Differentially Distributed IP₃ Receptors and Ca²⁺ Signaling in Rod Bipolar Cells

Abstract

purpose. Inositol (1,4,5)-trisphosphate receptors (IP₃Rs) contribute substantially to cytosolic free calcium ion (Ca²⁺) concentration transients and thereby modulate neuronal function. The present study was undertaken to determine the contribution of IP₃Rs to the function of rod bipolar cells in the retina.

methods. Immunoreactivity for IP₃Rs in rod bipolar cells from mouse retinas was detected by immunocytochemical methods. Intracellular Ca²⁺ concentrations were optically recorded in acutely isolated rod bipolar cells, and biophysical properties of IP₃Rs were analyzed with single channel electrophysiology.

results. The distribution of IP₃R isoforms was correlated with cytosolic Ca²⁺ transients induced by activation of group I metabotropic glutamate receptors (mGluRs) and with biophysical properties of differentially expressed IP₃Rs.

conclusions. The differential distribution of IP₃Rs is used by rod bipolar cells to convey Ca²⁺ signals that are distinct in their duration, amplitude, and kinetics at the subcellular level, and that serve the functions of individual subcellular compartments. IP₃R-mediated Ca²⁺ signaling indicates a potential mechanism for the adaptation of the ON-pathway of vision and for coincidence and threshold detection in retinal neurons.

In many neurons, transients in the intracellular Ca²⁺ concentration mediated by release of Ca²⁺ from intracellular stores functionally depend on the activity of inositol 1,4,5-trisphosphate (IP₃)-gated Ca²⁺ channels/ IP₃ receptors (IP₃R). Production of IP₃ through the stimulation of plasma membrane receptors and activation of the phospholipase C (PLC) pathway leads to the activation of IP₃Rs that release Ca²⁺ from intracellular Ca²⁺ stores. ² This second messenger signaling system has effects on global Ca²⁺ signaling as well as on local, spatially defined Ca²⁺ signals mediating a variety of physiological processes in many cells and particularly in neurons. ^{3 4 5 6}  

Peng et al. ⁷ have described the localization in the mammalian retina of type 1 IP₃R in synaptic terminals of photoreceptors and bipolar cells, as well as in amacrine cell processes, indicating a possible role in neurotransmission. Wang et al. ⁸ localize type 1 IP₃Rs to outer segments of cone photoreceptors and conclude a possible role of IP₃Rs in photoreceptor function. In vertebrate retina, IP₃Rs had been identified in photoreceptor cells and plexiform layers, ⁹ horizontal cells, bipolar cells, and Müller glia ¹⁰ and had been implicated predominantly in neuronal and Müller cell function. ¹¹ Indirect evidence from a number of studies indicates that IP₃Rs potentially play important roles in regulating neuronal function in the retina by affecting physiological processes governed by transients in the intracellular Ca²⁺ concentration. ^{12 13 14 15 16 17 18 19 20}  

Bipolar cells are the first interneurons in the glutamatergic vertical pathway of retinal information processing, and integrate signals in the two synaptic layers of the retina. ^{21 22} Besides evidence for the presence of type 1 IP₃R in these cells ^{7 8 9 10 11} and modulation of neurotransmitter release by transients in the intracellular Ca²⁺ concentration, ^{23 24 25 26 27 28 29 30} the function of intracellular Ca²⁺ channels in bipolar cell physiology, especially in the outer retina, remains elusive. The present study addresses this issue by investigating the distribution of the different types of intracellular Ca²⁺ channels and by analyzing IP₃R-mediated Ca²⁺ signaling and biophysical properties of individual IP₃Rs in an anatomically and physiologically well-characterized type of bipolar cells, rod bipolar cells. These cells express group I metabotropic glutamate receptors, ^{1 31} a group of G-protein-coupled glutamate receptors, that are coupled to stimulation of PLC and generation of IP₃ ³² . Rod bipolar cells also have physiologically relevant PLC activity ³³ and IP₃R-dependent Ca²⁺ stores. ^{7 9 10 11}  

Materials and Methods

All experiments described in the present study were carried out in accordance with the appropriate National Institutes of Health and University of North Texas Health Science Center Guidelines for the Welfare, Care and Use of Experimental Animals and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Isolation and Immunochemistry of Rod Bipolar Cells

Rod bipolar cells were isolated from retinas of outbred albino Swiss Webster mice (Harlan, Indianapolis, IN) by mechanical and enzymatic dissociation as described previously, and were identified using morphologic and/or immunochemical criteria. ^{34 35} Freshly isolated rod bipolar cells, which maintain not only their substructural morphology but also important physiological functional properties during enzymatic and mechanical dissociation, ^{68 69 70 71 72 73} were used in the study. To show preservation of subcellular protein distribution after enzymatic and mechanical dissociation of rod bipolar cells, we used stainings of metabotropic glutamate receptors 1 and 5, two upstream elements of IP₃R-mediated signaling in rod bipolar cells, to verify that the dissociation process did not lead to rearrangements of proteins (see Figs. 2G and 2H ). After plating on coverslips coated with poly-d-lysine, cells were processed for either immunocytochemistry or optical imaging of intracellular Ca²⁺ concentrations. Immunocytochemistry was performed as described previously. ^{34 35} Briefly, cells were fixed in 4% para-formaldehyde (w/v) in phosphate buffer and incubated in blocking, primary and secondary antibody solutions 2 hours each. Primary antibodies were used at concentrations identified below. Goat anti-mouse or anti-rabbit IgG labeled with Alexa Fluor 488 and Alexa Fluor 594 (diluted 1:500; Molecular Probes, Inc., Eugene, OR) were used to visualize immunoreactivity using standard immunofluorescence microscopy techniques. ^{34 35}  

For Western blot analyses and isolation of endoplasmic reticulum (ER) from bipolar cells, ON-bipolar cells which comprise both ON-cone and rod bipolar cells of which ON-cone bipolar cells make up approximately 36% and rod bipolar cells approximately 64% ^{36 37} were isolated by immunopanning. After dissociation of retinas, cells were incubated in vials (Corning, Acton, MA) coated with polyclonal antibody to the N-terminal extracellular region of metabotropic glutamate receptor 6 (mGluR6; rabbit polyclonal, 1:500; Novus Biologicals, Littleton, CO). mGluR6-positive ON bipolar cells that adhered to the vials were collected after centrifugation at 500g and used for Western blot analyses (WB) or ER isolation. Immunoreactivity for intracellular Ca²⁺ channels was assayed with the standard Western blotting technique. ¹ Primary antibodies were used at concentrations identified below. 

Antibodies

Intracellular Ca²⁺ channels were detected with isoform-specific antibodies: type 1 IP₃R (EMD Biosciences, La Jolla, CA) rabbit polyclonal, 1:1000 for WB, 1:100 for immunocytochemistry (ICC) ³⁸ ; type 2 IP₃R (EMD Biosciences) rabbit polyclonal, 1:1000 for WB, 1:100 for ICC ³⁹ ; type 3 IP₃R (BD Biosciences Pharmingen, San Diego, CA) mouse monoclonal, 1:1000 for WB, 1:100 for ICC ⁴⁰ ; pan RyR (Affinity BioReagents, Golden, CO) mouse monoclonal, 1:1000 for WB, 1:50 for ICC. ⁴¹ Additionally, specific antibodies were used to immunolabel mGluR1 (Chemicon, Temecula, CA; rabbit polyclonal, 0.2 μg/mL for ICC) ¹ mGluR5 (Chemicon; rabbit polyclonal, 1 μg/mL for ICC) ¹ to isolate ON bipolar cells using mGluR6 immunoreactivity (Novus Biologicals; rabbit polyclonal, 1:500 for immunopanning), and to characterize rod bipolar cells using protein kinase Cα immunoreactivity (Seikagaku, Tokyo, Japan; mouse monoclonal, 1:100 for ICC). ³⁵  

Optical Imaging of Intracellular Ca²⁺ Concentrations

Calcium imaging was performed as described previously. ^{34 42 43} Briefly, after the isolation of retinal neurons, cells were transferred to L-15 medium (Leibovitz medium; Sigma-Aldrich, St. Louis, MO) and were incubated in 4 μM cell permeant fluo-3 (fluo-3-acetoxymethylester; Molecular Probes) at 37°C for 30 minutes, washed, and recorded for up to 6 hours after dissociation. Changes in fluorescence intensity of the Ca²⁺ indicator fluo-3 in loaded cells were measured over time with time-lapse videomicroscopy (Olympus IX70, Olympus, Japan; Hamamatsu ORCA-ER, Hamamatsu, Japan; Lambda DG-4 Ultra High Speed Wavelength Switcher with appropriate filter sets; Sutter Instrument Company, Novato, CA; SimplePCI Imaging Software v. 5.2; Compix Inc., Imaging Systems/Hamamatsu Photonics Management Corporation, Bridgewater, NJ). Cells received fresh L-15 medium constantly using a gravity-fed perfusion system with a flow rate of 1 mL/min. Images were acquired every 0.5 second and Ca²⁺ transients were calculated as the ratio of the fluorescence intensity during drug application (F) over the average baseline fluorescence intensity 10 s before drug application (F/F ₀). Subcellular regions of interest of rod bipolar cells were analyzed independently and their spatiotemporal patterns of Ca²⁺ transients were analyzed separately. 

During the calcium imaging experiments, the group I metabotropic glutamate receptor agonist (S)-3,5-dihydroxyphenylglycine (S-DHPG) ^{44 45} (Sigma-Aldrich) was applied to the imaging chamber at pharmacologically relevant concentrations between 0.1 and 250 μM in a 0.5 seconds bolus application. Similarly, thapsigargin (10 μM; EMD Biosciences), an inhibitor of sarcoplasmic and endoplasmic reticulum Ca²⁺ ATPases, ⁴⁶ was bath-applied to release Ca²⁺ from intracellular stores at the end of experiments to test the functionality and Ca²⁺ levels of intracellular Ca²⁺ stores. In several experiments, 100 μM dantrolene and/or 1 μM xestospongin C (Sigma-Aldrich) were included in the perfusate to block ryanodine receptors and/or IP₃Rs, respectively. To test the dependence of measured Ca²⁺ transients on the extracellular Ca²⁺ concentration, Ca²⁺-free L-15 medium containing 5 mM EGTA to buffer trace amounts of free Ca²⁺ was used instead of regular L-15 medium. When cells were tested in the Ca²⁺-free medium environment they were equilibrated in Ca²⁺-free medium for 2 minutes before the application of mGluR1 agonists. No significant depletion of Ca²⁺ stores occurred during this time. 

Three or more independent experiments analyzing at least 5 cells each were performed to obtain data for each experimental condition, and statistical analyses used standard one-way or multiple ANOVA for comparisons of parametric populations. 

Single Channel Electrophysiology

Intracellular calcium channels, present in ER vesicles prepared from ON-bipolar cells that had been isolated with mGluR6 immunopanning, were measured as described previously ^{42 43} after incorporation into planar lipid bilayers. Bilayers had a 250 mM HEPES-Tris solution, pH 7.35 on the cytosolic side and a 250 mM HEPES, 55 mM Ba(OH)₂ solution, pH 7.35 on the ER lumen side of the channel. The identity of IP₃Rs was verified by activation with its ligand IP₃ and inactivation by 50 mg/L heparin or 1 μM xestospongin C. No RyRs were observed in the ER preparations. The activity of IP₃Rs was monitored over a range of cytosolic-free Ca²⁺ and IP₃ concentrations. Channel activity was recorded under voltage-clamp conditions, filtered at 3 kHz and digitized using a planar lipid bilayer workstation (Warner Instruments, Inc., Hamden, CT). Data acquisition and analysis were carried out with pClamp version 8.1 (Axon Instruments, Union City, CA) identifying mean dwell times, current amplitudes, and open probability (P _o). The identity of IP₃R subtypes was determined using known biophysical parameters particularly the channel activity dependence on cytosolic IP₃ and free Ca²⁺ concentrations ^{52 53} (reviewed in Refs. ² , ⁵⁴ , ⁵⁵ ). Based on these properties the group of IP₃Rs with high IP₃ sensitivity was identified as IP₃R2 and the group of IP₃Rs with intermediate IP₃ sensitivity as IP₃R1. 

The data for each experimental condition were obtained from three or more independent experiments, and statistical analyses used standard one-way or multiple ANOVA for comparisons of parametric populations. 

Results

In the present study, the distribution and function of intracellular calcium channel isoforms in rod bipolar cells of the mouse retina were investigated with specific antibodies, optical imaging of intracellular Ca²⁺ concentrations, and electrophysiology. We identified a correlation between the isoform-specific differential distribution of intracellular calcium channels and their isoform-specific biophysical and cellular Ca²⁺ signaling properties in neurons. Our results indicated that the differential distribution of intracellular Ca²⁺ channel isoforms with different biophysical properties can be used by neurons to produce cellular signaling patterns as described for other cell types. ^{47 48 49 50 51} Such functional distribution patterns in neurons potentially provide a basis for compartment-specific intracellular signaling mechanisms that convey temporally and spatially distinct signaling patterns. 

In mGluR6-immunopurified ON-bipolar cells (see Materials and Methods), which comprise both ON-cone and rod bipolar cells (approximately 36% and 64%, respectively ^{36 37} ), Western blot analyses showed the expression of two IP₃R isoforms, types 1 and 2 IP₃R (Fig. 1) . Both antibodies against type 3 IP₃R and RyRs showed no significant signals when compared to mouse control tissues (Fig. 1) . These results from immunoblotting experiments were confirmed by immunocytochemistry. Immunofluorescence staining of acutely isolated rod bipolar cells showed specific label with signals above control levels (Figs. 2A and 2B)only for type 1 and 2 IP₃R (Figs. 2C and 2D) . The stainings for type 1 IP₃R immunoreactivity indicate that type 1 IP₃R can be found throughout the entire rod bipolar cell. Highest levels of immunoreactivity, however, were found in somata and axon terminals (Fig. 1C) . In contrast, immunoreactivity for type 2 IP₃R was restricted to the dendrites and only very faint immunofluorescence label for type 2 IP₃R immunoreactivity was found in the distal portion of the rod bipolar cell soma (Fig. 1D) . 

When rod bipolar cells were stained with antibodies against type 3 IP₃R and RyRs that had been used for immunocytochemistry successfully in previous publications, ^{40 41} no specific immunofluorescence signals were detected (Figs. 2E and 2F) . However, with immunolabeling of acutely isolated rod bipolar cells we could corroborate findings from a previous study ¹ that had identified group I mGluRs in rod bipolar cells postsynaptically to rod photoreceptor cells using vertical cryosections of the retina and ultrastructural immunolocalization. Immunoreactivity for mGluR1 and mGluR5, two upstream elements of IP₃R-mediated signaling in rod bipolar cells, was found in the dendrites and distal one-quarter to one-third of rod bipolar cell somata (Figs. 2G and 2H) , consistent with synaptic release sites for the ligand of group I mGluRs, glutamate, by rod photoreceptor cells in the outer plexiform layer ¹ and, at the same time, indicating the major sites of IP₃ generation in the distal portion of rod bipolar cells. 

Based on these findings of the expression of specific isoforms of intracellular Ca²⁺ channels by rod bipolar cells (Fig. 1)and of the differential subcellular distribution of these isoforms (Fig. 2) , we hypothesized that stimulation of IP₃Rs would produce functionally different Ca²⁺ signaling patterns dependent on the amount of IP₃ generated by group I mGluR activation. 

Intracellular Ca²⁺ concentrations were recorded optically in rod bipolar cells after ester loading with fluo-3 in the absence of extracellular Ca²⁺. Changes in fluo-3 emission wavelength maximum as an indicator of changes in intracellular calcium concentrations were monitored. Fluorescence intensity of the Ca²⁺ indicator dye fluo-3 was normalized to the baseline intensity and was used as a measure of relative changes in the intracellular Ca²⁺ concentration and displayed as intensity change over time. Freshly isolated rod bipolar cells were stimulated with bolus applications of a specific agonist of group I mGluR, S-DHPG, to induce IP₃ generation in the distal portion of rod bipolar cells (Fig. 3) . Figure 3Ashows an image montage of a representative rod bipolar cell response to S-DHPG stimulation. Low concentrations of the IP₃-generating agonist (0.1–10 μM S-DHPG) produced a temporally and spatially well-defined increase in the cytosolic Ca²⁺ concentration that was restricted to the dendrites (Fig. 3B , black trace). The mean maximum amplitude at 10 μM S-DHPG was 9 ± 3% with an average signal duration of 7 ± 3 seconds (mean ± SEM; n = 17). No signals were observed in the soma at low agonist concentrations. At higher agonist concentrations (50–250 μM S-DHPG), Ca²⁺ transients with distinct spatial and temporal characteristics could be observed in dendrites and soma (Figs. 3A and 3B) . Whereas Ca²⁺ transients in the dendrites immediately followed the stimulus (Fig. 3B , gray trace), somata showed a delayed onset response of 22 ± 4 seconds (mean ± SEM; n = 17; Fig. 3B , gray trace). Ca²⁺ concentrations in both compartments returned to baseline levels during constant perfusion of the cell with medium. The mean maximum amplitude at 100 μM S-DHPG in the dendrites was 11 ± 4% with an average signal duration of 82 ± 7 seconds (mean ± SEM; n = 16). In the soma, mean maximum amplitude at 100 μM S-DHPG was 15 ± 3% with an average signal duration of 104 ± 9 seconds (mean ± SEM; n = 16). All measurements were obtained in Ca²⁺-free medium excluding a contribution of ligand- or store-operated Ca²⁺ channels located on the plasma membrane to the Ca²⁺ transients. The absence of the contribution of extracellular Ca²⁺ to the cytosolic Ca²⁺ transients through Ca²⁺ channels of the plasma membrane and the receptor-specific stimulation of group I mGluRs enabled us to focus on the spatio-temporal patterns of IP₃-induced intracellular Ca²⁺ release via mGluR1 in rod bipolar cells. 

Based on our findings from the immunochemistry experiments, we hypothesized that the observed Ca²⁺ transients induced by group I mGluR stimulation are mediated by IP₃R and not RyR. Preincubation and perfusion of rod bipolar cells for 5 minutes before and during group I mGluR stimulation with 100 μM dantrolene to inactivate RyRs ² did not alter signal amplitude or duration, whereas addition of 1 μM xestospongin C ² to the perfusion medium for 5 minutes before and during S-DHPG application completely abolished agonist-induced Ca²⁺ transients. Based on the same findings from the immunochemistry experiments, we further hypothesized the existence of IP₃Rs with distinct biophysical properties in rod bipolar cells that can be attributed to the differential expression of both types 1 and 2 IP₃R. To test this hypothesis, we analyzed single channel electrophysiological properties of intracellular calcium channels isolated from mGluR6-immunopurified ON-bipolar cells. Intracellular calcium channels with RyR characteristics (activation by Ca²⁺, cyclic ADP ribose, caffeine, and block by ryanodine, ruthenium red ² ) were not observed, whereas two populations of IP₃R distinct in their activity dependence on IP₃ concentrations were identified (Fig. 4) . A range of physiologically relevant concentrations of IP₃ on the cytoplasmic side of the channel was tested to evaluate activation of the channels by their ligand. Figures 4A and 4Bshow representative recordings from IP₃Rs with high and intermediate IP₃ sensitivity, respectively. Dose–response analyses showed that IP₃Rs with high IP₃ sensitivity had an EC₅₀ of 63 ± 4 nM and IP₃Rs with an intermediate IP₃ sensitivity had an EC₅₀ of 196 ± 7 nM. Based on previously published data on the IP₃ dependence of IP₃Rs, ^{52 53} reviewed in Refs. ⁵⁴ , ⁵⁵ , we tentatively identified the group of IP₃Rs with high IP₃ sensitivity as IP₃R2 and the group of IP₃Rs with intermediate IP₃ sensitivity as IP₃R1 (Figs. 4C and 4D) . 

Discussion

In neurons, intracellular Ca²⁺ signaling is critically determined by IP₃R-mediated release of Ca²⁺ from intracellular Ca²⁺ stores. ⁴ Especially for retinal neurons, the second messenger substance IP₃ has been implicated indirectly in several studies as an important physiological component of intracellular and neuronal signaling. ^{12 13 14 15 16 17 18 19 20} In bipolar cells that integrate signals in the two synaptic layers of the retina, ^{21 22} as the first interneurons in the glutamatergic vertical pathway of retinal information processing changes in the intracellular Ca²⁺ concentration have been shown to control neurotransmitter release. ^{23 24 25 26 27 28 29 30} Specifically, rod bipolar cells express group I metabotropic glutamate receptors, ^{1 31} and have physiologically relevant PLC activity ³³ and IP₃R dependent Ca²⁺ stores. ^{7 9 10 11} Therefore, these findings of IP₃-mediated Ca²⁺ signaling via group I mGluRs in rod bipolar cells set the stage for experimental analyses presented in the present study. Our findings indicated that, depending on the strength of glutamatergic input, more specifically group I mGluR activation, Ca²⁺ transients with differential varying temporal and spatial properties ensue. Using immunoblotting, immunocytochemistry, optical imaging of Ca²⁺ concentrations, and single channel electrophysiology, we identified the presence and differential distribution of types 1 and 2 IP₃R and potentially excluded the functional expression of other isoforms of intracellular Ca²⁺ channels in rod bipolar cells. 

Freshly isolated rod bipolar cells maintain not only their substructural morphology but also important physiological functional properties during enzymatic and mechanical dissociation. ^{68 69 70 71 72 73} Therefore, they were used as model systems in the present study to allow a detailed analysis of IP₃R localization and function in mammalian rod bipolar cells. Control experiments using the immunolocalization of metabotropic glutamate receptors 1 and 5 (Figs. 2G and 2H)indicate that the native subcellular protein distribution ^{1 69} was preserved after enzymatic and mechanical dissociation of rod bipolar cells and did not lead to rearrangements of proteins. 

These findings corroborated data related to the immunolocalization of type 1 IP₃R ^{7 9 10 11} and expanded these reports to include the localization of type 2 IP₃R. But more importantly, they incorporate functional properties as well as mechanisms of action of IP₃R-mediated Ca²⁺ signaling in rod bipolar cells. Our results also supported the notion that the differential localization of functionally distinct isoforms of intracellular Ca²⁺ channels determines cellular signaling patterns and functions. ^{47 48 49 50 51} The results also identify IP₃R-mediated signaling initiated by glutamatergic neurotransmission as a potential mechanism of action in rod bipolar cells. We determined that the differential distribution of IP₃R isoforms influences group I mGluR and IP₃-mediated Ca²⁺ signaling in mouse rod bipolar cells and might play a critical role in the modulation of signaling in these first interneurons in the glutamatergic vertical pathway of retinal information processing. 

Low agonist and therefore low cytosolic IP₃ concentrations would preferentially activate IP₃R isoforms with a high affinity for IP₃ ^{52 54 55} in the dendrites (Figs. 2D 3B 4A 4C 4D) . In contrast, higher agonist and therefore higher cytosolic IP₃ concentrations would also be able to activate IP₃R isoforms with a lower affinity for IP₃, ^{52 54 55} that in rod bipolar cells are found throughout the cell, but predominantly in the soma and the axon terminal (Figs. 2C 3 4B 4C 4D ). These correlations between localization of receptor isoforms and their functional and biophysical properties resulting in specific Ca²⁺ signaling patterns are further supported by the diffusion properties of IP₃ ⁵⁶ and spatial constraints of the group I mGluR ¹ and IP₃R signaling system (Figs. 2C 2D 2G 2H) . Allbritton and colleagues ⁵⁶ identified IP₃ as a global intracellular messenger measured in cytosolic extracts with a fast diffusion coefficient of 283 μm² and a long, effective range of 24 μm taking diffusion coefficient and lifetime into account. Therefore, even if one assumes that IP₃ is exclusively being generated at the tips of rod bipolar cell dendrites, which is not supported by the group I mGluR immunolocalization data ¹ (Figs. 2G 2H)showing extrasynaptic expression of group I mGluRs, IP₃Rs in the soma would still be activated if they had a similar affinity to IP₃ as IP₃Rs in the dendrite, (i.e., the same IP₃R isoforms in both soma and dendrites). However, the existence of IP₃Rs with different affinities for IP₃ in the soma and the dendrites (Figs. 2C 2D 4)allows rod bipolar cells to discriminate the strength of incoming glutamate/IP₃ signals (Fig. 3) . Recent studies show that mGluR 1 function in addition to binding of the receptor to l-glutamate also depends on the extracellular calcium concentration. ^{74 75 76} The present study used native group I mGluRs as an indirect means of stimulating IP₃ production in rod bipolar cells, allowing not only the control of the amount but more importantly the location of the IP₃ release better than other current pharmacological or cell biological techniques, and at the same time, limiting the observed effects to a ligand-specific interaction and subsequent second-messenger mobilization. Even though high S-DHPG doses used in the present study produced maximal receptor activity, future studies investigating the involvement of changes in extracellular calcium concentrations need to address modulating effects of extracellular calcium concentrations on the results of the present study. 

In ON-bipolar cells, glutamate released by photoreceptor cells produces hyperpolarization by binding to mGluR6, which closes a cGMP-gated cation channel on the plasma membrane. ⁵⁷ As reported previously ^{58 59 60} (reviewed in Ref. ⁶¹ ), the cytosolic Ca²⁺ concentration is critical for the regulation of this cation channel and thereby the mGluR6-mediated adaptation of the ON-pathway of visual neurotransmission to changing light levels. Results of the present study potentially indicate that the group I mGluR-mediated effects of glutamate could modulate the mGluR6 pathway through regulation of the cytosolic Ca²⁺ concentration ^{58 59 60} (reviewed in Ref. ⁶¹ ), and potentially adapt the sensitivity of rod bipolar cells to changing light levels. 

In addition to this potential function of the group I mGluR/IP₃R pathway, other reports also support the notion of group I mGluR activation resulting in hyperpolarization via activation of Ca²⁺-activated K⁺ channels as a result of IP₃R mediated Ca²⁺ signaling ^{62 63} (reviewed in Ref. ³² ), which is relevant in light of reported Ca²⁺-activated K⁺ channel activity in bipolar cells. ⁶⁴ However, it should be noted that depending on the system and physiological environment, other functions of group I mGluR including neuronal depolarization are also being discussed. ³²  

Similar to processes observed in other portions of the CNS, ^{65 66 67} the expression of a low affinity IP₃R in the soma of rod bipolar cells could also serve as a coincidence and threshold detector for elevated levels of IP₃ in the cytosol, providing signal integration in the soma. In contrast, the expression of a high affinity IP₃R in the dendrites of rod bipolar cells would help to maintain regulation of synaptic events as discussed above for a potential involvement in mGluR6-mediated signaling. ^{58 59 60 61} In summary, rod bipolar cells can use the differential distribution of IP₃R isoforms to produce spatio-temporally distinct cytosolic Ca²⁺ signals that contribute to Ca²⁺-dependent functions of subcellular compartments. 

 Supported in part by a German National Merit Scholarship Foundation student scholarship (CM), by the University of North Texas Health Science Center at Fort Worth Intramural Research Program (PK), a Young Investigator Award from The National Alliance for Research on Schizophrenia and Depression (PK), National Institutes of Health/National Institute on Aging Grant AG022550 (PK), and National Institutes of Health/National Eye Institute Grant EY014227 (PK).

 Submitted for publication August 3, 2004; revised August 25, 2004; accepted October 3, 2004.

 Disclosure: P. Koulen, None; J. Wei, None; C. Madry, None; J. Liu, None; E. Nixon, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 2.

 

Localization of IP₃R and RyR immunoreactivity in mouse rod bipolar cells. Acutely isolated rod bipolar cells were fixed and stained for type 1 (C), type 2 (D), and type 3 (E) IP₃R immunoreactivity, as well as for RyR immunoreactivity (F). Controls include secondary antibody controls for both types of secondary antibodies used [anti-mouse (A); anti-rabbit (B)], as well as labeling of mGluR1 (G) and mGluR5 (H) immunoreactivity. For each panel, differential interference contrast images show the typical rod bipolar cell morphology with dendritic tree, soma axon, and axon terminal system.

Figure 1.

 

Western blot showing the expression of IP₃Rs and RyRs in ON-bipolar cells (ONBC) and controls (C). Homogenates of enzymatically isolated mouse rod bipolar cells were analyzed. Mouse cerebellum, liver, pancreas, and brain homogenates were loaded as controls for IP₃R type 1, IP₃R type 2, IP₃R type 3, and RyRs, respectively. Twenty micrograms total protein per lane were loaded on 4%–12% gradient gels to detect individual intracellular calcium channels. IP₃R isoform specific antibodies detected approximately 250 kDa bands, the pan-RyR antibody detected a high molecular weight band of approximately 550 kDa (numbers and arrows, position and size of protein standards in kDa, respectively). Only IP₃Rs type 1 and 2 could be detected in RBCs.

Figure 3.

 

Group I mGluR agonists induce spatiotemporally differential Ca²⁺ transients in isolated mouse rod bipolar cells. (A) Montage of a representative imaging experiment visualizing the cytosolic Ca²⁺ concentration at 0 seconds, during 10 and 50 seconds, and after 200 seconds, application of 100 μM S-DHPG. A differential interference contrast image to the right shows the main subcellular regions of the rod bipolar cell, dendritic tree (D), soma (S), axon (A), and axon terminal system (AT). Although a Ca²⁺ transient in the dendrites was observed immediately after stimulation (10 seconds), a Ca²⁺ transient with larger amplitude was seen in the soma after a delay and after the peak of the dendritic Ca²⁺ transient (50 seconds). (B) Changes in the intracellular Ca²⁺ concentration in two regions of a representative rod bipolar cell, dendrites (black) and soma (gray). After a single 0.5 second bolus application of a low dose of an mGluR1 agonist (10 μM S-DHPG, indicated by arrow to the left), only the dendrites responded with a Ca²⁺ transient, whereas a higher concentration of mGluR1 agonist (100 μM S-DHPG, 0.5 second bolus application indicated by the arrow to the right) induced Ca²⁺ transients in both dendrites and soma.

Figure 4.

 

Two types of IP₃R with respect to their IP₃ sensitivity can be isolated from the ER of mouse rod bipolar cells. Activity of individual IP₃Rs was recorded with 0.5 μM free Ca²⁺, 500 μM ATP as IP₃R co-agonists, 10 μM ruthenium red to block RyRs present on the cytoplasmic side of the channel and Ba²⁺ as the current carrier on the luminal side of the channel. Bars to the right of each trace indicate zero current levels of channel activity and downward deflections indicate channel openings. (A) and (B) Electrophysiological recordings of two representative experiments with three traces at 0.01, 0.1, and 1 μM IP₃, exemplifying the two types of IP₃Rs that could be observed in mouse rod bipolar cells. (C) and (D) summarize the IP₃ dependence of the absolute and normalized open probability (P _o), respectively, for the two types of IP₃Rs isolated from mouse rod bipolar cells. Based on the biophysical properties, the group of IP3Rs with high IP3 sensitivity is identified as IP3R2 (type 2 IP₃R; n = 9) and the IP₃R with intermediate IP₃ sensitivity as IP₃R1 (type 1 IP₃R; n = 11).

The authors thank Margaret, Richard, and Sara Koulen for generous support and encouragement.