Anisopotentiality in retinal ganglion cells

kinases in the retinal On-bipolar cell

Results

 Discussion

 Summary

 Characterization of retinal P10 cultures

Characterization of retinal P0 cultures

Alcon collaborations

References

To determine whether retinal ganglion cells integrate light information at the level of the dendrites.

Following the elegant work of Regehr and Armstrong it has become clear that certain neurons such as Purkinje cells and hippocampal neurons are not isopotential. This property could be studied using of low resistance patch-clamp electrodes to minimise the attenuation of the dendritic signal through space-clamping artifacts. When the soma was voltage-clamped and depolarized by brief command potentials, the regenerative current spike caused by the opening of voltage-gated sodium channels that underlie the action potential could be seen to separate into distinct population peaks. Regehr and Armstrong inferred that because threshold potential in the soma/ axon hillock would result in an all-or-none spike of excitable sodium current if excitable sodium channels were present exclusively in the axon and soma, then excitable sodium channels must therefore also be present in the dendrites, supporting the convictions of Sugimori and Llinas.

Methods and Conclusions

Whole-cell voltage-clamp recordings from ganglion cells were performed initially in 150 micron slices of Tiger salamander (Ambystoma tigrinum) retina from 6-8" animals kept at 4C. Slices were cut on filter paper and were laid flat across tracks of vacuum grease. Later a flat-mounted preparation was used to prevent damage to the dendritic layer through sectioning. Voltage-clamp recordings were made at the level of the soma using 1-2 MW electrodes, and cells were routinely voltage-clamped at -50mV. Brief depolarising voltage steps (3-4 mV, 100ms) were applied in increments using P-clamp software and acquired at the limit of resolution of the board. The threshold for the main regenerative spike was reached at around -33 mV, but before this potential smaller regenerative inward currents were elicited with a latency that became shorter latency as the somatic threshold potential was approached. As this somatic threshold was passed the latency of the smaller regenerative current decreased so that the two peaks that were separated at or near threshold, eventually merging into one as the magnitude of the depolarizing pulse became large enough to activate all the excitable sodium channels simultaneously. This smaller spike most likely represents a population of excitable sodium channels present in the dendrites with a higher (more negative) threshold potential, or alternatively a population of excitable calcium channels. At certain resonant holding potentials close to the somatic threshold the unclamped region would spontaneously fire trains of regenerative spikes for the duration of the depolarising pulse, confirming that the retinal ganglion cell is most probably anisopotential.

Return to contents

To determine the second messenger pathways involved in the regulation of the L-APB metabotropic receptor cascade in retinal On-bipolar cells. The effects of modulation of the activity of specific protein kinases upon L-APB responses in the On-bipolar cell were tested to determine which, if any, pathways might mediate adaptation in this neuron.

Bipolar cells are interneurons in the retina that relay visual information from photoreceptors to ganglion cells. The On-bipolar cell, which depolarizes in response to light, expresses a metabotropic L-glutamate receptor whose activation produces membrane hyperpolarization and a decrease in membrane conductance (Shiells et al., 1981; Nawy and Copenhagen, 1987). This receptor has a high affinity for the L-glutamate receptor agonist L-2-amino-4-phosphonobutyrate (APB) (Slaughter and Miller, 1981). On the basis of previous studies (Nawy and Jahr, 1990; Shiells and Falk, 1990; Nawy and Jahr, 1991; Yamashita and Wassle, 1991; de la Villa et al., 1995) a model has been proposed to explain how L-glutamate hyperpolarizes the On-bipolar cell membrane. The receptor is positively coupled via a G-protein to a cGMP phosphodiesterase (PDE). Binding of L-glutamate to the APB receptor activates the PDE and produces hydrolysis of cGMP. As a result of the fall in cGMP concentration, a cGMP-dependent cation channel closes and hyperpolarizes the cell. In the absence of agonist, some of the cGMP-dependent channels are open and continuously depolarize the cell due to an influx of Na⁺ through the channel (Yamashita and Wassle, 1991). Under voltage-clamp, this depolarization is observed as a tonic inward current. When exogenous cGMP is applied through the recording pipette, the inward current becomes larger as cGMP diffuses into the cell and opens cGMP-dependent cation channels (Nawy and Jahr, 1990; de la Villa et al., 1995). The APB receptor pathway is reminiscent of the transduction cascade expressed by photoreceptors (reviewed in Lagnado and Baylor, 1992), the cell to which the bipolar cell is developmentally most closely related. This current is voltage-independent (Shiells and Falk, 1994) and responses to APB do not desensitize in either whole-cell or perforated-patch recordings where high access resistances are obtained (Nawy and Jahr, 1990; Shiells and Falk, 1990; Yamashita and Wassle, 1991; de la Villa et al., 1995).

A gene encoding a rat APB receptor that is expressed by On-bipolar cells has recently been cloned and classified as mGluR6 (Nakajima et al., 1993). It is closely related to mGluR4, mGluR7 and mGluR8, metabotropic receptors that have a high affinity for APB (Okamoto et al., 1994; Duvoisin et al, 1995). Behavioral and electrophysiological experiments with mice that do not express the mGluR6 receptor reveal a profound visual impairment (Masu et al., 1995). More detailed behavioral studies in goldfish (Bilotta et al, 1995) and monkey (Schiller, 1992; Dolan and Schiller, 1994), using intra-ocular application of APB to saturate the receptor, show profound deficits in basic visual performance such as spatial contrast or in the detection of incremental changes in background light. These experiments firmly establish an important role for the APB receptor in vision.

More subtle physiological modulation of this receptor by endogenous mechanisms would provide a means for fine-tuning of spatial contrast and absolute sensitivity in response to changes in ambient light levels. There is increasing evidence for modulatory roles of dopamine, nitric oxide and neuroactive peptides in regulating synaptic transmission between retinal neurons, allowing the retina to maintain its sensitivity over a wide range of light intensities by trading high absolute sensitivity in dim light for greater spatial and temporal resolution at higher light intensities. It is not yet clear how light, or other environmental cues might control the release of these modulatory substances. The intracellular targets and mechanisms of action of these compounds are not well defined either. In addition the pathways for terminating the responses of the beta-adrenergic receptor and rhodopsin have been extensively studied. Receptor kinases that are activated by free bg subunits are believed to phosphorylate the C-termini of these receptors allowing the subsequent binding of proteins of the arrestin family. This sequence of events terminates receptor activity in these systems, but so far evidence for the role of arrestin-like proteins or G-protein receptor kinases in the L-APB receptor pathway in the On-bipolar cell has not been forthcoming.

In this study we have attempted to define functional sites within the APB receptor pathway that are modulated by kinases. These sites that provide a potential mechanism for the regulation of the On-bipolar cell synaptic response. Specifically, we report that responses to exogenous APB are profoundly altered by inhibitors of CaMKII and PKA, and by alkaline phosphatase. On this basis we propose that both CaMKII and PKA regulate the APB receptor pathway, and that intra-or extracellular ligands that activate these kinases may have important roles in controlling the sensitivity of this synapse and the visual system as a whole. In addition we have provided preliminary evidence for a G-protein receptor kinase that terminates the activity of the agonist-stimulated receptor.

Return to contents

Slice preparation. Tiger salamanders 6 to 8 inches in length (Kons Scientific, Sullivan) were maintained in an aerated tank at 2-4^oC and were anaesthetized with 3-aminobenzoic acid prior to dissection. Retinal slices were prepared as described previously (Nawy & Jahr, 1990). The slices were viewed under a total magnification of 640x using a water immersion objective with Hoffman optics and an Optivar assembly on a Zeiss Axioskop. Slices were continuously superfused with an oxygenated Ringer at room temperature (70^oF) following a 10 minute pre-incubation with a Ringer solution containing 0.1 mM kainate which substantially depleted the OFF-bipolar cell population. The slices were subsequently treated for 5 minutes with 400 U/ml hyaluronidase and 120 U/ml type IA collagenase (both Sigma). The salamander Ringer contained (mM) 108 NaCl, 2 CaCl₂, 2.5 KCl, 10 HEPES, 10 glucose and the pH was adjusted to 7.6 with NaOH.

Electrophysiological recordings. Patch pipettes were fabricated from hematocrit glass tubing (VWR Scientific) using a Narishige PP-83 two-stage vertical puller. After fire polishing electrodes had resistances of 1-2 MW when filled with a N-methyl D-glucamine phosphate pipette solution which, when adjusted to pH 7.4 with CsOH, contained (mM) 52 N-methyl D-glucamine (NMG), 52 H₃PO₄, 10 EGTA, 60 Cs⁺, 1 cGMP and 0.1 GTP. Whole-cell recordings were made using an Axopatch-1D amplifier (Axon Instruments, Foster City CA), and were obtained after applying slight negative pressure to the pipette at a holding potential of -30mV. Following the establishment of the whole-cell configuration the cell was voltage-clamped at -50mV and the series resistance (R_s) and input resistance (R_i) of the cell were monitored using Axobasic software. Whole-cell recordings that were kept for further analysis typically had R_i and R_s values in the range of 1-3 GW and 7-11 MW respectively. Data were low-pass filtered at 2kHz and stored on computer and videotape for further analysis. Continuous recordings shown were acquired at 7Hz through a Digidata 1200 Interface (Axon Instruments).

Drug application. The cell was continuously superfused with control Ringer containing 0.1mM picrotoxin and 1mM CoCl₂ from the central barrel of a multi-barrel array of flow pipes mounted on a piezo-bimorph assembly. Timed applications of drug were controlled by computer. This allowed the on-line acquisition of drug applications (acquired at 40 Hz). 2mM APB (or 1mM L-Glu) was applied at timed intervals to the cell of interest.

Tip-filling technique. In tip-fill experiments the shank of the electrode tip was filled with a pre-filtered solution with the desired nucleotide composition. The electrode was then back-filled with 4-6ml of the nucleotide solution containing the inhibitor of interest, so that the concentration of inhibitor was not substantially altered upon complete equilibration of the pipette contents.

Materials. All reagents were obtained from the Sigma Chemical Co. (St.Louis) unless otherwise stated. Stock solutions of adenosine 5'-triphosphate (MgATP, 100mM), guanosine 5'-triphosphate (GTP, sodium salt, 10mM), guanosine 3':5'-cyclic monophosphate (cGMP, sodium salt, 100mM). Guanosine 5'-O-3-thiotriphosphate (GTPgS, 100mM) was dissolved in pipette solution and stored at -20^oC after adjusting the pH to 7.4 with KOH. Stock solutions of the inhibitors KN-62 (1mM), obtained from Calbiochem (La Jolla, CA), and 3-isobutyl-1-methylxanthine (IBMX, 100mM) were dissolved in dimethylsulphoxide (final concentration DMSO <0.1% v/v) and stored frozen. The peptide inhibitor PKI 5-24 amide (Peninsula Laboratories, Belmont, CA) was reconstituted in distilled water at 1mg/ml and stored at -80^oC. Alkaline phosphatase was prepared freshly from solid. Stock solutions of APB (2mM) and L-Glu (100mM) were made up in Ringer and stored at 4^oC. Bovine catalytic subunit of protein kinase A was obtained in purified form from Promega and freshly reconstituted and maintained on ice.

Analysis and Acquisition. Data are presented as the means of n experiments ± the standard error of the mean. Significance was determined by the use of the Student's two-sample unpaired t-Test assuming unequal variances. To determine the normalized holding current, measurements were taken at 10, 20 and 30s, and at 30s intervals thereafter. For control cells and recordings with PKI inhibitor, inward current measurements were binned at 1 minute intervals and then normalized to the peak holding current for each recording. To determine the mean time course for APB responses, agonist was applied at one minute intervals initially, and thereafter every two minutes. All other measurements of holding current and APB responses presented were taken 2 minutes after obtaining the whole-cell configuration. To determine the percent suppression of holding current, the magnitude of the APB response was divided by the value of the holding current just prior to agonist application.

 Return to contents

On-bipolar cells dialyzed with cGMP develop a persistent cation-selective inward current when voltage-clamped at negative potentials. Application of APB reversibly suppresses this cGMP-dependent current (Nawy and Jahr, 1990; Shiells & Falk, 1990; de la Villa et al., 1995). A representative recording showing the response to successive applications of APB is shown in Figure 1A. We routinely observed a run-down in the magnitude of the response to APB. The data are summarized in Figure 1B. Initially the amplitude of the mean APB response (circles) increased as cGMP entered the cell and augmented the magnitude of the APB-sensitive holding current (Nawy and Jahr, 1990). The APB response then decayed to an average of about 35% of the maximum over the next 10 minutes (n=12). Since the APB receptor is a metabotropic receptor that does not gate a channel directly, run-down in the response suggests that either the bound receptor can no longer signal to the rest of the pathway, or that there is a reduction in G-protein or PDE activity downstream of the receptor. Regardless of mechanism, dialysis of the cell during whole-cell recording produced a loss of functional coupling between the receptor and the cGMP-dependent cation conductance.

Responses to APB were normalized to the size of the holding current by dividing response amplitude by holding current at each time point for individual cells, and the mean data are plotted in Figure 1C. The normalization was performed as we noticed a small time-dependent decline in the holding current in 8 of the 12 cells tested, although no decline was observed in the trace illustrated in Figure 1A. This decline in holding current may reflect a decrease in cGMP-dependent current. The normalization procedure also takes into account the initial increase in holding current as cGMP enters the cell, as well as this more gradual decrease in holding current. This plot shows that the decline in receptor coupling began immediately after breaking into the cell, even when the absolute size of the APB response is still increasing, and that coupling continued to decline before reaching a steady-state after about 6 minutes.

The time-dependent rundown in the APB response suggests that a diffusible factor is required to support coupling between the receptor and the channel underlying the cGMP-dependent current. There is evidence that dopamine potentiates L-glutamate responses in retinal horizontal (Knapp and Dowling, 1987) and Off-bipolar cells via a PKA-dependent mechanism (Maguire and Werblin, 1994). Furthermore, there are consensus PKA phosphorylation sites on the C-terminus of the rat mGluR6 receptor (Nakajima et al., 1993). We therefore investigated the potential role of PKA in regulating the amplitude of the APB response.

The rate of rundown of the APB response was accelerated when a selective peptide inhibitor of protein kinase A, PKI 5-24 amide (Cheng et al., 1986), was included in the pipette (Figure 2A). Although the initial application of APB reversibly suppressed most of the holding current, in the recording shown the response to APB was almost completely abolished within two minutes in the presence of inhibitor. Overall, the proportion of the holding current that was suppressed by APB declined more rapidly and more completely in cells internally dialyzed with PKI 5-24 amide (n=8) than in control cells (Figure 2C), which did not show a complete loss of the APB response. The inhibition of receptor function was specific as PKI 5-24 amide did not change the rate at which the normalized holding current declined relative to the control population, nor did it significantly reduce the peak holding current (Figure 2B, P=0.34).

The enhanced rate of run-down in the presence of PKI suggests that the diffusible factor responsible for run-down in control recordings with ATP, GTP and cGMP may be protein kinase A. Further these experiments also suggest that the degree of coupling between the receptor and the phosphodiesterase activity might be increased by dialysing the cell with the constitutively active catalytic subunit of protein kinase A. Figure 3 shows the effect of addition of 2mg/ml of purified PKA catalytic subunit to the pipette. The presence of this subunit caused the agonist response to become prolonged, and the plateau response phase lasted up to 10 seconds after washout. Further in 4 experiments little or no run-down of the response was observed with 2-4 mg/ml of catalytic subunit in the pipette after 2 minutes (Table 1).

Table 1 summarises the percent suppression of holding current by L-APB with 5mM PKI, 2-4mg/ml PKA catalytic subunit or 2mM ATP alone in the pipette. The data indicates that PKI significantly enhances the rate of run-down and that PKA catalytic subunit prevents it. Experiments where 10mM EGTA was replaced with 10mM BAPTA were also performed (Table 1). BAPTA is a more rapid chelator than EGTA and has been shown to abolish fast calcium-dependent events in the squid giant synapse. The inclusion of 10mM BAPTA significantly enhanced run-down after 2 minutes suggesting that calcium may also be involved in the regulation of coupling of the L-APB receptor to the PDE.

We next tested whether addition of exogenous phosphatase would also increase the run down of the response. Alkaline phosphatase has been used previously to determine the effect of dephosphorylation upon whole-cell currents (Payet & Depuis, 1992; Kim, 1991; Chen et al., 1990). In the presence of alkaline phosphatase there was also a nearly complete run-down of the APB response within 4 minutes of establishing the whole-cell configuration (Figure 4A), that was significantly decreased with respect to the control population after 2 minutes (Figure 4B, right panel). The degree of run-down was comparable to the PKI-treated cells.

We also observed a second unexpected effect of alkaline phosphatase. A holding current initially developed when cGMP was included in the tip-filling solution, but thereafter started to decline rapidly. Holding current was reduced from an average of 125 pA in control cells to approximately 40 pA in alkaline phosphatase-treated cells, and it is possible that a substantial portion of the residual current was due to cGMP-independent current (Figure 4B, left panel). This experiment raised the possibility that phosphorylation is required for activation of the channels underlying the cGMP-dependent current.

It is known that Ca²⁺-calmodulin regulates the cGMP affinity of CNG channels (Hsu and Molday, 1993; Chen and Yau, 1994; Gordon et al., 1995), and it seemed reasonable to us that they may also play a role in regulation of the cGMP-dependent current in the On-bipolar cell. Calcium-Calmodulin activates both phosphatases and kinases, and thus addition of calmodulin inhibitors might simultaneously affect both phosphorylation and dephosphorylation. We therefore tested the idea that the reduction of the cGMP-dependent current observed in the presence of alkaline phosphatase might also be observed when a calmodulin-dependent protein kinase is inhibited. In particular, the type II form has been implicated in synaptic plasticity (Wang and Kelly, 1995). Consequently, the specific CaMK inhibitor KN-62 (Tokumitsu et al., 1990) was added to the recording pipette. In cells dialyzed with KN-62 (1 mM) the mean holding current was greatly reduced compared to control cells (Figure 5A). In all cells tested KN-62 reduced the holding current, measured after 2 minutes, to around 30% of the current in control cells (Figure 5B, left panel). The reduction in holding current observed in the presence of KN-62 or alkaline phosphatase was virtually identical. The amplitude of the APB response was decreased, but KN-62 did not significantly alter the proportion of current suppressed by APB relative to that in control cells (Figure 5B, right panel).

The possibility that KN-62 reduces the holding current by enhancing PDE activity was tested by introducing the broad-spectrum PDE inhibitor IBMX into the pipette. Patch-pipettes were tip-filled with KN-62 and back-filled with an identical solution containing 100mM IBMX. Introducing IBMX into the cell did not prevent suppression of holding current by KN-62 (Figure 5B left panel, n=6). The observation that cGMP hydrolysis is not required for inhibition of the cGMP-dependent current by the CaMK inhibitor suggests that the inhibitor acts downstream from the PDE, possibly at the level of the channel itself.

The validity of the assumption that IBMX was effective in inhibiting PDE activity was tested. It has been shown previously that dialyzing the On-bipolar cell with GTPgS leads to an irreversible suppression of the holding current by APB, presumably by the persistent activation of PDE (Nawy & Jahr, 1990). When we substituted GTPgS for GTP in the pipette solution, the holding current was suppressed by more than 50% following application of agonist (Figure 6B, n=5). Experiments were performed where the electrode was tip-filled with GTPgS and then back-filled with an identical solution supplemented with IBMX (100 mM). The initial application of APB evoked a sustained suppression of the holding current, followed by a progressive recovery of holding current as the intracellular concentration of IBMX increased. No further response to APB could be elicited (Figure 6A). The effect of IBMX on holding current is summarized in Figure 6B. In cells dialyzed with IBMX as well as GTPgS (n=7), the holding current amplitude did not differ significantly from control (cGMP alone) cells (n=12). These data suggest that the concentration of IBMX used in these experiments was sufficient to fully inhibit PDE activity.

The time-dependence of the decrease in holding current and in the response to L-APB indicate that cGMP is not in itself sufficient to support maximal activation of the cGMP dependent conductance and coupling between the mGluR6 receptor and PDE. These observations raise the possibility that the synaptic pathway of the ON-bipolar cell may be regulated through other mechanisms, such as protein phosphorylation or other ATP-dependent processes. Consequently, ATP was omitted from the pipette solution to determine whether either the holding current or the L-APB response were ATP-dependent in nature. Figure 7A shows the effect of removing ATP from the pipette solution. Neither the initial size of the holding current nor the proportion of holding current initially suppressed by L-APB were significantly affected by the removal of ATP (n=14), which may reflect the time required for cytosolic ATP levels to become depleted. However, successive applications of L-APB evoked progressively smaller responses associated with only a partial recovery of the holding current following L-APB washout. Although the suppression of the holding current and L-APB response probably reflect the run-down of an ATP-dependent process in these cells, in the absence of a competitive non-hydrolyzable ATP analog, such as ATPgS, it is not possible to mask the effects of any residual ATP that might be synthesised in the cell.

Figure 7B shows the effect of dialyzing the cell with 1mM ATPgS upon the magnitude of the holding current and response to L-APB. In the presence of ATPgS, L-APB evoked an irreversible suppression of the holding current, after which no further response to L-APB could be elicited (n=7). Figure 8A compares the responses to a single 5 second application of L-APB with 2mM ATP, 1mM ATPgS and ATP-free pipette solutions. The initial rate of suppression of the holding current by L-APB was not substantially affected by the removal of ATP, although ATPgS may slow the rate of suppression of the holding current. However, in the ATP-depleted cell population, the recovery of the holding current following L-APB washout was slower, and in the presence of ATPgS little or no recovery of holding current was observed following washout. These observations indicate that an adequate supply of a hydrolyzable ATP substrate is required for recovery of the holding current following L-APB application. As ATPgS only partially inhibited the peak holding current (P<0.05 vs control), then ATP-dependent events must presumably occur less frequently in the absence of agonist than in the presence of agonist.

IBMX reverses the effect of GTPgS, but not ATPgS, upon the holding current.

It is not apparent whether the effects of ATPgS result from the constitutive activation of PDE or an ATP-dependent channel opening process, or both. The broad-spectrum PDE inhibitor IBMX was used to discriminate between events that serve to regulate PDE activity, and those that act downstream of the PDE to modulate channel activity. It has been demonstrated that dialysis of the ON-bipolar cell with GTPgS leads to an irreversible suppression of the holding current, presumably due to the constitutive activation of PDE (Nawy & Jahr, 1990a). Thus, when 100mM GTPgS was included in the back-fill solution but excluded from the tip, the magnitude of the holding current two minutes following L-APB application was suppressed by more than 50% relative to the control population (Fig.8B, n=5), an effect that was fully reversed by back-filling with 100mM IBMX (n=7). The effectiveness of the tip-filling technique is shown in Fig.8A, where the tip-fill solution contained GTPgS and the back-fill solution contained IBMX and GTPgS. Under these conditions the initial application of L-APB evoked a sustained suppression of the holding current which gradually reversed towards a new steady-state, presumably due to the action of IBMX, the presence of which could be inferred by the inhibition of further responses to L-APB.

To address the site of action of ATPgS electrodes were tip-filled with a solution containing 1mM ATPgS and then back-filled with an identical solution containing 100mM IBMX. Figure 8B shows that back-filling pipettes with IBMX neither prevented nor reduced the extent of the ATPgS-mediated suppression of the holding current (n=5). These observations support the contention that GTPgS enhances PDE activity and indicate that ATPgS suppresses channel activity at one or more sites downstream from the PDE.

Heparin has been demonstrated to inhibit both rhodopsin kinase and b-adrenergic receptor kinase, preventing the shut-down of responses to these receptors. In five separate experiments dialyzing the On-bipolar cell with heparin at 1.5mg/ml appeared to progressively slow the rate of recovery of holding current following agonist washout. An example of this effect is shown in Fig.9. However, in three experiments dialysis with 1mM inositol hexakisphosphate (IP6), an inhibitor of members of the arrestin family of proteins, did not appear to alter the kinetics of the response to APB.

Figure 9. Heparin inhibits the recovery of holding current following agonist washout. Copy of chart-recording of cell voltage-clamped at -50mV with 1.5mg/ml heparin added to the pipette. The duration of applications is as indicated by the horizontal bars.

 Return to contents

Our observation that inhibitors of PKA block the APB response implies that transmitters that increase intracellular cAMP levels via the activation of adenylate cyclase should augment receptor coupling. Two potential transmitters are dopamine and vasoactive intestinal peptide (VIP), both released by subsets of amacrine cells (Brecha et al., 1984; Witkovsky et al., 1984). Neither substance gates ion channels directly, but instead modulate responses to other transmitters. Dopamine potentiates L-glutamate responses in both horizontal (Knapp & Dowling, 1987) and Off-bipolar cells (Maguire and Werblin, 1994), and closes gap junctions between horizontal (Teranishi et al, 1983) and amacrine cells (Hampson et al., 1992; Mills and Massey, 1995). These actions of dopamine are believed to be mediated by an increase in cAMP levels (Knapp and Dowling, 1987; DeVries and Schwartz, 1989; Maguire and Werblin, 1994). VIP potentiates GABA responses in both ganglion and bipolar cells (Veruki and Yeh, 1992) through PKA activation (Veruki and Yeh, 1994). Modulation of L-glutamate, APB or light responses by dopamine or VIP has not yet been reported in On-bipolar cells, and is probably best examined using perforated-patch or sharp electrodes to prevent run-down.

Although our data cannot distinguish between the potential targets of PKA phosphorylation, biochemical studies of other G-protein-coupled receptors provide a possible clue. Kinases regulate coupling of both the b₂-adrenergic receptor and rhodopsin to their respective G-proteins (reviewed by Palczewski and Benovic, 1991; Prement et al., 1995). In general, it appears that the receptors, rather than the G-protein or effector enzymes, are the targets for phosphorylation. Both PKA (Benovic et al., 1985) and b-adrenergic receptor kinase (BARK) phosphorylate the b₂-adrenergic receptor to uncouple it from the G-protein (Benovic et al., 1986). Likewise, inactivation of rhodopsin is initiated by multiple phosphorylations through rhodopsin kinase (Kuhn, 1984). Analysis of the deduced amino acid sequence of the rat APB receptor (Nakajima et al., 1993), mGluR6, reveals several consensus PKA phosphorylation sites on the carboxyl-terminal end of the receptor. It is tempting to speculate that PKA phosphorylates the APB receptor, here initiating rather than terminating receptor coupling. Phosphorylation of the PDE is also possible, and there is evidence for phosphorylation of the inhibitory g subunit of PDE in rod photoreceptors, although not by PKA (Hayashi et al., 1991).

Inhibition of the cGMP-dependent current by KN-62 and alkaline phosphatase suggests either that phosphorylation is an absolute requirement for channel opening, or that it dramatically increases channel affinity for cGMP. It is possible that an upstream element of the cascade, such as the APB receptor, PDE or G-protein, is the target of phosphorylation. This seems unlikely, since inhibition of the current persisted in the presence of a sufficient concentration of IBMX to block hydrolysis of cGMP in cells dialyzed with GTPgS. Our results suggest instead that the phosphorylation site(s) are located on the channel or an associated protein.

Since CaMK is activated by Ca²⁺-calmodulin, an increase in the intracellular Ca²⁺ concentration near the channel should augment the size of the cGMP-dependent current. This finding has been reported in isolated rat bipolar cells using the perforated-patch recording technique (Yamashita and Wassle, 1991), where it was observed that increasing intracellular Ca²⁺ with the Ca²⁺ ionophore A23187 potentiated both the inward current and the size of the APB response. The source of Ca²⁺ under physiological conditions is unclear. One potential route of Ca²⁺ entry is via an L-type current that begins to activate near -30 mV, and is fully activated at about -10 mV (Maguire et al., 1989). Light depolarizes the On-type bipolar cell from a dark potential of about -35 mV to about -20 mV (Nawy and Copenhagen, 1987), a depolarization that should be sufficient to activate this L-type current, in addition to increasing cGMP-dependent channel activity. The entry of Ca²⁺, either through cGMP- or voltage-dependent channel, might be expected to potentiate the effects of cGMP.

Evidence presented here and in previous studies clearly shows that there are major functional differences between the cGMP-dependent channel of the On-bipolar cell and CNG channels from photoreceptors and olfactory receptor neurons. In the rod photoreceptor, channel affinity for cGMP is decreased both by Ca²⁺ and phosphorylation, apparently through independent pathways. Phosphorylation by an as yet unidentified kinase(s) decreases the rod channel's affinity for cGMP nearly 10-fold in excised patches (Gordon et al., 1992). In addition raising Ca²⁺ in the presence of calmodulin lowers the channel's cGMP affinity 2-3 fold in both rod outer segment membrane vesicles (Hsu and Molday, 1993) and excised patches (Nakatani et al., 1995). Similar results have been obtained from patches containing an endogenous Ca²⁺-binding protein that is eluted from the patch at low Ca²⁺ concentrations (Gordon et al., 1995). The affinity of the olfactory receptor channel for cAMP has an even more pronounced Ca²⁺-calmodulin dependence (Chen and Yau, 1994; Liu et al., 1994). There also appear to be differences between the On-bipolar channel and CNG channels in their susceptibility to channel block by L-cis-diltiazem (Shiells and Falk, 1992), channel open probability and unitary conductance (de la Villa et al., 1995).

In the retina, synaptic transmission from photoreceptors to On-bipolar cells is mediated through the metabotropic glutamate receptor mGluR6. The binding of glutamate to the postsynaptic mGluR6 receptor leads to the activation of a cGMP phosphodiesterase and the closure of cGMP-dependent cation channels. We have obtained functional evidence that at least two different classes of protein kinase play an essential role in the regulation of this pathway. Using the conventional whole-cell patch-clamp recording technique we have demonstrated that intracellular dialysis with PKI 5-24 amide, a specific inhibitor of cAMP-dependent protein kinase (PKA), selectively abolishes coupling of the receptor to the channel. Up-regulation of the pathway by the inclusion of the catalytic subunit of PKA increases coupling between the receptor and the channel and delays termination of the response. In addition KN-62, a specific inhibitor of calmodulin-dependent protein kinase, selectively inhibits the cGMP-dependent conductance. The depletion of intracellular ATP, the addition of ATPgS, or the inclusion of alkaline phosphatase in the pipette all promote both the uncoupling of the receptor from the channel and the suppression of the cGMP-dependent cation conductance. These results suggest that protein kinases play an essential role in modulating synaptic transmission in the On-bipolar cell. The model in figure 11 illustrates the regulation of the APB receptor pathway by these three protein kinases.

Return to contents

To determine whether P10 bipolar cells in culture expressed the APB receptor cascade.

Rats age post-natal day 10 (P10) were sacrified with sodium pentobarbital and the retinas were dissested into ice-cold media. After trituration neurons were plated onto sterile cover slips and cultured at 37°C in serum-supplemented medium containing BDNF and various other growth factors (methods and culture courtesy of Dr.Scott Nawy). Voltage-clamp recordings were performed using both perforated-patch and conventional whole-cell modes of the patch-clamp technique were made using electrodes of 4-7 MW resistance at various stages after plating (days in vitro). Cells were viewed on an inverted Zeiss light microscope. Data was acquired and analysed as described above.

Bipolar cells identified by their morphology and PKC immunoreactivity showed no evidence of an APB response or of developing inward current when dialyzed with 1mM cGMP during the first 10 days in vitro (DIV, see earlier methods for APB receptor assay). However as shown in Fig.12, in recordings from days 11-20 and 21-30 around 50% of cells exhibited an APB response current suggesting that the combination of growth factors used were sufficient to allow processes to regenerate and to support the functional expression of all elements in the cascade. However, accompanying western blotting experiments were performed which failed to reveal a positive band for mGluR6 in the cultures using a polyclonal antibody provided by Professor S.Nakanishi (all western blotting experiments performed in laboratory of Dr.Zaven Kaprelian). Positive controls were performed with primary antibodies against PKC in culture and mGluR6 in homogenates from frehly dissociated retina. Although mGluR6 positive bands of the appropriate molecular weight were present in the homogenised control retina, the raised polyclonal detected many bands and all cells in P10 and P0 cultures were immunoreactive for the polyclonal (all culture work performed by S.Nawy).

Figure 12. Cultured retinal On-bipolar cells express the APB receptor cascade after 10 days in vitro. Top panel. Averaged response from five 3s applications of 2mM APB in a cell recording obtained from a P10 culture after 10 days in vitro. Bottom panel. Histogram showing number of cells responding (hatched bars) and failing to respond (hollow bars) to APB at different ages of P10 culture (days in vitro). Cells were selected on the basis of morphology.

Return to contents

To determine the glutamate receptor subtypes present in DIV 8-16 P0 retinal ganglion and amacrine cell cultures.

Ganglion and amacrine cells were identified by the transient or robust appearance of regenerative current spikes in response to trains of depolarizing pulses in voltage-clamp mode. All cells tested (n>20) responded to AMPA with a rapid initial desensitization and to kainate with a sustained non-desensitizing response. No cells responded to NMDA or APB under the conditions tested (internal solution contained ATP, GTP, cGMP and NMGCl). Data in possession of Dr.S.Nawy and all P0 cell cultures provided by Dr.S.Nawy and Mr.E.Wexler.

To test the effects of an Alcon NMDA receptor antagonist upon ganglion cell NMDA responses in Ambystoma.

The Alcon inhibitor irreversibly blocked ganglion cell NMDA responses in acute slices (n=3).

To develop a rat retinal slice culture system suitable for testing the efficacy of NMDA receptor antagonists upon hypoxia and kainate-induced cell death in the ganglion cell layer.

Retinal slice culture Retinas from post-natal day 10-15 rats were dissected in ice-cold Na⁺-free choline chloride rat Ringer containing high Mg²⁺ (4mM), low Ca²⁺ (100nM). Retinas were laid ganglion cell layer face down onto filter paper and 100mm slices were cut and submerged into glass-bottomed petri dishes with 1ml of DMEM medium containing 0.5% agarose. The agarose media was maintained at 50°C in a water bath to prevent gelling and allowed to cool for 2 minutes before orientating retinal slices below the surface. Two to four slices were laid per dish under sterile conditions and placed into a 5% CO₂ incubator at 37°C. After 30 minutes the agarose had gelled and 1ml of 10% serum-containing DMEM was pipetted onto the surface of the agarose (no added factors) and replaced every 2 days.

Live-dead stain The surface of the chambers were washed three times for 10 minutes with serum-free DMEM and then incubated for 30 minutes with serum-free DMEM containing 2ml/ml ethidium homodimer and 2mg/ml calcein-AM with or without 200mM kainate. The dishes were then given three five minute washes at 25°C with Dulbecco's phosphate-buffered saline. Slices were viewed under an upright microscope with Hoffmanised optics with appropriate filter sets.

In two separate sets of cultures morphology was retained in 0.5% agarose culture after 4-5 days and few dead cells were observed in any neuronal layer as judged by ethidium homodimer staining. The positive control (calcein-AM) stained all cells brightly. In contrast slices grown on filter supports lost all morphological integrity and laminar structure within 2 days. In conclusion this system provides an ideal method for the live-dead ratio imaging system (Molecular Probes). The system is also suitable for biochemical and cell biological assays. A thirty minute incubation with 200mM kainate resulted in extensive cell death in both the ganglion and inner nuclear cell layers as measured with propidium iodide or ethidium homodimer. Control slices exhibited little cell death as measured by propidium iodide staining. (Photographs taken by E.Wexler and Dr.S.Nawy).

Return to contents

Asano, T., Suzuki, T., Tsuchiya, M., Satoh, S., Ikegaki, I., Suzuki, Y., and Hidaka, H (1989). Vasodilator actions of HA1077 in vitro and in vivo putatively mediated by the inhibition of protein kinase. Br. J. Pharm. 98, 1091-1100.

Benovic, J.L., Pike, L.J., Cerione, R.A., Staniszewski, C., Yoshimasa, T., Codina, J., Caron, M.G., and Lefkowitz, R.J. (1985). Phosphorylation of the mammalian beta-adrenergic receptor by cyclic AMP-dependent protein kinase. Regulation of the rate of receptor dephosphorylation by agonist occupancy and effects on coupling of the receptor to the stimulatory guanine nucleotide regulatory protein. J. Biol. Chem. 260, 7094-7101.

Benovic, J.L., Strasser, R.H., Caron, M.G., and Lefkowitz, R.J. (1986). Beta-adrenergic receptor kinase: identification of a novel protein kinase that phosphorylates the agonist-occupied form of the receptor. Proc. Natl. Acad. Sci. USA 83, 2797-2801.

Bilotta, J., Demarco, P.J. Jr., and Powers, M.K. (1995) The contribution of ON- and OFF-pathways to contrast sensitivity and spatial resolution in goldfish. Vis. Research. 35, 103-108.

Brecha, N.C., Oyster, C.W. and Takahashi, E.S. (1984) Identification and characterization of tyrosine hydroxylase immunoreactive amacrine cells. Invest. Ophthal. and Vis. Sci. 25, 66-70.

Chen, T.Y., and Yau, K.-Y. (1994) Direct modulation by Ca²⁺-calmodulin of cyclic nucleotide-activated channel of rat olfactory receptor neurons. Nature. 368, 545-548.

Chen, Q.X., Stelzer, A., Kay, A.R., and Wong., R.K.S. (1990). GABA_A receptor function is regulated by phosphorylation in acutely dissociated guinea-pig hippocampal neurones. J. Physiol. 420, 207-221.

Cheng, H-C., Kemp, B.E., Pearson, R.B., Smith, A.J., Misconi, L., Van Patten, S.M., and Walsh, D.A. (1986). A potent synthetic peptide inhibitor of the cAMP-dependent protein kinase. J. Biol. Chem. 261, 3989-3992 .

DeVries, S.H., and Schwartz, E.A. (1989). Modulation of an electrical synapse between solitary pairs of catfish horizontal cells by dopamine and second messengers. J. Physiol. 414, 351-375.

Dolan, R.P., and Schiller, P.H. (1994) Effects of ON channel blockade with 2-amino-4-phosphonobutyrate (APB) on brightness and contrast perception in monkeys. Vis. Neurosci. 11, 23-32.

Duvoisin, R.M., Zhang, C., and Ramonell, K. (1995). A novel metabotropic glutamate receptor expressed in the retina and olfactory bulb. J. Neurosci. 15, 3075-3083.

Gordon, S.E., Downing-Park, J., and Zimmerman, A.L. (1995). Modulation of the cGMP-gated ion channel in frog rods by calmodulin and an endogenous inhibitory factor. J. Physiol. 486, 533-546.

Gordon, S.E., Brautigan, D.L., and Zimmerman, A.L. (1992). Protein phosphatases modulate the apparent agonist affinity of the light-regulated ion channel in retinal rods. Neuron 9, 739-748.

Gray-Keller, M. P., Polans, A.S., and Detwiler, P.B. (1993). The effect of ecoverin-like calcium-binding proteins on the photoresponse of retinal rods. Neuron 10, 523-531.

Hayashi, F., Lin, G. Y., Matsumoto, H., and Yamazaki, A. (1991) Phosphatidylinositol-stimulated phosphorylation of an inhibitory subunit of cGMP phosphodiesterase in vertebrate rod photoreceptors. Proc. Natl. Acad. Sci. USA 88, 4333-4337.

Hampson, E. C. G. M., Vaney, D. I., and Weiler, R. (1992) Dopaminergic modulation of gap junction permeability between amacrine cells in mammalian retina. J. Neurosci. 12, 4911-4922.

Hsu, Y.-T., and Molday, R. S. (1993) Modulation of the cGMP-gated channel of rod photoreceptor cells by calmodulin. Nature 361, 76-79.

Kawamura, S. (1993). Rhodopsin phosphorylation as a mechanism of cyclic GMP phosphodiesterase regulation by S-modulin. Nature 362, 855-857.

Kim, D. (1991). Modulation of acetylcholine-activated K⁺ channel function in rat atrial cells by phosphorylation. J. Physiol. 437, 133-55.

Knapp, A.G., and Dowling, J.E. (1987). Dopamine enhances excitatory amino acid gated conductances in cultured retinal horizontal cells. Nature 325, 437-439.

Kuhn, H., Hall S.W., and Wilden, U. (1984). Light-induced binding of 48-kDa protein to photoreceptor membranes is highly enhanced by phosphorylation of rhodopsin. FEBS Letts. 176, 473-478.

Lagnado, L., and Baylor, D. (1992). Signal flow in visual transduction. Neuron 8, 995-1002.

Liu, M., Chen, T-Y., Ahamed, B., Li., J., and Yau, K.-Y. (1994). Calcium-calmodulin modulation of the olfactory cyclic nucleotide-gated cation channel. Science 266, 1348-54.

Maguire G., Lukasiewicz, P., and Werblin, F. (1989). Amacrine cell interaction underlying the response to change in the tiger salamander retina. J. Neurosci. 9, 726-735.

Maguire G., and Werblin, F. (1994). Dopamine enhances a glutamate-gated ionic current in OFF bipolar cells of the Tiger salamander retina. J. Neurosci. 14, 6094-6101.

Masu, M., Iwakabe, H., Tagawa, Y., Miyoshi, T., Yamashita, M., Fukuda, Y., Sasaki, H., Hiroi, K., Nakamura, Y., Shigemoto, R., Takada, M., Nakamura, K., Nakao, K., Katsuki, M., and Nakanishi, S. (1995). Specific deficit of the ON response in visual transmission by targeted disruption of the mGluR6 gene. Cell 80, 757-765.

Mills, S.L., and Massey, S.C. (1995). Differential properties of two gap junction pathways made by AII amacrine cells. Nature 377, 734-737.

Nakajima, Y., Iwakabe, H., Akazawa, C., Nawa, H., Shigemoto, R., Mizuno, N., and Nakanishi, S. (1993). Molecular characterization of a novel retinal metabotropic glutamate receptor mGluR6 with a high agonist selectivity for L-2-amino-4-phosphonobutyrate. J. Biol. Chem. 268, 11868-11873.

Nakatani, K., Koutalos, Y., and Yau, K-W. (1995). Ca²⁺ modulation of the cGMP-gated channel of bullfrog retinal rod photoreceptors. J. Physiol. 484.1, 69-76.

Nawy, S. and Copenhagen, D.R. (1987). Multiple classes of glutamate receptor on depolarizing bipolar cells in retina. Nature 325, 56-58.

Nawy, S., and Jahr, C.E. (1990). Suppression by glutamate of cGMP-activated conductance in retinal bipolar cells. Nature 346, 269-271.

Nawy, S., and Jahr, C.E. (1991). cGMP-gated conductance in retinal bipolar cells is suppressed by the photoreceptor transmitter. Neuron 7, 677-683.

Okamoto, N., Hori, S.H., Akazawa, C., Hayashi, Y., Shigemoto, R., Mizuno, N., and Nakanishi, S. (1994) Molecular characterization of a new metabotropic glutamate receptor mGluR7 coupled to inhibitory cyclic AMP signal transduction. J. Biol. Chem. 269, 1231-1236.

Palczewski , K., and Benovic, J.L. (1991). G-protein-coupled receptor kinases. TIBS 16, 387-391.

Payet, M.D., and Depuis, G. (1992). Dual regulation of the n-type K⁺ channel in Jurkat T lymphocytes by protein kinases A and C. J. Biol. Chem. 267, 18270-18273.

Premont, R.T., Inglese, J., and Lefkowitz, R.J. (1995). Protein kinases that phosphorylate activated G protein-coupled receptors. FASEB 9, 175-182.

Sather, W.A., and Detwiler, P.B. (1987). Intracellular biochemical manipulation of phototransduction in detached rod outer segments. Proc. Natl. Acad. Sci. USA 84, 9290-9294.

Schiller, P.H. (1992). The ON and OFF pathways of the visual system. Trends in Neurosci. 15, 86-92.

Slaughter, M.M., and Miller, R.F. (1981). 2-amino-4-phosphonobutyric acid: a new pharmacological tool for retina research. Science 211, 182-185.

Shiells, R.A., Falk.G., and Naghshineh, S. (1981). Action of glutamate and aspartate analogs on rod horizontal and bipolar cells. Nature 294, 592-594.

Shiells, R.A., and Falk, G. (1990). Glutamate receptors of rod bipolar cells are linked to a cyclic cGMP cascade via a G-protein. Proc. R. Soc. Lond. B. 242, 91-94.

Shiells, R.A., and Falk, G. (1992). Properties of the cGMP-activated channel of retinal on-bipolar cells. Proc. R. Soc. Lond. B. 247, 21-25.

Shiells, R.A., Falk G. (1994). Responses of rod bipolar cells isolated from dogfish retinal slices to concentration-jumps of glutamate. Visual Neuroscience. 11 (6), 1175-83.

Teranishi, T., Negishi, K., Kato, S. (1983). Dopamine modulates S-potential amplitude and dye-coupling between horizontal cells in carp retina. Nature, 243-245.

Tokumitsu, H., Chijiwa, T., Hagiwara, M., Akihiro, M., Terasawa, M., and Hidaka, H. (1990). KN-62, 1-[N,O-Bis(5-isoquinolinesulphonylpiperazine)], a specific inhibitor of Ca²⁺-Calmodulin-dependent protein kinase II. J. Biol. Chem 265, 4315-4320.

de la Villa, P., Kurahashi, T., Kaneko, A. L (1995). Glutamate-induced responses and cGMP-activated channels in three subtypes of retinal bipolar cells dissociated from the cat. J. Neurosci. 15, 3571-3582.

Veruki, M. and Yeh, H. H. (1992). Vasoactive intestinal polypeptide modulates GABA_A receptor function in bipolar cells and ganglion cells of the rat retina. J. Neurophys. 67, 791-797.

Veruki, M. and Yeh, H. H. (1994). Vasoactive intestinal polypeptide modulates GABA_A receptor function through activation of cyclic AMP. Vis. Neurosci. 11, 899-908.

Witkovsky, P., Eldred, W., and Karten, H.J. (1984) Catecholamine- and indoleamine-containing neurons in the turtle retina. J. Comp Neurol. 228, 217-225.

Yamashita, M., and Wassle, H. (1991). Responses of rod bipolar cells isolated from the rat retina to the glutamate agonist 2-amino-4-phosphonobutyric acid (APB). J. Neurosci. 11, 2372-2382.

Yau, K-W., and Nakatani, K. (1985). Light-suppressible cyclic GMP-sensitive conductance in the plasma membrane of a truncated rod outer segment. Nature 317, 252-255.

Return to retina home page
Return to contents