| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
MS 8442 Received 7 July 1998; accepted after revision 8 December 1998.
 |
ABSTRACT |
|---|
 TopAbstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Transduction of light into an electrical signal takes place at the outer segment of retinal rods and cones. In this specialized cell compartment, an orderly sequence of molecular interactions is optimized to signal light changes over a wide range of light intensities (Sather & Detwiler, 1987). While the molecular steps of the phototransductive process have been detailed and explained by quantitative models (Lamb & Pugh, 1992; Nikonov et al. 1998), a much less exhaustive account is at present available for the subsequent events of signal processing in photoreception.
The observation that the waveform of the photocurrent (Baylor et al. 1979, 1984b) is different from the photovoltage (Detwiler et al. 1978; Schneeweis & Schnapf, 1995) indicates that the transduced signal is further elaborated by rods before its transmission to second-order neurones. In cold-blooded vertebrates, the difference between the kinetics of photovoltage and photocurrent has been attributed to gating by the light-induced hyperpolarization of two voltage-dependent conductances.
The first one is a caesium-sensitive conductance (Fain et al. 1978), subsequently characterized as a hyperpolarization-activated current (Ih; Bader et al. 1982), with a low selectivity between sodium and potassium (Bader & Bertrand, 1984; Wollmuth & Hille, 1992; Wollmuth, 1995). Ih has also been described in cones from salamander (Barnes & Hille, 1989) and monkey (Yagi & McLeish, 1994). A second voltage-dependent current, selective for potassium and blocked by barium, has been identified in salamander rods as Ikx (Beech & Barnes, 1989). A voltage-dependent potassium current had previously been suggested (Owen & Torre, 1983) to contribute to the high-pass filtering of small signals in the rod network of amphibians (Detwiler et al. 1978, 1980; Torre & Owen, 1983). Although Schneeweis & Schnapf (1995) have recently suggested the presence of voltage-dependent currents in monkey photoreceptors, the properties and functional roles of such currents in mammalian rods are still unknown.
We have investigated the properties and functional roles of the voltage-dependent conductances gated by membrane hyperpolarization in guinea-pig rods, by means of the perforated-patch-clamp technique (Horn & Marty, 1988). We have found that two distinct voltage-dependent currents are present in guinea-pig rods: (a) a Cs+-sensitive current activated by membrane hyperpolarization (Ih), and (b) a Cs+-insensitive current deactivated by membrane hyperpolarization and inhibited by Ba2+ (Ikx). Both Ih and Ikx contribute to the bandpass amplification observed in the voltage response of rods stimulated with sinusoidal currents of 0·2-40 Hz frequency. This bandpass amplification takes place before the light-induced hyperpolarization spreads to the rod synapse, and may therefore improve the temporal fidelity of signals transmitted through the rod-bipolar synapse.
The proposed role of Ih in the voltage response of rods is in agreement with the effect of Ih blockers on the temporal resolution of the first synapse in the rod pathway of mammals (Gargini et al. 1999).
Preliminary results of this study have been presented to the Biophysical Society (Demontis & Cervetto, 1995) and to the Physiological Society (Demontis & Cervetto, 1996).
|
METHODS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Preparation
Adult albino guinea-pigs (250-400 g) obtained from a local supplier (Stefano Morini S.a.S., S. Polo d'Enza, Italy) were kept on a 12:12 light:dark cycle, and reared in accordance with the rules of the local Animal Welfare Committee for the Care and Use of Laboratory Animals. On the day of the experiment, the animal was dark-adapted for 1 h and then anaesthetized by an initial intraperitoneal injection of 35 mg kg-1 pentothal sodium (Gellini S.p.A., Aprilia, Italy). After enough anaesthetic had been provided to fully suppress corneal reflexes, the eye was quickly (6 min) enucleated in dim red light, and the animal killed by an intraperitoneal lethal dose of anaesthetic (350 mg kg-1 pentothal sodium). After enucleation, the anterior pole was discarded and the posterior pole was bathed in Locke's solution, the composition of which was (mM): NaCl, 140; KCl, 3·6; CaCl2, 1·2; MgCl2, 2·4; glucose, 10; Hepes, 10; pH 7·6 with tetramethylammonium hydroxide (TMA-OH). After 15 min at 8°C, the retina was dissected free from the sclera with the help of an infrared converter and placed for 25 min in a solution containing 0·3 mg ml-1 hyaluronidase and 30 U ml-1 DNAse at 34°C. This enzymatic treatment does not affect the electrical properties of rods, since in a few cells we were able to record similar currents without enzymatic treatment. However, it was easier to obtain giga-seals in rods treated with hyaluronidase. After incubation, the retina was washed three times in cold Locke's solution and then placed in a plastic dish with 3 mg ml-1 ovomucoid. All chemicals were from Sigma (Sigma-Aldrich S.r.l., Milan, Italy).
Recordings
The retina was finely chopped with a razor blade, and an 80 µl aliquot of the preparation was transferred to the 200 µl recording chamber placed on the stage of an inverted microscope (Olympus IMT-2). Solutions were gravity-fed to the chamber at a rate of 0·8 ml min-1, and a six-way valve allowed selection among different solutions. Line purging by a four-way Hamilton microvalve close to the experimental chamber reduced the dead volume to a few microlitres. Removal of solution was by a peristaltic pump (P1-pump, Pharmacia Biotech S.p.A., Cologno Monzese, Italy). CsCl or BaCl2 were added to Locke's or tetraethylammonium (TEA)-Ba2+-K+ solution (see Fig. 4 legend) with small increases in osmolarity.
Patch pipettes were drawn on a two-stage horizontal puller (BB-CH, Mecanex, Nyon, Switzerland) from 100 µl haematocrit borosilicate tubes (Brand GMBH, Wertheim, Germany) to a bubble number of about 3·5-4 and coated with Sigmacote (Sigma-Aldrich). Their resistance in Locke's solution was between 5 and 10 M
, when filled with an intracellular solution of composition (mM): KCl, 140; Na2ATP, 5; LiGTP, 1; Pipes, 5; pH 7·2 with KOH (Schneeweis & Schnapf, 1995).
After obtaining a seal in excess of 10 G, neither further suction, nor an electrical pulse, provided electrical access to the cell and in most cases they led to seal breakdown. Electrical access was easily obtained by including Amphotericin B (3 mg in 100 µl DMSO) in the intracellular solution at a final concentration of 210 µg ml-1. The process of pore formation was followed in the current-clamp mode by the appearance of a progressively negative membrane potential, or by the presence of stable currents in the voltage-clamp mode. When the currents underwent abrupt changes, the recording was discontinued.
Perforated-patch recordings had a series resistance up to 50 M, which was 50 % compensated. This partial compensation is not expected to cause a large error in the applied voltage, since in most cases the maximal current recorded was around 100 pA; consequently, the associated error would be at most 2·5 mV (for steps from -35 to -110 mV), less than 5 %. Moreover, considering the rod capacity of 2-6 pF (corresponding to time constants of 100-300 µs), the partial compensation of the series resistance is not expected to affect the kinetics of voltage-dependent currents, which are activated with time constants in excess of 20 ms at room temperature. Junction potentials were measured in accordance with the method described by Neher (1992).
Data were recorded with an Axopatch-1D amplifier (Axon Instruments), and were filtered, unless specified, at 200 Hz with a 4-pole Bessel filter, acquired by a Digidata 1200 A/D board (Axon Instruments), at a sampling rate of 500 Hz, and stored on line on the hard disk of a 486-microprocessor-based personal computer.
Stimulation
Stimulation was either by 2 s steps or by sinusoidal currents whose frequency ranged from 0·1 to 50 Hz. For voltage-clamp experiments, the voltage steps ranged in amplitude from -110 to +50 mV, from a holding voltage of -35 mV, unless otherwise specified. For current step injections in current-clamp experiments, the injected current ranged from +2 to -18 pA, well in the range of dark-current amplitudes (up to -20 pA) in guinea-pig rods (Demontis et al. 1995). In most cases, in order to mimic the light response of rods, the current steps were applied to cells whose membrane potential was depolarized at about -40 mV by injecting a steady current of 8-16 pA.
The voltage response to sinusoidal current stimuli was recorded after the application of about 30 stimuli, in order to measure steady-state effects. In order to reduce noise, 16 responses were averaged off-line for both current and voltage.
Data analysis
Unless otherwise stated data are given as means ± S.E.M. Data were analysed by Mathcad 5.0 (MathSoft Inc., Cambridge, MA, USA) and plotted by Origin (MicroCal Software Inc., Northampton, MA, USA). Estimates of the parameters describing the activation of the ionic conductances responsible for the inward rectification were obtained by fitting eqn (1)) to the I-V data obtained from the last 10 ms of the 2 s activation steps:
where Gh, Eh, V½h and Sh are the maximal conductance, the reversal potential, the half-activation voltage and the inverse slope factor, respectively, for Ih; Gkx, Ekx, V½kx and Skx have the same meaning for Ikx. Unless otherwise specified, Eh and Ekx were kept fixed at -31 and -87 mV, respectively. Equation (1) provides estimates for Ih and Ikx activation parameters in close agreement with those produced by the Cs-difference method (Ih isolated by subtracting the recordings in 3 mM CsCl from control recordings). As the parameters describing Ih and Ikx activation in 30 cells are not normally distributed, the Wilcoxon rank test for pairs was adopted for the statistical comparison of Ih and Ikx activation parameters in Tables 1 and 2. GL and EL are the conductance and reversal potential of the linear component of the membrane current. This linear component tends to increase during recording, and has an apparent reversal potential close to -20 mV. As detailed in the Appendix, the properties of this linear current are expected from the combination of the current through the seal resistance and the electrogenic current generated by the Na+-K+ pump in response to the influx of sodium through the recording pipette.
For the data presented in Fig. 5, the linear resistive and capacity currents were removed by subtracting on-line a 1/5 scaled version of the stimulus (P/5 protocol). The parameters describing the instantaneous I-V relationship of rods were estimated by fitting a Goldman-Hodgkin-Katz (GHK) model, described by eqn (2)), which includes terms for sodium and potassium permeability, to the I-V curves of Fig. 5:
where PNa and PK are the sodium and potassium permeability, respectively, z, F, R and T have their usual meanings, and 
= zF(R T)-1.
Guinea-pig rods are modelled as the parallel combination of an inductance (L), a purely resistive arm (R1) and the resistive component of the inductance (R2) (see Fig. 9) as described in Detwiler et al. (1980) and also by considering the membrane capacitance (Cm) (Baylor et al. 1984a). R1, R2 and L were estimated from the voltage response to the inward current step injection, in accordance with Baylor et al. (1984a).
Briefly, R1 was estimated from the equation R1 = 
Vpeak/I, where Vpeak = Vpeak - Vbaseline. Vpeak is the voltage at peak, Vbaseline is the voltage before stimulation and I is the stimulus intensity. R2 was estimated at a steady state by considering the resistance of the parallel combination (Rtot) of R1 and R2, in accordance with the equation R2 = (RtotR1)/(R1 - Rtot), where Rtot = Vss/I. Vss is the amplitude of the response when the voltage has relaxed to a steady state (Vss - Vbaseline). L was estimated from the time constant (
ex) of the relaxation from Vpeak to Vss, in accordance with the equation L = ex(R1 + R2).
The modulus of the complex impedance (|Z(f)|) of this circuit was computed by eqn (3):
where 
= 2
f, and f is the stimulation frequency. As the values for the modulus of the complex impedance include a constant, the experimental and theoretical data plotted in Fig. 7 were normalized to 0·2 Hz.
The theoretical phases (
(f)) of the voltage signals generated by sinusoidal current injections into the circuit shown in Fig. 9 are described by:
where Cm, L, R1, R2 and have the same meaning as above.
The experimental phases ('(f)) are computed by:
'(f) = 2f(tI,peak - tV,peak), (5)
where f is the stimulation frequency, tI,peak the peak time of current and tV,peak the peak time of voltage.
|
RESULTS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Effects of caesium, TEA and barium on inward rectification in rods
Guinea-pig rods had an average membrane potential of -52·6 ± 1·5 mV (n = 30) when recorded in bright light. In these conditions, their input resistance (Rin), measured in the linear region of the I-V relationship (between -35 and -30 mV), was 2·5 ± 0·15 G (n = 30).
Linearity in the I-V relationship broke down for hyperpolarizations negative to -45 mV, with time- and voltage-dependent membrane rectification becoming manifest, as shown in Fig. 1 for both current-clamp and voltage-clamp recordings in the same cell. Traces in Fig. 1A are current-clamp recordings of voltage changes elicited by 2 s steps of inward current of graded amplitude. A transient peak, followed by relaxation to a plateau characterizes the voltage response. This plateau becomes more prominent in response to large current steps. At the end of the 2 s stimulation, a transient depolarization develops, peaking at -38·8 mV and eventually decaying to the resting membrane potential of -45·5 mV.
 |
View larger version [in this window] [in a new window] |
|
|
A, current-clamp responses to 2 s current injections, ranging in amplitude from 0 to 17·5 pA in 2·5 pA increments. Membrane potential in bright light was -45·5 mV. B, voltage-clamp responses to 2 s hyperpolarizing and depolarizing voltage steps, from the same cell as in A. Holding voltage was -35 mV; voltage stimuli ranged from -110 to +50 mV in 10 mV increments, before stepping back to -60 mV for 1 s. Time calibration applies to both A and B. The dotted line in B is the zero current level. Two points have been removed from the capacity transients. C, | ||
The recordings in Fig. 1B show the time course of currents elicited by depolarizing and hyperpolarizing voltage steps of increasing amplitude recorded from the same rod as in Fig. 1A, but in the voltage-clamp mode. Note the slow inward relaxation for voltages ranging from -40 to -110 mV. This inward relaxation is evident when the steady-state current amplitudes (measured at the end of stimulation) are plotted against voltage in Fig. 1C (filled symbols). The smooth line is a best fit of eqn (1)) to the data points (see figure legends). The good fit to data obtained with eqn (1) supports the idea that at least two voltage-dependent conductances contribute to the electrical properties of the rod membrane in bright light. The average values of the parameters obtained by eqn (1) from the I-V relation of 30 cells recorded by the perforated-patch technique were Gh = 0·98 ± 0·04 nS, V½h = -74·8 ± 1·3 mV, Sh = 8·2 ± 0·4 mV, Gkx = 0·2 ± 0·01 nS, V½kx = -52·9 ± 1·7 mV and Skx = 5·9 ± 0·4 mV.
The voltage-dependent currents underlying rod inward rectification were further investigated by analysing the effect of different pharmacological blockers. The responses of a guinea-pig rod to steps of inward currents of two different amplitudes are shown in Fig. 2A, before, during and after application of 3 mM CsCl. CsCl reversibly blocked the initial transient component elicited by a 15 pA inward current injection, and affected the time course of membrane potential recovery at the end of the stimulus, by reducing and delaying the amplitude of the transient component (see Discussion). On the other hand, CsCl caused a slight reduction in the relaxation induced by an inward current of 5 pA, and had a marginal effect on the rate of potential recovery at the end of the stimulus. Note that CsCl reversibly shifted the membrane potential from -41·5 to -43·5 mV; the junction potential associated with the CsCl-containing solution does not account for this effect.
 |
View larger version [in this window] [in a new window] |
|
|
A, current-clamp responses to 2 s inward current steps of -5 and -15 pA, before (thin traces), during (thick traces) and after (dotted traces) application of 3 mM CsCl. The resting potential in bright light was -53 mV and was brought to -41 mV by the injection of a steady outward current of 5·1 pA. B, voltage-clamp responses to 2 s voltage steps to -50 and -80 mV, before stepping back to -70 mV, from the same cell as in A. Holding voltage was -35 mV. Control (thin traces), during application of 3 mM CsCl (thick traces) and after washing out CsCl (dotted traces) sweeps are plotted. Two points have been removed from the capacity transients. C, open and filled symbols are average current amplitudes, measured during the last 10 ms of the 2 s voltage steps (thick dot in B) in control and CsCl, respectively. The smooth line through the experimental points in control is a best fit of eqn (1) with: Gh, 0·76 nS; V½h, -75·7 mV; Sh, 7·3 mV; Gkx, 0·18 nS; V½kx, -58·9 mV; Skx, 8·8 mV; GL, 0·068 nS; VL, -20 mV. The data in CsCl were fitted by setting Gh to 0 nS; the best-fit values for the other parameters are: Gkx, 0·19 nS; V½kx, -48·3 mV; Skx, 5·6 mV; GL, 0·102 nS; VL, -20 mV. T = 23 °C. | ||
Responses recorded in the voltage-clamp mode before, during and after 3 mM CsCl, are illustrated in Fig. 2B. The inward rectification activated by stepping to -80 mV was fully blocked by 3 mM CsCl, and the block was reversible upon removal of CsCl. In contrast, the inward rectification activated by voltage steps to -50 mV appears to be Cs+ independent. Note that during perfusion with CsCl, the leakage increased slightly and consequently the response to -50 mV in CsCl appears larger than the control recordings. The selective block of inward relaxation by CsCl is illustrated in Fig. 2C, where control (open symbols) and CsCl (filled symbols) data are compared. In three cells, analysis by eqn (1) of the I-V data in control conditions and in the presence of 3 mM CsCl indicates that CsCl does not significantly change the parameters describing the gating of the current (Ikx) responsible for most of the inward rectification measured at -50 mV, as shown in Table 1.
Table 1. Ih and Ikx activation parameters from eqn (1) in control conditions and in the presence of 3 mM CsCl
| V½ (mV) | G (nS) | S (mV) | ||||
| Cell | Control | 3 mM CsCl | Control | 3 mM CsCl | Control | 3 mM CsCl |
| Ih | ||||||
| 1 | -66 | - | 1·23 | - | 6·8 | - |
| 2 | -74 | - | 1·14 | - | 7·8 | - |
| 3 | -76 | - | 0·76 | - | 7·3 | - |
| Mean | -72·0 ± 3·1 | - | 1·04 ± 0·14 | - | 7·3 ± 0·29 | - |
| Ikx | ||||||
| 1 | -49·5 | -59·6 | 0·24 | 0·25 | 3·0 | 3·1 |
| 2 | -43·2 | -55·6 | 0·16 | 0·28 | 8·8 | 10·0 |
| 3 | -58·9 | -53·2 | 0·18 | 0·21 | 7·4 | 7·2 |
| Mean | -50·5 ± 4·6 | -56·1 ± 1·9 | 0·19 ± 0·02 | 0·25 ± 0·02 | 6·4 ± 1·8 | 6·8 ± 2·0 |
The dashed line in Fig. 2C is a best fit to measurements in 3 mM CsCl for voltages negative to -70 mV, and crosses the voltage axis at -20 mV, suggesting, in agreement with eqn (1), the presence of a linear component of the membrane current reversing close to -20 mV. This current is abolished by omitting Na+ from the recording pipette (data not shown), and probably represents the combined contribution of the seal resistance and the electrogenic current generated by the sodium pump (see Appendix).
The partial suppression of the inward rectification by Cs+ at -50 mV may indicate that either an additional, Cs+-insensitive ionic conductance is present, or that the partial block of the inward rectification is a result of the voltage dependence of the CsCl blockade.
In order to investigate these possibilities, additional blockers of voltage-dependent conductances, such as TEA and Ba2+, were tested. As shown in Fig. 3, the inward rectification recorded at -50 mV was only marginally affected by 10 mM TEA, a well-known blocker of a variety of potassium-selective channels. TEA did not affect the inward rectification for voltage steps to -90 or -70 mV, but partially blocked the outward current elicited by voltage steps to +50 mV, and induced a membrane depolarization from -41 to -32 mV. The small difference between the measurements in controls and TEA, for voltages negative to -40 mV, shown in Fig. 3B, can be accounted for by an increase in the leakage conductance (see legend). In four cells, TEA (10 mM) reduced Gkx from 0·21 ± 0·04 to 0·10 ± 0·02 nS (P < 0·05, Wilcoxon rank test for pairs), with no effect on V½kx, Skx or Ih activation parameters. Similar results were obtained in different cells with higher TEA concentrations, up to 30 mM (data not shown).
 |
View larger version [in this window] [in a new window] |
|
|
A, the responses to 2 s voltage steps to -90, -70, -50 and +50 mV from a holding voltage of -35 mV are shown before (thin traces) and during (thick traces) exposure to 10 mM TEA chloride. Two points have been removed from the capacity transients. B, | ||
Note that TEA affects a noisy component of the membrane current, suggesting that it blocks some large-conductance ionic channels. In a few cases, when accidental dislodging of the pipette resulted in a perforated vesicle, it was possible to record the gating of single channels of large conductance, whose amplitude increased with membrane depolarizations above -20 mV (data not shown). Similar TEA-sensitive currents and channels have previously been reported in amphibian rods (Bader, et al. 1982; Rispoli et al. 1995).
The limited effect of 10 mM TEA suggests that ionic channels insensitive to Cs+ and with a low affinity for TEA contribute to the inward rectification observed at -50 mV. A Cs+- and TEA-insensitive current, whose voltage dependence is shifted by Ba2+ to more depolarized potentials has been described in salamander rods as Ikx (Beech & Barnes, 1989).
The data in Fig. 4 show the block of the inward rectification at -50 mV by a modified saline containing 1 mM BaCl2, 20 mM TEA and 6 mM KCl (TEA-Ba2+-K+ solution, see figure legend; Fig. 4A) or by a saline containing 1 mM BaCl2, 20 mM TEA but normal KCl (TEA-Ba2+ solution, see figure legend; Fig. 4B). Traces have been aligned for clarity, to compensate for the reduction in the holding current.
 |
View larger version [in this window] [in a new window] |
|
|
A, response to voltage steps to -70 and -50 mV, from holding values of -30 mV, in Locke's solution (thin traces) and in the presence of saline containing (mM): 120 NaCl, 20 TEA, 1 BaCl2, 6 KCl, 1·2 CaCl2, 5 Hepes, pH 7·4 (TEA-Ba2+-K+ saline) (thick traces). B, response to voltage steps to -70 and -50 mV, from holding values of -40 mV, in Locke's solution (thin traces) and in the presence of saline containing (mM): 120 NaCl, 20 TEA, 1 BaCl2, 3·6 KCl, 1·2 CaCl2, 5 Hepes, pH 7·4 (TEA-Ba2+ saline) (thick traces). The recordings in A and B have been aligned for clarity; the holding current was 9·6 pA in controls and -6·2 pA in the presence of TEA-Ba2+-K+ in A and 7·6 pA in controls and -0·1 pA in TEA-Ba2+ in B. Two points have been removed from the capacity transients in A and B. The vertical calibration bar is 20 pA for A and 15 pA for B. C and D, current amplitudes, averaged over the last 10 ms of the 2 s voltage steps (thick dots in A and B), are plotted for controls ( | ||
The complete I-V curves from Fig. 4A and B are shown in Fig. 4C and D, respectively. Open symbols in Fig. 4C and D are controls, while filled symbols are data in TEA-Ba2+-K+ solution (C) and in TEA-Ba2+ solution (D). Best fits to data based on eqn (1) suggest that both TEA-Ba2+-K+ and TEA-Ba2+ solutions halved Gkx and depolarized V½kx by about 40 mV. The increase in the inward rectification for voltages negative to -60 mV in TEA-Ba2+-K+ saline is expected on the basis of the known sensitivity of Ih to extracellular K+ (Bader & Bertrand, 1984). The reversibility of TEA-Ba2+-K+ saline was not systematically investigated, but in three cells the I-V relation reverted back to control values upon washing out with Locke's solution. The average values for the parameters describing the I-V relationship, estimated in seven rods by eqn (1), in control conditions and in the presence of TEA-Ba2+-K+ solution, are reported in Table 2A. Moreover, control experiments with TEA-Ba2+ saline indicate that activation of Ikx is shifted to more depolarized potentials independently of the external K+ concentration, as shown in Fig. 4B and D and as reported in Table 2B for the averaged values from three experiments.
Table 2. Ih and Ikx activation parameters from eqn (1) in control conditions and in the presence of TEA Ba2+-K+ or TEA-Ba2+ saline
| V½ (mV) | G (nS) | S (mV) | ||||
| Ih | Ikx | Ih | Ikx | Ih | Ikx | |
| A | ||||||
| Control | -66·1 ± 2·8 | -46·4 ± 3·8 | 1·09 ± 0·06 | 0·18 ± 0·02 | 8·0 ± 0·5 | 5·6 ± 0·3 |
| TEA-Ba+-K+ | -67·9 ± 3·2 | -4·7 ± 11·6 *** | 2·29 ± 0·39 *** | 0·08 ± 0·03 ** | 7·8 ± 0·6 | 7·6 ± 2·2 |
| B | ||||||
| Control | -74·8 ± 3·1 | -49·5 ± 4·1 | 0·98 ± 0·08 | 0·21 ± 0·03 | 7·3 ± 1·1 | 5·8 ± 0·5 |
| TEA-Ba2+ | -72·0 ± 2·1 | -5·6 ± 9·9 *** | 1·04 ± 0·14 | 0·10 ± 0·04 ** | 7·3 ± 0·3 | 6·1 ± 0·4 |
These data suggest that most of the inward rectification observed at -50 mV in the presence of Cs+ is a consequence of Ikx deactivation, with only a minimal contribution from the unblocked Ih. The current measured in BaCl2, reversing close to -20 mV, might result from the presence of a linear component of the membrane current reversing at about -20 mV (see also Fig. 2C and Fig. 4C and D).
Ionic selectivity of Ih in guinea-pig rods
One distinctive feature of Ih is its high permeability to both Na+ and K+. The ionic selectivity of the inward current active at -90 mV can be estimated from the instantaneous I-V relationship, as shown in Fig. 5 for a rod bathed in TEA-Ba2+-K+ saline. The recordings in Fig. 5A were leak subtracted on-line by a P/5 protocol, and the instantaneous I-V relationship from the recordings in Fig. 5A is plotted in Fig. 5B (open symbols).
 |
View larger version [in this window] [in a new window] |
|
|
A, the voltage was stepped for 2 s to -90 mV, before stepping back to voltages ranging from -85 to -15 mV, in 10 mV steps. Holding voltage was -30 mV. Traces have been leak-subtracted on-line by a P/5 protocol. The dotted line is the zero current level. B, | ||
The smooth curve is the best fit of eqn (2) to data, with a PNa/PK of 0·331 (see legend). The average PNa/PK and reversal potential in TEA-Ba2+-K+ saline were 0·26 ± 0·21 and -33·4 ± 2·8 mV, respectively (n = 3). Similar estimates were obtained in Locke's solution after off-line Ikx subtraction, with -31·0 ± 1·2 mV and 0·29 ± 0·1 for average reversal potential and PNa/PK, respectively (n = 3). While the small differences between reversal potential and PNa/PK ratio observed in Locke's solution and in TEA-Ba2+-K+ saline were not significant (Mann-Whitney U test, P > 0·2), the increase in PNa/PK with an external potassium concentration higher than Locke's solution has previously been reported in salamander rods (Hestrin, 1987).
Functional roles of Ih and Ikx in guinea-pig rods
An effective way of investigating the role of Ih and Ikx in rods is that of studying the effect of current stimuli modulated sinusoidally in time. Figure 6 illustrates the responses of a rod to both step and sinusoidal current injection. Figure 6A shows the responses to 2 s current steps ranging in amplitude from +2 to -16 pA for a rod whose membrane potential was depolarized from -65·5 to -39·9 mV by injecting a steady current of +17 pA. The response to the -4 pA step had a peak amplitude of 9·8 mV and relaxed by 3·45 mV with a time constant of 0·383 s. The response to the -16 pA step had a peak amplitude of 33 mV and relaxed by 12·4 mV with a time constant of 0·204 s. The open symbols in Fig. 6B are current amplitudes measured at a steady state for voltage-clamp steps ranging from -110 up to -20 mV, and the corresponding current traces are shown in the inset. Note that in this particular cell, Ih and Ikx are activated at rather negative voltages, as shown by the absence of a clear sag in the voltage response to ±2 pA current steps and by estimates provided by fitting the steady-state current values in Fig. 6B with eqn (1)
 |
View larger version [in this window] [in a new window] |
|
|
A, current-clamp responses to 2 s inward current steps ranging from +2 to -16 pA, in 2 pA steps. The cell was depolarized from -65 to -39·9 mV by a steady-current injection of 17 pA. B, | ||
Traces in Fig. 6C are voltages measured in current-clamp conditions in response to the sinusoidal currents in Fig. 6D. The thick trace in Fig. 6C is the voltage response measured when the current stimulus was applied from a membrane potential close to -65 mV, while the thin trace illustrates the voltage response recorded when the stimulus was applied from -52 mV. The dashed vertical lines in Fig. 6C and D are drawn through the peaks of the current stimuli, and clearly show that the peak of the voltage response recorded at a starting potential of -65 mV (thick trace in Fig. 6C) precedes that of the stimulus (thick trace in Fig. 6D). However, the peak advance was reduced (thin trace in Fig. 6C) when the current stimulus (thin trace in Fig. 6D) was applied to the cell from a starting potential of -52 mV. Note that when the stimulus was applied to the cell at -65 mV, the voltage response deviated from a simple sinusoid, with the slope of the depolarizing phase being 106·5 mV s-1, which is steeper than that of -96·5 mV s-1 of the hyperpolarizing phase. On the other hand, when the stimulus was applied from a membrane potential of -52 mV the rate of depolarization and hyperpolarization were similar, being 61·2 and -63·2 mV s-1, respectively.
Differences in voltage amplitude measured in response to similar stimulus intensities are also apparent, with a rod impedance of 4·2 and 2·7 G at -65 mV and -52 mV, respectively. The dependence of impedance on voltage might indicate that Ih and Ikx gating endows rods with inductive-like properties, generating a resonant peak in the impedance-frequency relationship.
To test the hypothesis that voltage-dependent currents contribute to bandpass amplification properties of rods, we measured the relationship between impedance and stimulation frequency, as shown in Fig. 7.
 |
View larger version [in this window] [in a new window] |
|
|
A, current-clamp responses to 2 s inward current steps ranging from | ||
Voltage responses of a rod to current steps ranging from 0 to -12 pA (step increment 2 pA) are plotted in Fig. 7A. In this cell, voltage-dependent currents are activated in the range of potentials between -40 and -50 mV, as indicated by the presence of a clear sag in the voltage response to -2 pA. Responses to stimulation frequencies of 0·4 (thick trace) and 4·1 Hz (thin trace) are illustrated in Fig. 7C. Note that voltage response amplitudes are different at 0·4 and 4·1 Hz, despite the similar stimulus amplitudes at 4·1 Hz (thin trace in Fig. 7D) and 0·4 Hz. The ratio between rod impedance at 4·1 and 0·4 Hz is about 1·13, suggesting that in fact rod impedance is frequency dependent. The bandpass properties of rods are better illustrated in Fig. 7B, where open circles plot the ratios between the impedance measured at different stimulating frequencies and that measured at 0·2 Hz in the same rod. A clear dependence of rod impedance on the stimulation frequency is apparent, with a peak of 1·6 at 1·9 Hz. In three rods, in which the relationship between impedance and stimulation frequency was investigated using 6-11 different stimulation frequencies, the normalized impedance peak was 1·5 ± 0·2 (n = 3). Membrane hyperpolarization by up to 10 mV increased the gain to 1·7 ± 0·3 (n = 3) and shifted the peak from 1·9 to 2·3 Hz. The continuous line plots the modulus of the complex impedance, computed in accordance with eqn (3) and normalized at 0·2 Hz, for the circuit in Fig. 9 (see legend in Fig. 9). The parameters were obtained from the response to the -4 pA current step shown in Fig. 7A, using the procedure described in Baylor et al. (1984a) for their 'economy model' (see also Methods).
Note that the voltage at 0·4 Hz (thick trace in Fig. 7C) leads the current stimulus (thick trace in Fig. 7D), while at 4·1 Hz, the voltage (thin trace in Fig. 7C) lags behind the current (thin trace in Fig. 7D). The starting voltage cannot account for this difference, because the 0·4 Hz stimulus (thick trace), if anything, was applied from a slightly more depolarized potential than the 4 Hz stimulus (thin trace). The experimental phase shifts (see eqn (5) in Methods) are plotted as filled circles in Fig. 7B. The continuous line through the filled circles is the theoretical phase computed for the circuit of Fig. 9 in accordance with eqn (4), with the same values for L, Cm, R1 and R2 used to compute the normalized impedance. The dotted line has been drawn for reference at zero phase shift.
The role of Ih and Ikx in the bandpass amplification of the voltage response was further investigated by using CsCl-containing saline or TEA-Ba2+ saline, as shown in Fig. 8.
 |
View larger version [in this window] [in a new window] |
|
|
A, current-clamp responses in control Locke's solution (thin trace) and in the presence of 3 mM CsCl (thick trace), to sinusoidal current stimuli. B, sinusoidal current stimuli frequency was 0·26 Hz. In A and B, traces are the average of 16 stimuli. Individual sweeps were sampled at 67 Hz and filtered at 20 Hz. T = 22 °C. C, current-clamp responses recorded in the presence of TEA-Ba2+ saline at two different starting voltages of -70·6 mV (thin trace) and -98·2 mV (thick trace). The arrow points to the harmonic distortion in the response recorded at the more hyperpolarized potential. D, sinusoidal current stimuli frequency was 0·23 Hz. In A and B, traces are the average of 16 stimuli. Individual sweeps were sampled at 31 Hz and filtered at 10 Hz. T = 21 °C. | ||
In Fig. 8A, 3 mM CsCl reduced the amplitude of the voltage response, despite the slight hyperpolarization, with a reduction of rod impedance from 2·12 to 1·95 G. Moreover, CsCl reduced the phase lead from 280 ms in control conditions to 220 ms. Qualitatively similar results were observed in a second rod. These observations suggest that Ih contributes to both gain and phase lead in the voltage response of rods addition, Ih gating may be responsible for the harmonic distortion observed upon membrane hyperpolarization.
The idea that the harmonic distortion is associated with Ih gating is also supported by the records illustrated in Fig. 8B, from a rod in TEA-Ba2+ solution. In this rod, in which phase lags were recorded at every stimulus frequency between 0·2 and 30 Hz, inward current injections bringing the membrane potential into the Ih activation range (V½h of -89 mV in this particular rod) induced a marked harmonic distortion, as indicated by the leftward-pointing arrow in Fig. 8C. This observation suggests that Ih gating is responsible for the harmonic distortion observed in the voltage response of rods, independently of Ikx gating.
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The aim of the present work was to investigate the properties and the functional roles of ionic currents activated by membrane hyperpolarization in retinal rods from the guinea-pig.
We will now discuss the evidence for the gating of two distinct voltage-dependent currents by membrane hyperpolarization, before discussing the functional roles of these currents in rods.
Voltage response and inward rectification in rods: contribution of Ih and Ikx
The present results show that guinea-pig rods respond to inward current injection with a membrane hyperpolarization characterized by a transient peak followed by relaxation to a plateau. Voltage-clamp recordings in the same rods, as shown in Figs 1, 2 and 6, suggest that an inward relaxation of the membrane current causes the sag from peak to plateau in the voltage responses.
CsCl abolishes this inward relaxation in the range of membrane potentials between -110 and -70 mV, and blocks the initial transient depolarization of the voltage response. In contrast, CsCl fails to block the inward relaxation and the peak to plateau transition for voltages less negative than -60 mV. BaCl2 blocks the inward rectification gated by membrane hyperpolarizations ranging from -40 to -60 mV, suggesting that two distinct voltage-dependent conductances are present in guinea-pig rods.
The poor discrimination between sodium and potassium and the sensitivity to external potassium (see Figs 5 and 4, respectively), along with the steep voltage-dependent activation with membrane hyperpolarization, suggest that the CsCl-sensitive conductance is akin to the current (Ih) described in other preparations (Bader et al. 1982; Bader & Bertrand, 1984; Pape, 1996 for review). The Cs+-resistant current blocked by Ba2+, with only a moderate sensitivity to TEA, corresponds to Ikx, an M-like current deactivated by membrane hyperpolarization (Beech & Barnes, 1989).
According to our results, for rods exposed to saturating lights, the voltage at which the sizes of opposing currents (Ih, depolarizing inward current and Ikx, hyperpolarizing outward current) match is the steady rod membrane potential in bright light. At this potential, Ih and Ikx are the only currents present, the linear component of the membrane current probably representing leakage through the cell-glass giga-seal (see Appendix).
Ih, Ikx and bandpass amplification in rods
As originally reported by Cole & Baker (1941), non-linearity of membrane currents due to their time and voltage dependence may provide cells with inductive-like properties. In rods from amphibians, gating of voltage-dependent currents has previously been reported to endow the rod network with bandpass properties (Detwiler et al. 1978, 1980). Furthermore, it has been reported that a Cs+-insensitive potassium conductance is responsible for the acceleration of the response to dim light stimuli and for the inductive properties of the rod network (Owen & Torre, 1983; Torre & Owen, 1983) in amphibian rods. This conductance, later identified as Ikx (Beech & Barnes, 1989), may therefore play a similar role in guinea-pig rods, leading to bandpass amplification of the voltage response.
Data in Fig. 7B provide direct support for the hypothesis that mammalian rods closely approximate the bandpass filtering with an amplification expected theoretically for the circuit of Fig. 9. In this model, a resistance (R2) is in series with an inductance (L) and is in parallel with a purely resistive element (R1) and a capacitor (Cm).
 |
View larger version [in this window] [in a new window] |
|
|
OS, IS and ST indicate the outer segment, the inner segment and the synaptic terminal, respectively. I(t) denotes that the outer segment generated current changes with time, but is independent of voltage (constant current generator). R1, R2, L and Cm are the elements of the LRC circuit of the inner segment. Vm is the presynaptic voltage, originating from the filtering by the LRC circuit at the inner segment of the current generated at the outer segment level. The arrow at the bottom indicates the flow of information, spreading from the outer segment to the synaptic terminal via the inner segment. | ||
Figure 7B shows that the agreement between experimental impedances (open circles) and the theoretical curve (see eqn (3)) is satisfactory for current stimuli applied close to -50 mV. Furthermore, analysis of the phase of the voltage response, as shown in Fig. 7B, provides further support for the idea that rods are electrically equivalent to the circuit of Fig. 9. Indeed, the theoretical phase of the circuit of Fig. 9, defined by eqn (4), correctly describes experimental phase leads and lags for this rod.
It should be noted that current stimuli of 4-8 pA, which induce responses over a range of membrane potentials from -55 to -30 mV, mainly gate Ikx. Therefore, in agreement with the data of Owen & Torre (1983), the potassium current carried by Ikx is the main conductance controlling the inductive properties of rods in response to small current stimuli, such as those generated by dim light illumination. These data suggest that when current stimuli mainly modulate Ikx, rods behave like bandpass amplifiers, in accordance with the model presented in Fig. 9.
However, when current stimuli are imposed from a membrane potential close to that in bright light, where Ih and Ikx have a similar size, a further increase in membrane impedance is observed. Moreover, the voltage waveform becomes distorted, with the rate of depolarization being higher than that of hyperpolarization, as shown in Fig. 6C. In this rod, the impedance measured for stimuli applied from a voltage close to the potential in bright light (-65 mV) is 4·2 G, about 1·6 times that of 2·7 G at -52 mV. The increase in rod impedance with membrane hyperpolarization was observed in three rods, with an average increase in the peak value of the normalized impedance from 1·5 to 1·7. Moreover, an increase in the phase lead of voltage was observed for potentials close to those in bright light, an effect associated with a harmonic distortion.
Analysis of voltage-clamp recordings in Fig. 6B by eqn (1) indicates that when this particular cell is depolarized to -52 mV, Ih is no longer active, while Ikx is almost fully activated. On the other hand, at -65 mV, both Ih and Ikx are active and of similar size, suggesting that both currents contribute to the inductive behaviour of the cell. Therefore, both Ih and Ikx may contribute to bandpass amplification, but the gain increases when Ih is further activated. This observation suggests a specific role for Ih in accelerating the rate of voltage recovery after bright, saturating lights.
This idea is in general agreement with the data of Fig. 2A, showing that 3 mM CsCl slows the kinetics of voltage recovery at the end of -15 pA current stimulus bringing the membrane potential into the Ih activation range. Note that CsCl hardly affects the rate of voltage recovery when the stimulus intensity is -5 pA, which does deactivate Ikx but fails to appreciably activate Ih. This suggests that, at the end of a large hyperpolarization, when Ikx is fully deactivated, the standing inward Ih will accelerate the recovery of the membrane potential to its dark level.
Further evidence for the role of Ih is provided by the data illustrated in Fig. 8A, which suggest that 3 mM CsCl reduces the harmonic distortion as well as the rod impedance. Moreover, the data in Fig. 8B show that the harmonic distortion may occur in rods upon membrane hyperpolarization that activates Ih independently of Ikx gating.
Bandpass amplification and synaptic transmission in rods: role of Ih
The role of Ih has been considered to be of minor importance in the voltage response of rods. The main reason is that the transient peak generated by Ih activation in response to bright light stimuli is clipped by the rod synapse (Attwell et al. 1987; Witkovsky et al. 1997). In addition, no bandpass amplification has previously been reported for signals transmitted through the rod-horizontal and rod-hyperpolarizing bipolar synapses (Belgum & Copenhagen, 1988), which has been considered to be an indication of no filtering at the first stage of synaptic transmission in the retina. In the above work, however, the contribution of the current to voltage conversion was not investigated. Intracellular recordings from cat horizontal cells in response to flickering lights suggest the presence of bandpass amplification of the first harmonic of the response, peaking at about 3-5 Hz (Lankheet et al. 1991). Moreover, analysis of the first harmonic of the aminophosphobutyrate (APB)-sensitive component of the cat electroretinogram (ERG), which mainly reflects synaptic transmission from rods to depolarizing bipolar cells, indicates the presence of bandpass amplification with a peak in the response at about 3-5 Hz (Gargini et al. 1999). In this respect, it is intriguing that both CsCl and zatebradine, an organic blocker of hyperpolarization-gated currents (Van Bogaert et al. 1990; Goethals et al. 1993; DiFrancesco, 1994), attenuate the gain of the rod-depolarizing bipolar synapse and delay both the onset and the recovery of the ERG response to steps of light (Gargini et al. 1999). At variance with the data from paired recordings in amphibians, these effects of the blockers of hyperpolarization-activated currents suggest a role for Ih in the filtering of signals transmitted from rods to second-order neurones.
It is possible that the different conclusions reached by different groups reflect differences in signal processing between the retina of the cat and those of cold-blooded vertebrates. However, this possibility seems unlikely considering that rods and second-order neurones in the vertebrate retina appear to share a common design in terms of membrane properties and cellular connectivity. The possibility that the high-pass filtering reported in the cat may result from the spreading of small voltage signals across the rod network, as originally reported in the amphibian retina (Attwell et al. 1987), is unlikely considering that large field stimuli were used in the studies done in the cat retina.
An alternative explanation, which is in general agreement with our results, is presented in schematic form in Fig. 9. In the scheme, the outer segment works like a constant-current generator to inject current into the LRC circuit, which is localized at the inner segment level. After high-pass filtering and amplification by the LRC circuit, the resulting voltage spreads to the synaptic terminal, where it controls transmission from rod to second-order neurones.
The idea of the outer segment being a constant current generator is rooted in the work of Bader et al. (1979), who showed that the amplitude of the light-sensitive current is approximately constant between -20 and -80 mV. According to our data, inner segment conductances are modelled as an LRC circuit and confer bandpass amplification to the voltage generated in response to outer-segment current injection. After this stage of bandpass amplification, synaptic transmission from rods to second-order neurones is modelled by a stage of low-pass filtering (Belgum & Copenhagen, 1988).
Activation of Ih may affect synaptic transmission by letting the voltage, which controls transmitter release by the rod synapse, recover to its dark level in advance of the dark current. The proposed role for Ih is in agreement with data from salamander rods, where simultaneous intracellular recordings of membrane voltage and light-sensitive current measurements were obtained by the suction pipette technique (Baylor et al. 1984a). These experiments showed that voltage recovered faster than dark current after bright illumination, as expected from Ih activation (see Fig. 1 in Baylor et al. 1984a). In addition, the data illustrated in Figs 6 and 8 indicate that gating of Ih during repolarization from the membrane potential in bright light causes a phase advance so marked that it results in a harmonic distortion. It is intriguing that these phenomena of phase advance and harmonic distortion of the voltage recovery occur for stimulus frequencies of 0·3-0·9 Hz, which approximate the kinetics of dark-current recovery after bright illumination (Demontis et al. 1999).
It is important to note that, although the Ih-generated sag in the voltage response of rods may be clipped by the rod synapse (Attwell et al. 1987; Witkovsky et al. 1997), Ih will nevertheless cause the photovoltage to lead the dark-current recovery. Considering that the photovoltage controls the operation of the rod synapse, the phase advance of the photovoltage associated with Ih activation might improve the temporal fidelity of signals transmitted through the rod synapse.
It must be noted that the amplitude of the first harmonic of the response of both rod-horizontal and rod-depolarizing bipolar synapses in the cat peaks at 3-5 Hz. This value is somewhat higher than that of 1·9 Hz reported in the present study for the voltage response of guinea-pig rods. It is perhaps appropriate to note that in monkey rods at body temperature (Schneeweis & Schnapf, 1995), the relaxation from peak to plateau measured in response to bright flashes is faster ( around 30 ms) than in guinea-pig rods at room temperature ( 
200 ms). This faster kinetics at body temperature is accounted for by a Q10 higher than 3 (in the range 25-35°C), in agreement with that of 4·3 reported for Ih activation in bullfrog sympathetic neurones (Tokimasa & Akasu, 1990). Therefore, at body temperature Ih and Ikx may provide amplification of visual signals tuned at higher frequencies than those found in the present work at room temperature.
Origin of leakage in patch-clamp recordings from guinea-pig rods
Data shown in Figs 2C and 4C and D include a linear component of the membrane current. This linear component has a reversal potential close to -20 mV, suggesting either the presence of a single current with a low ionic selectivity or the lumped contribution from several currents.
The tendency of this linear conductance to increase in the course of prolonged recordings, its lack of sensitivity to known blockers of ionic channels and its dependence on sodium in the recording pipette, suggests that it is an artefact of the recording procedure. As detailed here, a linear current reversing at about -20 mV is expected to result from the current through the leakage resistance and the electrogenic current contributed by the Na+-K+ pump.
As a consequence of its 3 Na+ :2 K+ stoichiometry, the Na+-K+-ATPase operation generates an electrogenic current (Ipump), defined by:
Ipump(t) = {\123}-Kpump([Na+]i(t) - [Na+eq]i)/3}F
, (A1)
where F is Faraday's constant (23480 C M-1) and is the rod free volume (23 pl). Kpump, the rate constant of the pumping process at 22°C, is expected to be 0·413 s-1, a value adequate to match the dark current at this temperature (7 pA with about 80 % of the current carried by sodium and the remaining 20 % by calcium) (Robinson et al. 1993; Demontis et al. 1997). The internal sodium concentration attained at thermodynamic equilibrium ([Na+eq]i), when the free energy of ATP hydrolysis (GATP) balances the free energy of ion transport against the gradient (GNa + GK), will change with the voltage, as described by:
With [ATP] = 5 mM, [ADP] = 100 µM and [Pi] = 1·5 mM, then GATP = -58·1 kJ M-1. With Fcal = 96·5 kJ V-1 mol-1, Rcal = 8·28 × 10-3 kJ mol-1 K-1, T = 295·15 K and V = -50 mV, then [Na+eq]i = 1·017 mM.
In the presence of sodium influx, the sodium pump will generate an electrogenic current if [Na+]i > [Na+eq]i.
In bright light, the recording pipette will provide the cell with a constant sodium load proportional to the difference between the free sodium concentration in the pipette ([Na+]p) and [Na+]i. At pH 7·2, sodium will be partly bound to ATP in the pipette, and [Na+]p 7·5 mM.
Diffusion of ions from the pipette to the cell has been reported to conform to a single exponential process (Oliva et al. 1988). According to this notion, the rate constant for the diffusion (Kdiff) from the pipette to the cell is given by:
Kdiff = DNa
/Ra, (A3)
where DNa is the diffusion coefficient in water for sodium (1·33 × 10-5 cm2 s-1), is the resistivity of saline (60 cm); and Ra is the access resistance of the pipette (50 M). The rate of sodium diffusion (
Na,pip) from the pipette to the cell at a steady state is given by:
where [Na+]i(
The value of [Na]i(
Equation (A5) gives [Na+]i(
The estimates provided for the pump current are approximate, because no allowance has been made for its known voltage dependence (Gadsby et al. 1985; Gadsby & Nakao, 1989), apart from that intrinsic in [Na+eq]i. Moreover, the contribution of Ih to the sodium turnover of rods has been omitted, and consequently the estimated pump current will strictly apply to voltages where Ih is not active, or will approximate experiments performed in the presence of 3 mM CsCl. The results for several voltages ranging from -120 to 0 mV are plotted as open circles in Fig. 10, assuming Ih,Na = 0.
The open diamonds are the values for the seal current reversing at 0 mV. The filled circles are the sum of the seal current and the electrogenic sodium pump current. The dotted line through the filled circles represents a regression line with slope (GL) of 0·069 nS, to be compared with the GL values provided in Figs 2 and 4. The computed reversal potential is -21 mV.
We would like to thank W. G. Owen and D. DiFrancesco for critical reading and helpful comments on an earlier version of this paper. Funding for this study was provided by grants from the Ministero Universita' e Ricerca Scientifica e Tecnologica (40 % and 60 %) and the European Union to L.C. and by a grant from the Consiglio Nazionale della Ricerca to G.C.D.
Corresponding author
G. C. Demontis: Dipartimento di Psichiatria, Neurobiologia, Farmacologia e Biotecnologie, Universita' di Pisa, Via Bonanno Pisano, 6 I-56126 Pisa, Italy.
Email: demontis{at}farm.unipi.it
This article has been cited by other articles:
Na,pip = Kdiff([Na+]i(
View larger version
[in this window]
[in a new window]
Figure 10. Computation of conductance and reversal potential of leakage from the seal resistance and the estimated Na+-K+ electrogenic current
, the pump current computed from eqn (A1); 
, the current through the seal resistance of 15 G; 
the net current resulting from the sum of the seal and the pump current. The straight dotted line through the filled circles is a regression line with a slope of 69 pS and a reversal potential of -21 mV, to be compared with the estimates for GL provided by eqn (1) for the data in Figs 2 and 4. The vertical dotted line has been drawn for reference at -20 mV.
REFERENCES
Top
Abstract
Introduction
Methods
Results
Discussion
References
Acknowledgements
Attwell, D., Borges, S., Wu, S. M. & Wilson, M. (1987). Signal clipping by the rod output synapse. Nature 328, 522-524
[Medline]
Bader, C. R. & Bertrand, D. (1984). Effect of changes in intra- and extracellular sodium on the inward (anomalous) rectification in salamander photoreceptors. The Journal of Physiology 347, 611-631
[Abstract]
Bader, C. R., Bertrand, D. & Schwartz, E. A. (1982). Voltage-activated and calcium-activated currents studied in solitary rod inner segments from the salamander retina. The Journal of Physiology 331, 253-284
[Medline]
Bader, C. R., MacLeish, P. R. & Schwartz, E. A. (1979). A voltage-clamp study of the light response in solitary rods of the tiger salamander. The Journal of Physiology 296, 1-26
[Medline]
Barnes, S. & Hille, B. (1989). Ionic channels of the inner segment of tiger salamander cone photoreceptors. Journal of General Physiology 94, 719-743
[Abstract]
Baylor, D. A., Lamb, T. D. & Yau, K. W. (1979). The membrane current of single rod outer segments. The Journal of Physiology 288, 589-611
[Abstract]
Baylor, D. A., Matthews, G. & Nunn, B. J. (1984a). Location and function of voltage-sensitive conductances in retinal rods of the salamander Ambystoma tigrinum. The Journal of Physiology 354, 203-223
[Abstract]
Baylor, D. A., Nunn, B. J. & Schnapf, J. L. (1984b). The photocurrent, noise and spectral sensitivity of rods of the monkey Macaca fascicularis. The Journal of Physiology 357, 575-607
[Abstract]
Beech, D. J. & Barnes, S. (1989). Characterization of a voltage-gated K channel that accelerates the rod response to dim light. Neuron 3, 573-581
[Medline]
Belgum, J. H. & Copenhagen, D. R. (1988). Synaptic transfer of rod signals to horizontal and bipolar cells in the retina of the toad (Bufo marinus). The Journal of Physiology 3, 225-245.
Cole, K. S. & Baker, R. F. (1941). Longitudinal impedance of the squid giant axon. Journal of General Physiology 24, 771-788.
Demontis, G. C. & Cervetto, L. (1995). Hyperpolarization-activated current (Ih) and its modulation in mammalian retinal rods. Biophysical Journal 68, A23 (Abstract).
Demontis, G. C. & Cervetto, L. (1996). Control of hyperpolarization-activated current (Ih) by dopamine and cyclic nucleotides in mammalian retinal rods. The Journal of Physiology 491.P, 12S.
Demontis, G. C., Longoni, B., Barcaro, U. & Cervetto, L. (1999). The transfer function of guinea-pig rods: role of the hyperpolarization-activated current Ih. Biophysical Journal (in the Press).
Demontis, G. C., Longoni, B., Gargini, C. & Cervetto, L. (1997). The energetic cost of phototransduction in retinal rods of mammals. Archives Italiennes Biologie 135, 95-109.
Demontis, G. C., Ratto, G. M., Bisti, S. & Cervetto, L. (1995). Effect of blocking the Na/K ATPase on Ca extrusion and light adaptation in mammalian retinal rods. Biophysical Journal 69, 439-450
[Abstract]
Detwiler, P. B., Hodgkin, A. L. & McNaughton, P. A. (1978). A surprising property of electrical spread in the network of rods in the turtle's retina. Nature 274, 562-565
[Medline]
Detwiler, P. B. A., Hodgkin, A. L. & McNaughton, P. A. (1980). Temporal and spatial characteristics of the voltage response of rods in the retina of the snapping turtle. The Journal of Physiology 300, 213-250
[Medline]
DiFrancesco, D. (1994). Some properties of the UL-FS49 block of the hyperpolarization-activated (IF) current in SA node myocytes. Pflügers Archiv 427, 64-70
[Medline]
Fain, G. L., Quandt, F. N., Bastian, B. L. & Gerschenfeld, H. M. (1978). Contribution of caesium-sensitive conductance increase to rod response. Nature 272, 467-469.
Gadsby, D. C, Kimura, J. & Noma, A. (1985). Voltage dependence of Na/K pump current in isolated heart cells. Nature 315, 63-65
[Medline]
Gadsby, D. C. & Nakao, M. (1989). Steady-state current-voltage relationship of the Na/K pump in guinea pig ventricular myocytes. Journal of General Physiology 94, 511-537
[Abstract]
Gargini, C., Demontis, G. C., Bisti, S. & Cervetto L. (1999). Effects of blocking the hyperpolarization-activated current (Ih) on the cat electroretinogram. Vision Research (in the Press).
Goethals, M., Raes, A. & Van Bogaert, P. (1993). Use-dependence block of the pacemaker current IF in rabbit sinoatrial node cells by zatebradine (UL-FS49). On the mode of action of sinus node inhibitors. Circulation 88, 2389-2401
[Abstract]
Hestrin, S. (1987). The properties and function of inward rectification in rod photoreceptors of the tiger salamander. The Journal of Physiology 390, 319-333
[Abstract]
Horn, R. & Marty, A. (1988). Muscarinic activation of ionic currents measured by a new whole-cell recording method. Journal of General Physiology 92, 145-159
[Abstract]
Lamb, T. D. & Pugh, E. N. (1992). A quantitative account of the activation steps involved in phototransduction in amphibian photoreceptors. The Journal of Physiology 449, 719-757
[Abstract]
Lankheet, M. J. M., van Wezel R. J. A. & van de Grind, W. A. (1991). Light adaptation and frequency transfer properties of cat horizontal cells. Vision Research 31, 1129-1142
[Medline]
Neher, E. (1992). Correction for liquid junction potentials in patch clamp experiments. Methods in Enzymology 207, 123-131
[Medline]
Nikonov, S., Engheta, N. & Pugh, E. N. Jr (1998). Kinetics of recovery of the dark-adapted salamander rod photoresponse. Journal of General Physiology 111, 7-37
[Abstract/Full Text]
Oliva, C., Cohen, I. S. & Mathias, R. T. (1988). Calculation of time constant for intracellular diffusion in whole cell patch clamp configuration. Biophysical Journal 54, 791-799
[Abstract]
Owen, G. W. & Torre, V. (1983). High-pass filtering of small signals by retinal rods. Ionic studies. Biophysical Journal 41, 325-339
[Abstract]
Pape, H. C. (1996). Queer current and pacemaker: the hyperpolarization-activated cation current in neurons. Annual Review of Physiology 58, 299-327
[Abstract]
Rispoli, G., Navangione, A. & Vellani, V. (1995). Transport of K+ by Na+-Ca2+, K+ exchanger in isolated rods of lizard retina. Biophysical Journal 69, 74-83
[Abstract]
Robinson, D. W., Ratto, G. M., Lagnado, L. & McNaughton, P. A. (1993). Temperature dependence of the light response in rat rods. The Journal of Physiology 462, 465-481
[Abstract]
Sather, W. A. & Detwiler, P. B. (1987). Intracellular biochemical manipulation of phototransduction in detached rod outer segment. Proceedings of the National Academy of Sciences of the USA 84, 9290-9294
[Medline]
Schneeweis, D. M. & Schnapf, J. L. (1995). Photovoltage of rods and cones in the macaque retina. Science 268, 1053-1056
[Medline]
Tokimasa, T. & Akasu, T. (1990). Cyclic AMP regulates an inwardly rectifying sodium-potassium current in dissociated bull-frog sympathetic neurones. The Journal of Physiology 420, 409-429
[Abstract]
Torre, V. & Owen, W. G. (1983). High-pass filtering of small signals by the rod network in the retina of the toad, Bufo marinus. Biophysical Journal 41, 305-324.Van Bogaert, P., Goethals M. & Simoens, C. (1990). Use- and frequency-dependent blockade by UL-FS49 of the pacemaker current in sheep Purkinje fibres. European Journal of Pharmacology 187, 241-256.
Van Bogaert, P., Goethals M. & Simoens, C. (1990). Use- and frequency-dependent blockade by UL-FS49 of the pacemaker current in sheep Purkinje fibres. European Journal of Pharmacology 187, 241-256
[Medline]
Witkovsky, P., Schmitz, Y., Akopian, A., Krizaj, D. & Tranchina D. (1997). Gain of rod to horizontal cell synaptic transfer: relation to glutamate release and a dihydropyridine-sensitive calcium current. Journal of Neuroscience 17, 7297-7306
[Abstract/Full Text]
Wollmuth, L. P. (1995). Multiple ion binding sites in Ih channels of rod photoreceptors in tiger salamanders. Pflügers Archiv 430, 34-43
[Medline]
Wollmuth, L. P. & Hille, B. (1992). Ionic selectivity of Ih channels of rod photoreceptors in tiger salamanders. Journal of General Physiology 100, 749-765
[Abstract]
Yagi, T. & McLeish, P. (1994). Ionic conductances of monkey solitary cone inner segments. Journal of Neurophysiology 71, 656-665
[Medline]
