the pore-forming agent used in the pipette was allowed to permeabilize the cell membrane under the patch pipette (monitored by the slowly decreasing voltage deflection), usually in 5-10 minutes, to access resistances of 30-40 Mfl (nystatin) or 20-30 M fl (amphotericin B). During the experiment there was usually a further decrease in access resistance of around 10 MH when using either agent.
Neurones were voltage-clamped in discontinuous single electrode voltage-clamp mode (Finkel and Redman, 1984). The duty cycle had a 30% period of current passing (the amplitude being related to the difference between the recorded voltage and the command voltage, according to the gain), and the sampling frequency was set so that the voltage response across the electrode had decayed prior to voltage sampling (this was monitored using a second oscilloscope). Typically, the sampling frequency was 5- 7 kHz for whole-cell recordings and 3-5 kHz for perforated patch recordings. To establish optimum voltage-clamp conditions, electrode capacitance (monitored from the electrode voltage response throughout a duty cycle) was first optimally compensated (sampling frequency was then adjusted at this stage). A 10 mV (hyperpolarizing) step was then applied (20 ms, every 100 ms), and the gain (for the current-passing circuit) was increased until the voltage response was square, with no overshoot. Phase was adjusted if it offered improvement to the voltage response, and the antialias filter was increased to reduce noise on the voltage and current signals without altering the settling characteristics during the duty cycle.
SCG membrane resistance (Rm), calculated from the current response to a Is, 10 mV voltage step from the zero current potential (in voltage clamp) using Ohm’s Law (V = I.R), was 350 ± 30 MO (n = 11 cells). Membrane capacitance (Cm) was estimated from the slow membrane time constant (im), determined from a fit of a two exponential function to the decay of the capacitance current during a -10 mV step from
the zero current potential in voltage-clamp mode (t^ = Rm-Cm). Average membrane capacitance was 92 ± 11 pF (n = 11).
2.2.2. Solutions.
Modified Krebs' solution: NaCl, 120 mM; KCl, 3 mM; CaCl2, 2.5 mM; MgCl2, 1.2
mM; NaHCOs, 22.6 mM; HEPES, 5 mM; D-glucose, 11.1 mM; tetrodotoxin, 0.5 pM; pH 7.4 approximately, when bubbled with 95% O2/ 5% CO2. Osmolarity was between 270-290 mOsm. In some experiments the [KCl] was increased to 15 mM. The flow rate was 7 ml/ minute when using the local perfusion system, and 13 ml/ minute when using the distant perfusion system in conjunction with the U-tube-1 ike device (see below).
K acetate-1 K gluconate-based whole-cell intracellular solution: Potassium acetate (or
potassium gluconate) 80 mM; KCl, 30 mM; MgCl2, 3 mM; HEPES, 40 mM; EGTA, 3 mM (estimated Ca^"^ concentration is less than 1 nM*). For some experiments, ATP and GTP (Mg^"^ salts) were added to give a final concentration of 2 mM ATP and 0.5 mM GTP.
*Using the Ca^'*’-sensitive dye, indo-1, the basal Ca^^ concentration in the K rebs’ solution has been estimated to be up to 10 pM: addition o f 3 mM EGTA buffers this basal Ca^^
concentration to almost zero (Dr. S. J. Marsh, personal communication).
Intracellular solution fo r perforated patch recordings: nystatin or amphotericin B were
suspended in dimethyl sulphoxide (0.5% and 0.2% respectively) by sonication, freshly each day, and then diluted to a final concentration of 0.25 mg/ ml nystatin or 50 pg/ ml amphotericin B in a K acetate-based solution similar to above, but containing no EGTA, ATP or GTP.
All intracellular solutions were adjusted to around pH 7.2 with a 10 M solution of KOH, and to between 270 and 280 mOsm using 1 M solutions of K acetate or K gluconate, giving a total intracellular [K^] of 125 mM. When using the K acetate- based solution (and accounting for the ionic activity of K acetate), the equilibrium potential for K"^ ions (calculated from the Nernst equation) was -94 mV in 3 mM
extracellular [K^] and -53 mV in 15 mM extracellular [K^]; the equilibrium potential for Cl ions was -33 mV.
The liquid junction potential between the intracellular and extracellular solutions was estimated from the method of Neher (1992). Using the KCl reference electrode, the pipette voltage was zeroed with intracellular solution in the pipette and in the bath. The bath was then superfused with the extracellular Krebs’ solution. Using the K gluconate intracellular solution, a liquid junction potential of -7 mV was measured (outside with respect to inside). Using the K acetate solution, a liquid junction
potential of -6 mV was measured on superfusing Krebs’ solution containing 3 mM KCl, and -5 mV on superfusing Krebs’ solution containing 15 mM KCl. These values have not been subtracted from measurements of membrane potential.