Voltage-Dependent Na⫹ Currents in Mammalian
Retinal Cone Bipolar Cells
ZHUO-HUA PAN AND HUI-JUAN HU
Department of Anatomy and Cell Biology, Wayne State University School of Medicine, Detroit, Michigan 48201
Received 14 December 1999; accepted in final form 8 August 2000
and Raviola 1986; Greferath et al. 1990). On the other hand,
multiple subtypes of cone bipolar cells have been reported
(Euler and Wässle 1995; Famiglietti 1981; Kolb et al. 1981;
Pourcho and Goebel 1987). Most of the previous studies of
membrane currents in bipolar cells were performed in lower
vertebrates (Connaughton and Maguire 1998; Kaneko and
Tachibana 1985; Lasater 1988; Maguire et al. 1989; TessierLavigne et al. 1988). Studies of mammalian bipolar cells were
mainly limited to rod bipolar cells (Gillette and Dacheux 1995;
Karschin and Wässle 1990). The properties of voltage-activated
membrane currents in mammalian cone bipolar cells are less clear.
We previously reported the capability of distinguishing mammalian rod and cone bipolar cells after enzymatic dissociation
(Pan 2000). We characterized and compared voltage-activated
membrane channels between these two types of bipolar cells.
Here we report that a portion of cone bipolar cells in the rat retina
display voltage-dependent Na⫹ currents. A brief report of this
work has been presented in abstract form (Pan 1999).
METHODS
Retinal bipolar cells are second-order neurons that relay the
signal from photoreceptors to amacrine and ganglion cells in
the retina. It is commonly believed that bipolar cells are nonspiking neurons and that signals propagate in a passive manner
in these cells. However, the recent findings of spontaneous and
light-evoked Ca2⫹-spikes in Mb1 retinal bipolar cells in the
goldfish suggest that bipolar cells may be more excitable than
what has been previously thought (Burrone and Lagnado 1997;
Protti et al. 2000; Zenisek and Matthews 1998). Nevertheless
voltage-dependent Na⫹ currents have not been previously reported in retinal bipolar cells. However, an early study found
that a portion of mammalian bipolar cells displayed immunoreactivity to Na⫹ channel ␣ subunits (Miguel-Hidalgo et al.
1994).
Bipolar cells are classified into ON and OFF types based on
their response polarity to light (Kaneko 1970; Werblin and
Dowling 1969). Bipolar cells are also divided into rod and cone
bipolar cells based on their synaptic inputs. In mammals, only
a single type of rod bipolar cells, ON type, has been reported
(Boycott and Dowling 1969; Boycott and Kolb 1973; Dacheux
Bipolar cells were isolated from ⱖ4-wk-old Long Evans rats by
dissociation methods previously described (Pan 2000; Pan and Lipton
1995). In brief, animals were deeply anesthetized with CO2 and killed
by decapitation. Retinas were removed and placed in a Hanks’ solution (normal Hanks’) containing (in mM) 138 NaCl, 1 NaHCO3, 0.3
Na2HPO4, 5 KCl, 0.3 KH2PO4, 1.25 CaCl2, 0.5 MgSO4, 0.5 MgCl2,
5 HEPES, and 22.2 glucose, with phenol red, 0.001% vol/vol; adjusted to pH 7.2 with 0.3 N NaOH. The retinas were incubated for
⬃50 min at 34 –37°C in an enzyme solution that consisted of the
Hanks’ solution, supplemented with 0.2 mg/ml DL-cysteine, 0.2 mg/ml
bovine serum albumin, and 1.6 U/ml papain, adjusted to pH 7.2 with
0.3 N NaOH. Following several rinses in Hanks’ solution, the retinas
were mechanically dissociated by gently triturating with a glass pipette. The resulting cell suspension was plated onto culture dishes.
Cells were used for recordings within 5 h after dissociation.
Bipolar cells were identified based on their characteristic morphology (Karschin and Wässle 1990; Pan and Lipton 1995; Yeh et al.
1990). Identification of rod and cone bipolar cells has been previously
described (Pan 2000). In brief, rod bipolar cells had long and thick
axons and usually retained axon terminals with relative large synaptic
boutons. Their dendritic trees were thick and bush-like. Cone bipolar
cells, on the other hand, had sparser dendritic trees and thinner axons.
Synaptic boutons were small or absent.
Recordings with patch electrodes in the whole cell configuration
were made by standard procedures (Hamill et al. 1981) at room
Address for reprint requests: Z.-H. Pan, Dept. of Anatomy and Cell Biology,
Wayne State University School of Medicine, 540 E. Canfield Ave., Detroit, MI
48201 (E-mail: [email protected]).
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked ‘‘advertisement’’
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
INTRODUCTION
2564
0022-3077/00 $5.00 Copyright © 2000 The American Physiological Society
www.jn.physiology.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on June 17, 2017
Pan, Zhuo-Hua and Hui-Juan Hu. Voltage-dependent Na⫹ currents
in mammalian retinal cone bipolar cells. J Neurophysiol 84:
2564 –2571, 2000. Voltage-dependent Na⫹ channels are usually expressed in neurons that use spikes as a means of signal coding. Retinal
bipolar cells are commonly thought to be nonspiking neurons, a
category of neurons in the CNS that uses graded potential for signal
transmission. Here we report for the first time voltage-dependent Na⫹
currents in acutely isolated mammalian retinal bipolar cells with
whole cell patch-clamp recordings. Na⫹ currents were observed in
⬃45% of recorded cone bipolar cells but not in rod bipolar cells. Both
⫹
ON and OFF cone bipolar cells were found to express Na channels.
The Na⫹ currents were activated at membrane potentials around ⫺50
to ⫺40 mV and reached their peak around ⫺20 to 0 mV. The
half-maximal activation and steady-state inactivation potentials were
⫺24.7 and ⫺68.0 mV, respectively. The time course of recovery from
inactivation could be fitted by two time constants of 6.2 and 81 ms.
The amplitude of the Na⫹ currents ranged from a few to ⬎300 pA
with the current density in some cells close or comparable to that of
retinal third neurons. In current-clamp recordings, Na⫹-dependent
action potentials were evoked in Na⫹-current-bearing bipolar cells by
current injections. These findings raise the possibility that voltagedependent Na⫹ currents may play a role in bipolar cell function.
NA⫹ CURRENTS IN RETINAL BIPOLAR CELLS
RESULTS
Na⫹ currents in morphologically identified cone bipolar
cells
When cone bipolar cells were depolarized to ⫺50 mV or
more positively from the holding potential of ⫺70 or ⫺80 mV,
fast transient inward currents were frequently observed. An
example of a morphologically identified cone bipolar cell that
displayed such currents is shown in Fig. 1A.
The fast transient currents were still present in the recording
solution containing 4 mM Co2⫹ and without added Ca2⫹ (Fig. 1,
B and C). The currents were activated around ⫺50 to ⫺40 mV
and reached their peak around ⫺20 to 0 mV from the holding
potential of ⫺80 mV (Fig. 1B). The rapid activation and inactivation of the currents resemble the typical pattern of Na⫹ currents.
Furthermore the currents were reversibly blocked by TTX (0.5–1
␮M) as exemplified in Fig. 1C (n ⬎ 10). These results indicate
that the fast transient currents are Na⫹ currents.
GABAC receptor-mediated currents in Na⫹-current-bearing
bipolar cells
To further ensure that the recorded Na⫹-current-bearing
cells were bipolar cells, we examined GABAC receptor-mediated currents in these cells. This is because bipolar cells but not
FIG. 1. Properties of voltage-activated fast inward transient
currents in morphologically identified cone bipolar cells. A: an
example of fast transient inward currents evoked from a morphologically identified cone bipolar cell (left). Right: the current
traces shown were evoked by test pulses of ⫺60 to ⫺20 mV
from the holding potential of ⫺80 mV. Recordings were made
in normal Hanks’. B: current-voltage (I-V) relationships of the
fast transient inward currents. Current traces were evoked by
test pulses of ⫺60 to 20 mV at increments of 10 mV from the
holding potential of ⫺80 mV (top). The currents measured at
the peak were plotted vs. test potentials (bottom). C: the transient inward current was reversibly blocked by TTX (500 nM).
Current traces were evoked by a test pulse at ⫺10 mV from the
holding potential of ⫺70 mV. Recordings in B and C were
made in normal Hanks’ containing 4 mM Co2⫹. Recordings in
A–C were made from different cone bipolar cells, respectively.
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on June 17, 2017
temperature (20 –25°C) with an EPC-9 amplifier and PULSE software
(Heka Electronik, Lambrecht/Pfalz, Germany). Electrodes were
coated with silicone elastomer (Sylgard, Dow Corning, Midland, MI)
and fire-polished. The resistance of the electrode was 7–14 M⍀.
Series resistance ranged from 12 to 40 M⍀ and was not routinely
compensated. Cell capacitance was canceled and recorded by PULSE
software. In some recordings, leak currents were subtracted with an
on-line P/4 protocol provided by PULSE software. Most recordings
were made either in the normal Hanks’ solution described in the
preceding text with or without Co2⫹ (4 mM) or in a high-Ca2⫹ (10
mM) solution. Ca2⫹ was not added in normal Hanks’ when Co2⫹ was
used. The high Ca2⫹ solution contained (in mM) 95 NaCl, 5 KCl, 10
CsCl, 20 TEA-Cl, 1 MgCl2, 10 CaCl2, 5 HEPES, and 22.2 glucose,
with phenol red, 0.001% vol/vol; pH 7.2. The electrode solution
contained (in mM) 40 CsCl, 80 Cs-acetate, 20 TEA-Cl, 1 MgCl2, 0.5
CaCl2, 5 EGTA, 10 HEPES, 0.5 Na-GTP, and 2 Na-ATP, pH adjusted
with CsOH to 7.4. In some recordings, Cs-acetate was replaced by
CsCl. In the recordings of action potentials, the electrode solution
contained (in mM) 133 K-gluconate, 7 KCl, 1 MgCl2, 0.5 CaCl2, 5
EGTA, 10 HEPES, 0.5 Na-GTP, and 2 Na-ATP, pH adjusted with
KOH to 7.4. Liquid junction potentials were measured according to
the procedure described by Neher (1992) and corrected. Chemical
agents were applied by local perfusion by gravity-driven superfusion
pipettes placed ⬃200 –300 ␮m away from the cell being recorded.
Tetrodotoxin (TTX) was purchased from Research Biochemicals
(Natick, MA). All other chemicals were purchased from Sigma (St.
Louis, MO). Each result reported in this study was based on the observations obtained from at least five cells unless otherwise indicated.
2565
2566
Z.-H. PAN AND H.-J. HU
were blocked under our recording conditions. The third-order
neurons were identified by their multipolar morphology and,
usually, the presence of much larger voltage-activated Na⫹
currents.
Ca2⫹ currents in Na⫹-current-bearing bipolar cells
third-order neurons in the rat retina were reported to express
GABAC receptors, which are insensitive to bicuculline, a
GABAA receptor antagonist (Euler and Wässle 1998; Feigenspan et al. 1993; Pan 2000). Over 80 Na⫹-current-bearing cone
bipolar cells were examined. All of them displayed GABAevoked currents in the presence of bicuculline. As a typical
example shown in Fig. 2, the presence of Na⫹ currents in this
cone bipolar cell was confirmed by depolarizing voltage pulses
(Fig. 2A). In the same cell, co-application of GABA (100 ␮M)
and bicuculline (200 ␮M) elicited a sustained inward current
when the membrane potential was held at ⫺80 mV (Fig. 2B).
No significant current was observed when GABA was coapplied with bicuculline in retinal third-order neurons (data not
shown) (see Pan 2000), confirming that GABAA receptors
3. Comparison of Na⫹ and Ca2⫹ currents in cone bipolar cells. A: the slow component of inward currents evoked from a cone
bipolar cell could be blocked by Co2⫹ (4 mM;
middle) but not by TTX (500 nM; bottom).
Current traces were evoked by a test pulse at
⫺10 mV from the holding potential of ⫺70
mV. Recordings were made in normal Hanks’.
B: Na⫹ and Ca2⫹ currents evoked by test
pulses of ⫺60 to 0 mV at increments of 10
mV from the holding potential of ⫺80 mV
from another cone bipolar cell in normal
Hanks’. C: the I-V relationships for the peak
Na⫹ (Œ) and Ca2⫹ (●) currents shown in B. D:
currents recorded in high-Ca2⫹ (10 mM) solution from the same cell shown in B and C.
Note that the amplitude and time scales in D
are different from B to emphasize the inactivation time course of the Ca2⫹ currents.
FIG.
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on June 17, 2017
FIG. 2. Demonstration of the presence of GABAC receptor currents in a
Na⫹-current-bearing cone bipolar cell. A: the presence of voltage-dependent
Na⫹ currents was confirmed by test pulses of ⫺50 to ⫺10 mV. B: coapplication of GABA (100 ␮M) and bicuculline (200 ␮M) evoked a sustained
inward current. Recordings were made in normal Hanks’ containing Co2⫹ (4
mM). The cell was held at ⫺80 mV.
Mammalian cone bipolar cells have been reported to express
both low-voltage-activated (LVA) Ca2⫹ currents and L-type
high-voltage-activated Ca2⫹ channels (de la Villa et al. 1998;
Hartveit 1999; Pan 2000; Satoh et al. 1998). We have previously reported that the LVA Ca2⫹ currents of rod and cone
bipolar cells display distinct activation and inactivation kinetics (Pan 2000). When recordings were made in normal Hanks’,
inward currents with slower activation and inactivation than
those of above-described Na⫹ currents were also observed in
cone bipolar cells. These were Ca2⫹ currents since the currents
were blocked by Co2⫹ but not by TTX (Fig. 3A). In many of
these cells, however, Na⫹ currents were significantly larger
than Ca2⫹ currents when cells were depolarized from ⫺80 mV
and recorded in the physiological Ca2⫹ concentration. A typical example is shown in Fig. 3B. The I-V relationships for the
peak Na⫹ (Œ) and Ca2⫹ (F) currents are shown in Fig. 3C. The
amplitude of the peak Ca2⫹ currents did not find to show too
much variation among cone bipolar cells with the average
value of 17.4 ⫾ 5.6 pA (mean ⫾ SD; n ⫽ 9). When recordings
were made in high-Ca2⫹ (10 mM) solution, as expected, Ca2⫹
currents become much larger (also see Pan 2000). For comparison, the currents shown in Fig. 3D were recorded in highCa2⫹ from the same cell shown in Fig. 3B. For this cell, the
Na⫹ currents were masked by the Ca2⫹ currents and were
barely noticed. But, for many other cells, the presence of Na⫹
currents could still be observed in high-Ca2⫹ solution (data not
shown). Voltage-activated Ca2⫹ currents in Na⫹-current-bearing bipolar cells appeared to be activated at slightly more
negative potentials (around ⫺60 to ⫺50 mV) than that of Na⫹
currents (around ⫺50 to ⫺40 mV as described in the preceding
text). Furthermore the inactivation kinetics of the LVA Ca2⫹
currents resembles that of the LVA Ca2⫹ current of cone
bipolar cells previously described (Pan 2000).
NA⫹ CURRENTS IN RETINAL BIPOLAR CELLS
Na⫹ currents in both
ON
and
OFF
cone bipolar cells
Since Na⫹ currents were observed only in a portion of
cone bipolar cells, we investigated whether Na⫹ channels
were specifically expressed only in the ON or OFF type of
bipolar cells. It has been known that ON and OFF bipolar cells
express different glutamate receptors: metabotropic glutamate receptors in ON bipolar cells and ionotropic glutamate
receptors in OFF bipolar cells. Particularly activation of the
metabotropic glutamate receptors by L-2-amino-4-phosphonobutyric acid (L-AP4), a selective agonist of metabotropic
glutamate receptors, closes cGMP-gated channels (Nawy
and Jahr 1990; Shiells and Falk 1990; Slaughter and Miller
1981). Mammalian OFF cone bipolar cells have been reported
to express kainate subtype of glutamate receptors (DeVries
and Schwartz 1999). Thus we examined the response of
Na⫹-current-bearing cone bipolar cells to kainate, L-AP4, or
FIG. 6. Properties of voltage-dependence of activation and inactivation. A:
voltage dependence of activation and steady-state inactivation. Activation was
determined by test pulses depolarized from the holding potential of ⫺80 mV.
Activation measured by conductance was calculated from the I-V relationships,
from which the conductance was obtained by using the equation g ⫽ I/(V –
VNa). VNa is the Na⫹ current reversal potential which was ⬃40 –50 mV.
Activation data (●, n ⫽ 5) were normalized at the test potential of 30 mV.
Steady-state inactivation was determined by a series of conditioning pulses
from ⫺100 to ⫺30 mV for 1 s followed by a test pulse at ⫺10 mV.
Inactivation was measured by the amplitude of the peak-current evoked by the
test pulse. Inactivation data (䡲, n ⫽ 6) were normalized at the condition pulse
at ⫺100 mV. B: voltage dependence of time to peak (n ⫽ 8). C: voltage
dependence of decay time constant (n ⫽ 8). The decay time constants were
obtained by fitting the decay current course with a single exponential. All
recordings were made in normal Hanks’ containing Co2⫹ (4 mM). The data
points are mean values and the error bars represent SE from the indicated
number of cells.
glutamate. In these recordings, cGMP (1 mM) was added in
the electrode solution. Cells were usually held at ⫺70 or
⫺80 mV. Both types of responses were observed among
Na⫹-current-bearing cone bipolar cells. Figure 4A illustrates
a typical example of Na⫹-current-bearing cone bipolar cells
responding to L-AP4 (2 ␮M) or glutamate (1 mM) with an
outward current or a decrease of the holding inward current
(n ⫽ 31). A typical example of Na⫹-current-bearing cone
bipolar cells responding to kainate (300 ␮M) with an inward
current is shown Fig. 4B (n ⫽ 19). These results indicate
that Na⫹ channels are expressed in both ON and OFF cone
bipolar cells.
Na⫹-current distribution and amplitude
⫹
FIG. 5. The distribution of the peak amplitude and current density of Na
currents. Most of the recordings were made in normal Hanks’ containing Co2⫹
(4 mM). Some recordings were made in normal Hanks’. In the latter case, the
peak Na⫹ currents were measured after subtracting the slower component of
Ca2⫹ currents. Na⫹ currents were evoked from the holding potential of ⫺80
mV. The distribution of peak current (●) and current density (E) for 28 cone
bipolar cells was plotted over the cell number.
Na⫹ currents were observed in ⬃45% of recorded cone
bipolar cells. For example, among 493 recorded cone bipolar
cells, 223 of them displayed Na⫹ currents. In contrast, Na⫹
currents were never observed in rod bipolar cells (n ⬎ 700).
Although it was not possible for us to tell whether a cone
bipolar cell could have Na⫹ currents by its morphology, Na⫹
currents were more frequently observed in the cone bipolar
cells with a smaller soma and longer axon. Consistent with this
observation, the average whole cell capacitance of cone bipolar
cells with Na⫹ currents was 2.46 ⫾ 0.41 pF (mean ⫾ SD; n ⫽
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on June 17, 2017
⫹
FIG. 4. Response of Na -current-bearing cone bipolar cells to L-AP4 or
kainate. A: a typical response of a cone bipolar cell to L-AP4 (2 ␮M). 䡠 䡠 䡠 , the
0 current level. B: a typical response of a cone bipolar cell to kainate (300 ␮M).
In each case, the Na⫹ current trace is shown in the right panel. Recordings
were made in normal Hanks’. Electrode solution contained cGMP (1 mM).
2567
2568
Z.-H. PAN AND H.-J. HU
Voltage dependence of activation and inactivation
Voltage dependence of activation was calculated from the
I-V relationships of the Na⫹ currents. The averaged data (F)
for five cells were shown in Fig. 6A (right). The half-maximum
activation potential and the slope factor were ⫺24.7 and 5.9
mV, respectively. Steady-state inactivation was determined by
conditioning pulses ranging from ⫺100 to ⫺30 mV followed
by a test pulse at ⫺10 mV. The averaged data (䡲) from six cells
were also shown in Fig. 6A (left). The half-maximum inactivation potential and the slope factor were ⫺68 and 11.8 mV,
respectively.
The activation and inactivation kinetics of the Na⫹ currents
were determined by measuring the time to peak and the decay
⫹
FIG. 8. Demonstration of Na -dependent spikes in a cone bipolar cell. A:
the presence of voltage-dependent Na⫹ currents was confirmed in voltageclamp recordings. B: in current-clamp mode, spikes were evoked by current
pulses (400 ms) from the holding current of ⫺3.8 pA to 5 and 10 pA. C: after
the application of TTX (500 nM), no spikes were evoked by the same current
pulses. Recordings were made in normal Hanks’ and electrode solution contained K-gluconate.
time constant at different test potentials. The decay time constants were obtained by fitting the decay current with a single
exponential. The average values of the time to peak (n ⫽ 8)
and decay time constant (n ⫽ 8) were plotted versus the test
potentials in Fig. 6, B and C.
Recovery from inactivation
⫹
FIG. 7. Recovery of Na currents from inactivation. A: paired-pulse protocols used to determine the time course of recovery from inactivation. In each
paired pulse, cells were first depolarized from the holding potential of ⫺80 to
⫺10 mV for 100 ms (condition pulse) followed by a test pulse (20 ms to ⫺10
mV) with a varied time delay. Paired pulses were applied once every 5 s. B: the
time course of Na⫹ current recovery from inactivation. The ratio values of the
peak Na⫹ currents evoked by test pulse and condition pulse are plotted in
relation to time delay. The data points are mean values and the error bars
represent SE obtained from 8 cells. —, the fitting to the sum of 2 exponential
functions with time constant of 6.2 and 81 ms.
The recovery of Na⫹ currents from inactivation was determined by a series of paired pulses (Fig. 7A). In each pairedpulse, cells were first depolarized from the holding potential of
⫺80 to ⫺10 mV for 100 ms (condition pulse) to fully inactivate the Na⫹ currents. Then, after a varied time delay, a second
test pulse (20 ms at ⫺10 mV) was applied. The time course of
recovery was constructed by plotting the ratios of the peak Na⫹
currents evoked by the test pulse to that evoked by the condition pulse versus time delays (n ⫽ 8; Fig. 7B). The time course
of recovery could be fitted by a sum of two exponential
functions with time constants of 6.2 and 81 ms.
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on June 17, 2017
121). In contrast, the average whole cell capacitance of cone
bipolar cells without Na⫹ currents was 2.74 ⫾ 0.44 pF (n ⫽
109). These two values are significantly different (P ⬍ 0.001;
by t-test).
Na⫹ currents were observed in cone bipolar cells that did not
retain axon terminals. In a few cases, we were able to identify
cone bipolar cells without axon and Na⫹ currents were also
observed in some of these cone bipolar cells. Furthermore we
did not notice there were any correlations between the amplitude of the Na⫹ current and the presence or absence of axon or
axon terminals. Our results suggest that the Na⫹ channels are
at least located in the soma.
When recordings were made in normal Hanks’, the peak
Na⫹ current of cone bipolar cells ranged from a few to ⬎300
pA. Figure 5 shows the distribution of the peak Na⫹ currents
and the current densities for 28 cone bipolar cells. The Na⫹
current density was obtained by dividing the peak current by
the cell membrane capacitance. The average Na⫹ current is
93.1 ⫾ 95.8 pA (n ⫽ 28). The average Na⫹ current density is
39.1 ⫾ 44.6 pA/pF (n ⫽ 28). As shown in Fig. 5, in some cells,
Na⫹ current densities are ⬎100 pA/pF.
NA⫹ CURRENTS IN RETINAL BIPOLAR CELLS
Na⫹-dependent action potentials in cone bipolar cells
DISCUSSION
Voltage-dependent Na⫹ currents in cone bipolar cells
In this study, we report voltage-dependent Na⫹ currents in
rat retinal bipolar cells with patch-clamp recordings. Bipolar
cells were identified by their characteristic morphology. Two
lines of evidence further ensure that the Na⫹-current bearing
cells recorded in this study were bipolar cells. First, these cells
showed bicuculline-insensitive GABA, or GABAC receptormediated, responses. GABAC receptors have been reported to
be expressed in bipolar cells but not in third-order neurons in
the rat retina (Euler and Wässle 1998; Feigenspan et al. 1993;
Pan 2000). Second, we also showed that a portion of these cells
responded to L-AP4 or glutamate with a decrease in conductance, a unique property of ON bipolar cells. Thus the Na⫹
currents observed in this study could not be recorded from
retinal third-order neurons.
The bipolar cells displaying Na⫹ currents were morphologically identified as cone bipolar cells, which have been previously described in detail (Pan 2000). In addition, the LVA
Ca2⫹ currents in these Na⫹-current bearing bipolar cells displayed the characteristic properties of the LVA Ca2⫹ currents
of cone bipolar cells (Pan 2000), further supporting that Na⫹
currents are expressed in cone bipolar cells.
It should be mentioned that the present study was performed
on acutely dissociated bipolar cells. Large Na⫹ currents were
also observed shortly after the dissociation (⬍20 min). Moreover, Na⫹ currents were never observed in rod bipolar cells
even though a large number of rod bipolar cells were recorded.
Therefore it is very unlikely that the Na⫹ current could arise as
some type of culture artifact.
Taken together, our results indicate that voltage-dependent
Na⫹ currents are expressed in cone bipolar cells of the rat
retina. Markedly, a significant portion (⬃45%) of cone bipolar
cells was observed to show Na⫹ currents in this study. The
magnitude of the Na⫹ currents among cone bipolar cells varied
widely. The variation was not found to be correlated to the
presence or absence of the axon and axon terminals. Furthermore the Na⫹ currents were found to be expressed in both ON
and OFF of cone bipolar cells. However, it remains to be
determined whether Na⫹ channels are expressed in specific
subtypes of ON and OFF cone bipolar cells.
The finding of Na⫹ currents is consistent with an early
report of immunolocalization of Na⫹ channel ␣ subunits in a
portion of cat and monkey bipolar cells (Miguel-Hidalgo et al.
1994). Surprisingly, however, this is the first report of voltageactivated Na⫹ currents in retinal bipolar cells with electrophysiological recordings. Why have voltage-dependent Na⫹ currents of bipolar cells not been observed in previous studies?
There are several possible reasons. First, because only a portion of cone bipolar cells express Na⫹ currents, these bipolar
cells might have been missed in previous recordings. Second,
bipolar cells displaying Na⫹ currents recorded in previous
studies might have been considered to be third-order neurons.
Finally, voltage-dependent Na⫹ currents may only be expressed in mammalian cone bipolar cells. In fact, most of the
previous studies of membrane currents of bipolar cells were
carried out in lower vertebrates (Connaughton and Maguire
1998; Kaneko and Tachibana 1985; Lasater 1988; Maguire et
al. 1989; Tessier-Lavigne et al. 1988). Studies of mammalian
retinal bipolar cells were previously mainly performed on rod
bipolar cells (Gillette and Dacheux 1995; Karschin and Wässle
1990) which do not express Na⫹ currents as further confirmed
in this study.
Properties of Na⫹ currents
The biophysical properties of the Na⫹ currents in cone
bipolar cells are largely similar to that of Na⫹ currents reported
in retinal third-order neurons (Barnes and Werblin 1986;
Hidaka and Ishida 1998; Kaneda and Kaneko 1991; Lipton and
Tauck 1987). The potential range of the peak current occurring
from ⫺20 to 0 mV is also similar to that of retinal third-order
neurons (Barnes and Werblin 1986; Lipton and Tauck 1987).
Both the activation and inactivation are rapid and voltage
dependent. In addition, there is an overlap in the activation and
steady-state inactivation curves (from ⫺40 to ⫺30 mV), suggesting Na⫹ currents may be constantly activated within this
potential range. Recovery of Na⫹ currents from inactivation is
rapid. Thus a brief hyperpolarization of membrane potential
could partially remove the inactivation of Na⫹ channels.
Under normal physiological conditions, the amplitude of
Na⫹ currents of cone bipolar cells is at least more than one
order of magnitude smaller than that of retinal ganglion cells.
However, since bipolar cells are much smaller than retinal
ganglion cells, the difference in the Na⫹ current density between cone bipolar cells and retinal third-order cells, such as
ganglion cells, does not appear to be so large. Na⫹ current
densities in retinal ganglion cells were reported to be 100 –300
pA/pF (Hidaka and Ishida 1998; Lipton and Tauck 1987). In
fact, Na⫹ current densities observed in some cone bipolar cells
in this study are close to or comparable to these values (see Fig.
5). Bipolar cells with such a large Na⫹ current density would
be expected to generate Na⫹ spikes. Indeed, we showed in this
study that Na⫹-dependent action potentials could be evoked by
current injections.
However, under our vitro current-clamp recording conditions, only a single spike could be observed. Interestingly, such
property resembles that of amacrine cells in vivo. As having
been demonstrated previously for retinal amacrine cells (Eliasof et al. 1987), the lack of the capability to fire a series of
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on June 17, 2017
We further determined whether Na⫹-dependent spike activities could be present in cone bipolar cells. The recordings
were made in normal Hanks’. K-gluconate was used in intracellular solution without blocking K⫹ currents (see METHODS
for details). Under these recording conditions, Na⫹-dependent
action potentials were observed. An example of such recordings is shown in Fig. 8. The cone bipolar cell was first recorded
in the voltage-clamp mode to confirm the presence of voltagedependent Na⫹ currents in this cell (Fig. 8A). Then the recordings were switched to the current-clamp mode resulted in the
cell being clamped at ⫺3.8 pA. Stepwise current pulses were
applied to depolarize the cell from the holding current of ⫺3.8
pA to 5 and 10 pA for 400 ms. A single spike was observed at
the beginning of the depolarization evoked by current injections (Fig. 8B). After an application of 500 nM TTX, the same
current injections did not evoke any spike (Fig. 8C). Similar
results were observed in four other cone bipolar cells.
2569
2570
Z.-H. PAN AND H.-J. HU
action potentials could be due to the intrinsic properties of the
channel itself and/or other membrane currents, such as voltageactivated K⫹ and Ca2⫹-activated currents.
The dark membrane potentials of bipolar cells are believed
to be around ⫺40 to ⫺45 mV with ON bipolar cells depolarizing and OFF bipolar cells hyperpolarizing in responding to
light stimulation. The steady-state inactivation curve of the
Na⫹ current suggests that most of the Na⫹ channels in ON
bipolar cells would be in the inactivated state(s). On the other
hand, the light-evoked hyperpolarization may partially remove
the inactivation of the Na⫹ channels in OFF bipolar cells.
Therefore the voltage-activated Na⫹ channels would be more
likely to be activated in OFF cone bipolar cells in vivo.
Functional implications of expression Na⫹
channels in bipolar cells
This research was supported by National Eye Institute Grant EY-12180.
REFERENCES
BARNES S AND WERBLIN F. Gated currents generated single spike activity in
amacrine cells of the tiger salamander retina. Proc Natl Acad Sci USA 83:
1509 –1512, 1986.
BOYCOTT BB AND DOWLING JE. Organization of the primate retina: light
microscopy. Philos Trans R Soc Lond B Biol Sci 255: 109 –184, 1969.
BOYCOTT BB AND KOLB H. The connections between the bipolar cells and
photoreceptors in the retina of the domestic cat. J Comp Neurol 148:
91–114, 1973.
BURRONE J AND LAGNADO L. Electrical resonance and Ca2⫹ influx in the
synaptic terminal of depolaring bipolar cells from the goldfish retina.
J Physiol (Lond) 505: 571–584, 1997.
CONNAUGHTON VP AND MAGUIRE G. Differential expression of voltage-gated
K⫹ and Ca2⫹ currents in bipolar cells in the zebrafish retinal slice. Eur
J Neurosci 10: 1350 –1362, 1998.
DACHEUX RF AND RAVIOLA E. The rod pathway in the rabbit retina: a depolarizing bipolar and amacrine cell. J Neurosci 6: 331–345, 1986.
DE LA VILLA P, VAQUERO CF, AND KANEKO A. Two types of calcium currents
of the mouse bipolar cells recorded in the retinal slice preparation. Eur
J Neurosci 10: 317–323, 1998.
DEVRIES SH AND SCHWARTZ EA. Kainate receptors mediate synaptic transmission between cones and “off” bipolar cells in a mammalian retina.
Nature 397: 157–160, 1999.
ELIASOF S, BARNES S, AND WERBLIN F. The interaction of ionic currents
mediating single spike activity in retinal amacrine cells of the tiger
salamander. J Neurosci 7: 3512–3524, 1987.
EULER T AND WÄSSLE H. Immunocytochemical identification of cone bipolar
cells in the rat retina. J Comp Neurol 361: 461– 478, 1995.
EULER T AND WÄSSLE H. Different contributions of GABAA and GABAC
receptors to rod and cone bipolar cells in rat retinal slice preparation.
J Neurophysiol 79: 1384 –1395, 1998.
FAMIGLIETTI EV. Functional architecture of cone bipolar cells in mammalian
retina. Vision Res 21: 1559 –1564, 1981.
FEIGENSPAN A, WÄSSLE H, AND BORMANN J. Pharmacology of GABA receptor
Cl⫺ channels in rat retinal bipolar cells. Nature 361: 159 –161, 1993.
GILLETTE MA AND DACHEUX RF. GABA- and glycine-activated currents in the
rod bipolar cell of the rabbit retina. J Neurophysiol 74: 856 – 875, 1995.
GREFERATH U, GRUNERT U, AND WÄSSLE H. Rod bipolar cells in the mammalian retina show protein kinase C-like immunoreactivity. J Comp Neurol
301: 433– 422, 1990.
HAMILL OP, MARTY A, NEHER E, SAKMANN B, AND SIGWORTH FJ. Improved
patch-clamp techniques for high-resolution current recording from cells and
cell-free membrane patches. Pflügers Arch 391: 85–100, 1981.
HARTVEIT E. Functional organization of cone bipolar cells in the rat retina.
J Neurophysiol 77: 1716 –1730, 1997.
HARTVEIT E. Reciprocal synaptic interactions between rod bipolar cells and
amacrine cells in the rat retina. J Neurophysiol 81: 2923–2936, 1999.
HIDAKA S AND ISHIDA AT. Voltage-gated Na⫹ current availability after stepand spike-shaped conditioning depolarization of retinal ganglion cells. Eur
J Physiol 436: 497–508, 1998.
KANEDA M AND KANEKO A. Voltage-gated sodium currents in isolated retinal
ganglion cells of the cat: relation between the inactivation kinetics and the
cell type. Neurosci Res 11: 261–275, 1991.
KANEKO A. Physiological and morphological identification of horizontal, bipolar, and amacrine cells in the goldfish retina. J Physiol (Lond) 207:
623– 633, 1970.
KANEKO A, PINTO LH, AND TACHIBANA M. Transient calcium current of retinal
bipolar cells of the mouse. J Physiol (Lond) 410: 613– 629, 1989.
KANEKO A AND TACHIBANA M. A voltage-clamp analysis of membrane currents in solitary bipolar cells dissociated from Carassium auratus. J Physiol
(Lond) 358: 131–152, 1985.
KARSCHIN A AND WÄSSLE H. Voltage- and transmitter-gated currents in isolated rod bipolar cells of rat retina. J Neurophysiol 63: 860 – 876, 1990.
KOLB H, NELSON R, AND MARIANI A. Amacrine cells, bipolar cells and
ganglion cells of the cat retina: a Golgi study. Vision Res 21: 1081–1114,
1981.
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on June 17, 2017
What might be the possible role of expression Na⫹ channels
in bipolar cells? First, as mentioned in the preceding text,
because there is an overlap in the activation and steady-state
inactivation curves, Na⫹ channels might be constantly activated around bipolar cell resting membrane potentials, which
could support cells’ electrical excitability. Second, activation
of Na⫹ channels, or Na⫹ spikes, in bipolar cells may speed up
the membrane depolarization and shape light-response waveform and, in turn, affect Ca2⫹ current activation and transmitter
release at the axon terminals.
Furthermore Na⫹ channels might play a role in bipolar cell
signal propagation. It is commonly thought that the space
constant of bipolar cells is much longer than the axons of
bipolar cells and, thus the membrane potentials can propagate
from dendrites to axon terminals passively without significant
loss. However, at least in rats, both the dendrites and axons of
most cone bipolar cells after dissociation appeared to be markedly thinner than that of rod bipolar cells (Pan 2000). In
addition, membrane resistance of bipolar cells in vivo has been
reported to be lower than the value obtained in vitro probably
due to cell coupling and constant activation of membrane ion
channels (Tessier-Lavigne et al. 1988). Together these factors
could significantly reduce the space constant for certain groups
of bipolar cells. Particularly, as mentioned in the preceding
text, voltage-dependent Na⫹ currents were more frequently
observed in cone bipolar cells with smaller soma and longer
axon. Therefore it is possible that the expression of Na⫹
channels in bipolar cells may serve as a mechanism for boosting or facilitating membrane potential propagation.
However, a previous study reported that co-application of
TTX and cadmium was not found to show effect on the
multi-quantal release of glutamate from mouse bipolar cells in
light-adapted retinal slices (Tian et al. 1998). This appears to
suggest that retinal bipolar cells don’t generate spontaneous
Na⫹ or Ca2⫹ activities under the studied conditions. On the
other hand, a recent study reported that Mb1-type bipolar cells
in dark-adapted goldfish retinal slices were capable of generating light-evoked Ca2⫹ spikes (Protti et al. 2000). Further
study will be needed to determine the functional role of the
expression of voltage-gated Na⫹ currents in bipolar cells.
Particularly it would be interesting to determine whether Na⫹
currents could shape the light response or generate lightevoked Na⫹ spikes in mammalian cone bipolar cells in vivo.
Furthermore, experiments with paired recordings of bipolar
cells and third-order neurons in retinal slices may be able to
determine whether Na⫹ currents play a role in bipolar cell
signal propagation.
NA⫹ CURRENTS IN RETINAL BIPOLAR CELLS
POURCHO RG AND GOEBEL DJ. A combined Golgi and autoradiographic study
of 3H-glycine-accumulating cone bipolar cells in the cat retina. J Neurosci
7: 1178 –1188, 1987.
PROTTI DA, FLORES-HERR N, AND VON GERSDORFF H. Light evokes Ca2⫹
spikes in the axon terminal of a retinal bipolar cell. Neuron 25: 215–227,
2000.
SATOH H, AOKI K, WATANABE S-H, AND KANEKO A. L-type calcium channels
in the axon terminal of mouse bipolar cells. Neuroreport 9: 2161–2165,
1998.
SHIELLS PA AND FALK G. Glutamate receptors of rod bipolar cells are linked to
a cyclic GMP cascade via a G-protein. Proc R Soc Lond B Biol Sci 242:
91–94, 1990.
SLAUGHTER MM AND MILLER RF. 2-Amino-4-phosphonoburyric acid: a new
pharmacological tool for retinal research. Science 211: 182–185, 1981.
TESSIER-LAVIGNE M, ATTWELL D, MOBBS P, AND WILSON M. Membrane
current in retinal bipolar cells of the axolotl. J Gen Physiol 91: 49 –72, 1988.
TIAN N, HWANG TN, AND COPENHAGEN DR. Analysis of excitatory and inhibitory spontaneous synaptic activity in mouse retinal ganglion cells. J Neurophysiol 80: 1327–1340, 1998.
WERBLIN FS AND DOWLING JE. Organization of the retina of the mudpuppy,
Necturus maculosus. II. Intracellular recording. J Neurophysiol 32: 339 –
355, 1969.
YEH HH, LEE MB, AND CHEUN JE. Properties of GABA-activated whole-cell
currents in bipolar cells of the rat retina. Vis Neurosci 4: 349 –357, 1990.
ZENISEK D AND MATTHEWS G. Calcium action potentials in retinal bipolar
neurons. Vis Neurosci 15: 69 –75, 1998.
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on June 17, 2017
LASATER EM. Membrane currents of retinal bipolar cells in culture. J Neurophysiol 60: 1460 –1480, 1988.
LIPTON SA AND TAUCH DL. Voltage-dependent conductances of solitary ganglion cells dissociated from the rat retina. J Physiol (Lond) 385: 361–391,
1987.
MAGUIRE G, MAPLE B, LUKASIEWICZ P, AND WERBLIN F. ␥-aminobutyrate type
B receptor modulation of L-type calcium channel current at bipolar cells
terminals in the retina of the tiger salamander. Proc Natl Acad Sci USA 86:
10144 –10147, 1989.
MIGUEL-HIDALGO JJ, SNIDER CJ, ANGELIDES KJ, AND CHALUPA LM. Voltagedependent sodium channel alpha subunit immunoreactivity is expressed by
distinct cell types of the cat and monkey retina. Vis Neurosci 11: 219 –228,
1994.
NAWY S AND JAHR CE. Suppression by glutamate of cGMP-activated conductance in retinal bipolar cells. Nature 346: 269 –271, 1990.
NEHER E. Correction for liquid junction potentials in patch clamp experiments.
Methods Enzymol 207: 123–131, 1992.
PAN Z-H. Differential expression of high- and two low voltage-activated
calcium currents, and sodium currents in rod and cone bipolar cells of the rat
retina (Abstract). Invest Opthalmol Vis Sci Suppl 40: 791, 1999.
PAN Z-H. Differential expression of high- and two types of low-voltageactivated calcium currents in rod and cone bipolar cells of the rat retina.
J Neurophysiol 83: 513–527, 2000.
PAN Z-H AND LIPTON SA. Multiple GABA receptor subtypes mediate inhibition of calcium influx at rat retinal bipolar cell terminals. J Neurosci 15:
2668 –2679, 1995.
2571