Retinal Voltage-Dependent Anion Channel: Characterization and Cellular Localization

Abstract

purpose. To characterize and localize retinal voltage-dependent anion channel (VDAC) and to understand its possible contribution to mitochondrial function and dysfunction.

methods. VDAC was characterized by a method involving purification from isolated mitochondria and reconstitution into a planar lipid bilayer (PLB). The permeability transition pore (PTP) was monitored by Ca²⁺ accumulation in isolated mitochondria and swelling of mitochondria. Localization was studied by immunocytochemistry and in situ hybridization.

results. Retinal VDACs exhibited the electrophysiological fingerprint of the VDAC superfamily. It had a maximal chord conductance of 3.7 ± 0.1 nanosiemens (nS) in 1 M NaCl, and a voltage-dependent conductance that was highest at transmembrane potential close to zero. It was modulated by glutamate, which decreased the channel’s open probability, and by La³⁺ and ruthenium amine binuclear complex (Ru360), which closed the channel. Energized and freshly prepared retinal mitochondria accumulated Ca²⁺ that is inhibited by La³⁺ ruthenium red and Ru360. Subsequent to Ca²⁺ accumulation, mitochondria released the accumulated Ca²⁺, probably through activation of the PTP. Ru360 inhibited Ca²⁺ release and mitochondrial swelling. VDAC was present in mitochondria of all retinal cell types: photoreceptor, bipolar, horizontal, amacrine, and ganglion cells. Most cells primarily expressed VDAC-1, but they also expressed VDAC-2 and -3.

conclusions. These results suggest that VDAC is involved in PTP activity and/or regulation and thus is an important player in retinal degeneration associated with PTP-mediated mitochondrial dysfunction.

Mitochondria are essential for cellular metabolism, energy production, modulation of cell redox potential, osmotic regulation, pH control, and calcium homeostasis. ^{1 2} Therefore, it is not surprising that mitochondrial dysfunction can cause cell death through depletion of adenosine triphosphate (ATP) and deregulation of Ca²⁺. ^{3 4 5 6} Involvement of mitochondria in apoptosis is associated with the opening of the permeability transition pore (PTP) induced by a variety of conditions, such as elevated matrix Ca²⁺, oxidants, and thiol reagents. This leads to dissipation of the membrane potential (ΔΨ) and release of mediators of apoptosis such as cytochrome c, apoptosis-inducing factor (AIF), and procaspase. ⁶  

The inner mitochondrial membrane contains numerous α-helical transport proteins, whereas the outer membrane contains only the mitochondrial β-strand porin, also called VDAC (voltage-dependent anion channel). Thus, the mitochondrial outer membrane provides a barrier between the mitochondrial matrix and the cell cytoplasm. VDAC mediates complex interactions between mitochondria and other parts of the cell by transporting anions, cations, ATP, and certain metabolites. ^{7 8 9 10 11 12}  

Three human VDAC cDNA clones with 75% to 94% homology have been identified. ^{13 14} Computer modeling of the primary amino acid sequence predicts that VDAC consists of a single α-helix and 12 to 16 transbilayer β-strands. ^{9 15 16} These β-strands are connected by several peptide loops of different sizes on both sides of the membrane. 

Purified VDAC incorporated into planar lipid bilayer (PLB) forms a large voltage-gated pore (∼3 nm) that can exist in multiple conformational states with different ionic selectivity and permeability. ^{9 10 11 12 13} In recent years, evidence has emerged that VDAC in the outer mitochondrial membrane is normally closed and highly regulated by metabolites, substrates, nucleotides, and associated proteins. VDAC has binding sites for hexokinase, ^{17 18} creatine kinase, ¹⁸ benzodiazepine receptor, ¹⁹ adenine nucleotide translocator (ANT), ²⁰ and the outwardly rectifying depolarization-induced chloride channel (ORDIC). ¹⁴ Reduced nicotinamide adenine dinucleotide (NADH), ²¹ and ATP ²² may also regulate VDAC. Recently, we have shown that VDAC possesses a Ca²⁺ binding site, is permeable to Ca²⁺, ²³ and is modulated by glutamate. ²⁴  

The retina has a very high metabolic demand with one of the highest rates of oxygen consumption. ^{25 26} Approximately 55% to 65% of retinal mitochondria are located in the photoreceptor’s inner segment, which has an oxygen consumption two times greater than other retinal regions. ²⁷ It is now clear that mitochondria in the photoreceptor’s inner segment are involved in retinal degeneration due to Ca²⁺ overload that induces PTP formation. ²⁸ To further understand factors and events that may contribute to retinal degeneration, we localized VDAC in the retina, purified it, characterized its channel activity, and studied its involvement in Ca²⁺ transport and PTP formation. 

Materials and Methods

Materials

Tris, HEPES, asolectin, phenylmethyl sulfonyl fluoride (PMSF), leupeptin, BSA, Triton X-100, Nonidet P-40, and reactive red-agarose were purchased from Sigma Chemical Co. (St. Louis, MO). Anti-VDAC was a mouse monoclonal that was prepared against the N terminus of 31HL human porin ²⁹ (clone no. 173/045; cat. no. 529538; Calbiochem, La Jolla, CA). Alkaline phosphatase-conjugated goat anti-mouse IgG was obtained from Promega (Madison, WI), and horseradish peroxidase (HRP)-conjugated anti-mouse from Protos Immunoresearch (San Francisco, CA). Hydroxyapatite (Bio-Gel HTP) was purchased from Bio-Rad (Richmond, CA), and Celite from BDH Chemical, Ltd. (Poole, UK). The synthesis of ruthenium amine binuclear complex (Ru360) was performed as described by Ying et al. ³⁰  

Preparation of Mitochondria

Mitochondria were isolated from bovine retina as described by Mederano and Fox. ³¹ Retinas obtained from eight bovine eyes were washed in buffer A containing 225 mM mannitol, 75 mM sucrose, l mM EGTA, and 5 mM HEPES (pH 7.0). Retinas were placed in a glass Teflon homogenizer containing 15 mL buffer A and were homogenized (six strokes). Buffer A (20 mL) containing 2 mM EGTA and 1% BSA was added and centrifuged at 1000g for 5 minutes. The supernatant was diluted 1:1 with buffer A and centrifuged again at 1000g for 5 minutes. The pellet was discarded, and the supernatant was centrifuged at 11,500g for 15 minutes. The pellet was resuspended in buffer A (without EGTA) and centrifuged again at 11,500g for 15 minutes. Finally, the pellet containing the mitochondria was resuspended in buffer A (without EGTA) and either used on the same day or frozen in liquid nitrogen for storage at −70°C. The freshly isolated mitochondria were approximately 80% intact, as indicated by their succinate-cytochrome c oxidoreductase activity before and after hypotonic shock (for detail see Ref. ²³ ). This mitochondrial preparation is well coupled, with a respiration control ratio of 2.44 ± 0.2 (n = 3), as determined from the rate of succinate-supported electron transport in the absence and presence of gramicidin. Protein concentration was determined according to Lowry et al. ³²  

Ca²⁺ Accumulation and PTP Formation

Ca²⁺ uptake by freshly prepared retinal mitochondria (0.5 mg/mL) was assayed for 1 to 20 minutes at 30°C in the presence of 225 mM mannitol, 75 mM sucrose, 120 μM CaCl₂ (containing 3 × 10⁴ cpm/nmol ⁴⁵Ca²⁺), 5 mM HEPES/KOH [pH 7.0], and 5 mM succinate with 0.l mM PI, or 4 mM MgCl₂ with 3 mM ATP. Uptake was terminated by rapid filtration (Millipore, Bedford, MA) followed by a wash with 5 mL 0.15 M KCl. It should be noted that in the presence of succinate and Pi, Ca²⁺ accumulation is transient: it reaches a maximal level and then rapidly decreases to approximately 30% to 40% of its maximal value (reflecting PTP opening). Mitochondrial swelling was also used to observe PTP formation. Swelling of cyclosporin A (CsA)-sensitive retinal mitochondria was measured as described by He et al., ³³ except that swelling was estimated by the decrease in light scattering at 540 nm. Thus, PTP opening was followed by accumulation of Ca²⁺ and swelling of mitochondria. 

Purification of VDAC

VDAC was purified as previously described. ²⁴ Briefly, bovine retina (5–20 mg protein) was incubated for 30 minutes at 0°C (at 5 mg/mL) in a solution containing 10 mM Tris (pH 7.0), 0.15 mM PMSF, 0.5 μg/mL leupeptin (pH 7.0), and 0.2% Triton X-100. After centrifugation at 44,000g for 20 minutes, the pellet was resuspended at 5 mg/mL in the above solution, but with a 3% Triton X-100 concentration. After centrifugation at 44,000g for 30 minutes, the Triton X-100 extract was applied to a dry hydroxyapatite/Celite (2:1 wt/wt) column (0.l g/mg protein) and eluted with a buffer containing 10 mM Tris (pH 7.4), and 3% Triton X-100. The VDAC-containing fractions were collected, pre-equilibrated with 10 mM Tris/HCl (pH 7.4), and 0.4% Nonidet P-40, and loaded onto a reactive red-agarose column (0.1 mL/mg protein). The loaded column was washed with the buffer, and VDAC was eluted with the same buffer with 0.4 M NaCl added. 

Gel Electrophoresis and Immunoblot Analyses

Analysis of the protein profile was performed using SDS-PAGE with the discontinuous buffer system of Laemmli, ³⁴ using 1.5-mm thick slab gels of 11% and 3.5% acrylamide for separating and stacking gels, respectively. Gels were stained with Coomassie brilliant blue. Western blot analysis was performed by standard procedures. ³⁵ The separated proteins from SDS-PAGE were electrophoretically transferred onto nitrocellulose membranes. For immunostaining, the membranes were blocked with 3% nonfat dry milk and 0.1% Tween-20 in Tris-buffered saline, incubated with monoclonal anti-VDAC antibodies (1:7000), and then with alkaline-phosphatase-conjugated anti-mouse IgG as a secondary antibody (1:10,000). The color was developed with 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium. 

Single-Channel Recording and Analysis

Reconstitution of purified VDAC into PLB, single-channel current recording, and data analysis were performed as previously described. ^{23 24 36} Briefly, PLB was prepared from soybean asolectin dissolved in n-decane (50 mg/mL). Only PLB with a resistance greater than 100 GΩ was used. Purified VDAC (approximately 1 ng) was added to the cis chamber. After one or a few channels were inserted into the PLB, excess protein was removed by washing the cis chamber with 20 volumes of solution to prevent further incorporation. Currents were recorded under voltage-clamp using an amplifier (Bilayer Clamp BC-525B; Warner Instruments, Hamden, CT). The currents were measured with respect to the trans side of the membrane (ground). The currents were low-pass filtered at 1 kHz (Bessal; Frequency Devices, Haverhill, MA), and digitized online using an interface board (model 1200; Digi-Data, Jessup, MD) and (with PCLAMP 6 software; Axon Instruments, Inc., Union City, CA). Curve fitting was done with scientific software (Sigma Plot 6.0; Jandel Scientific, SPSS Inc., Chicago, IL). 

Light Microscope Immunocytochemistry

Eyes from an adult rat, guinea pig, or rabbit were removed in animals under deep anesthesia (for rat and rabbit, pentobarbital at 45 μg/g; for guinea pig, ketamine at 40 μg/g, xylazine at 8 μg/g, and pentobarbital at 35 μg/g). Anesthesia was injected intraperitoneally. Animals were killed by anesthetic overdose (three times the initial dose). Research animals were obtained and cared for in accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals, Institute of Laboratory Animal Resources and University of Pennsylvania policy. Animal experimentation was conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The eyecups were removed and fixed in 4% (wt/vol) paraformaldehyde in phosphate buffer (PB; 0.1 M, pH 7.4) for 1 hour and cryoprotected by incubation overnight in 30% sucrose in PB at 4°C. Retina was separated from the back of the eye, embedded in a mixture of tissue-freezing medium (Electron Microscopy Sciences, Fort Washington, PA) and 20% sucrose (1:2), frozen in liquid nitrogen, and cryosectioned vertically at 12 μm. Sections were incubated in blocking solution (phosphate buffer containing 5% sucrose [SPB], 10% normal goat serum, and 0.1% saponin) for 1 hour, and in the primary antibody (diluted in blocking solution) for 16 to 20 hours at 4°C. After three washes with SPB, the sections were incubated for 2 hours in secondary antibody conjugated to HRP, and developed with 3,3′-diaminobenzidine tetrahydrochloride (DAB). The sections were washed again, mounted in glycerol, and imaged with a conventional light microscope. 

In Situ Hybridization

Fixed paraformaldehyde (4%) and cryoprotected retinas were cryosectioned at 20-μm. cDNAs for mouse VDAC-1, -2, and -3 were kindly provided by William J. Craigen. ¹⁴ The template plasmid for antisense and sense riboprobes for the three cDNAs were subcloned (pBluescript SK+; Stratagene, La Jolla, CA) and were linearized with XhoI and SacI, respectively. The antisense and sense transcripts were obtained by in vitro transcription, using [³³P]UTP and T3 or T7 RNA polymerase (Promega). The transcription reaction was performed by incubating approximately 1 mg linearized plasmid with 10 U T3 or T7 RNA polymerase for 2 hours at 37°C. Alkaline hydrolysis of the probes was performed for 30 minutes at 60°C. In situ hybridization was performed as described. ³⁷ Prehybridization was performed in equilibration solution containing 50% (vol/vol) formamide, 10% (wt/vol) dextran sulfate, and 1× Denhardt solution, at 56°C for 2 hours. Hybridization was performed in prehybridization solution with the addition of 0.5 mg/mL transfer (t)RNA, and 10⁷ cpm/slide of RNA probe and was performed overnight at 56°C. Sections were washed with 2× SSC at 56°C for 30 minutes, followed by 0.1× SSC at 56°C for 3 hours, and then dehydrated, dried, and exposed at 4°C to photograph emulsion (NBT-2; Eastman Kodak, Rochester, NY) for 2 weeks. 

Results

Purification of VDAC from Bovine Retina

VDAC was purified from mitochondria isolated from bovine retina with a simple two-step chromatography method: a hydroxyapatite column followed by a reactive red-agarose column (Fig. 1) . Using this method, we have purified VDAC fromliver mitochondria, ²⁴ brain synaptosomes, ²³ and sarcoplasmic reticulum. However, in retina, before VDAC was solubilized with 3% Triton X-100, mitochondria were treated with 0.2% Triton X-100; this step extracted two proteins (30 kDa and approximately 100 kDa), which otherwise would copurify with VDAC in the hydroxyapatite column. Approximately 90% of the 3% Triton-X-100-extracted proteins bound to the hydroxyapatite column, but VDAC (a 32-kDa protein identified by a specific monoclonal antibody) was eluted (Fig. 1B) . For further purification, the VDAC-containing fractions from the hydroxyapatite column were combined and applied to a reactive red-agarose column. VDAC bound to the reactive red- agarose was eluted with 0.4 M NaCl (Fig. 1) . The purity of VDAC was estimated at approximately 98% by densitometric scanning of a Coomassie-stained VDAC (after SDS-PAGE). 

Characterization of VDAC Channel Activity

Purified retinal VDAC was reconstituted into PLB and studied under voltage clamp conditions. Figure 2A shows current traces from a single channel in response to voltages stepped from a holding potential of 0 mV to the potentials indicated in each current trace. Similar to all known VDACs, VDAC was fully open at 0 mV; however, no net current was measured at this potential (Fig. 2A , dashed line), because symmetrical solutions were used. The chord conductance of the main conductance state was assessed in symmetrical solutions by applying either −10 or +10 mV to the membrane. At these relatively small membrane potentials, the conductance remained constant for up to 120 minutes of recording. In symmetrical solutions of 1 M NaCl, the chord conductance at 10 mV was estimated to be 3.7 ± 0.l nanosiemens (nS; n = 3). 

In contrast to the constant currents observed at −10 or +10 mV in NaCl solutions, when larger voltage steps were applied, the channel converted to a lower conductance state within seconds (Fig. 2A , lower trace). Transitions between the main conductance state and the subconductance state also occurred when a voltage ramp between −60 and +60 mV was applied (Fig. 2B , solid bars). As expected for symmetrical solutions, the current reversed at 0 mV (Fig. 2B , open arrow). 

Modulation of Retinal VDAC Channel Activity

Recently, we demonstrated that glutamate ²⁴ and La^{3+ 23} modulate the channel activity of VDAC isolated from rat liver or sheep brain synaptosomes. To determine whether retinal VDAC is similarly modulated, we tested the effect of these reagents on VDAC channel activity. At 10 mV, and in the absence of glutamate, the channel remained stable in the full conducting state for more than 30 minutes. When glutamate was added, the channel displayed fast alternations between the fully open state, subconductance states, and the fully closed state (Fig. 3A) . Figure 3B shows that when La³⁺ was added to either side of the reconstituted VDAC, the open channel was stabilized in the low-conducting state. The inhibitory effect of La³⁺ was reversed by EGTA. As we reported previously, in some cases, La³⁺ inhibited the channel activity by stabilizing the channel in a long-lived closed state with zero conductance (not shown). ²³ Ruthenium amine binuclear complex (Ru360) is a trivalent cation, shown to inhibit mitochondrial Ca²⁺ transport ³⁰ and PTP opening. ²³ Ru360, at relatively low concentrations, decreased VDAC conductance in single-channel (Fig. 4A) and multichannel (Fig. 4B) experiments. In single channel measurements, in the presence of Ru360, VDAC was stabilized at the low-conducting state, with fast fluctuations between the substate and closed state (Fig. 4A) . 

Ca²⁺ Accumulation in Isolated Bovine Retinal Mitochondria

Because VDAC is permeable to Ca²⁺, ²³ VDAC may control the permeability of the outer mitochondrial membrane to this cation. We therefore tested whether inhibition of VDAC channel activity also inhibits Ca²⁺ transport in isolated mitochondria energized by succinate or ATP. As expected, ATP-stimulated Ca²⁺ uptake was inhibited by the proton ionophore FCCP and oligomycin (Table 1) . In addition, La³⁺, RuR, and Ru360, which inhibit VDAC channel activity, also inhibited Ca²⁺ uptake (Table 1) . The inhibition of Ca²⁺ accumulation by La³⁺, RuR, and Ru360 was obtained at the same range of concentrations reported previously for the inhibition of Ca²⁺ uptake by isolated rat liver mitochondria. ^{30 38}  

Inhibition of PTP Opening by Ru360

We followed PTP opening by monitoring the Ca²⁺ accumulation of energized mitochondria as a function of time. When freshly prepared mitochondria were incubated in a solution containing succinate, a transient Ca²⁺ accumulation was observed that reached a peak of 80 nmol/mg protein and then rapidly decreased to 30 nmol/mg protein (Fig. 5A) . This transient Ca²⁺ retention (uptake followed by release) requires the presence of Pi ³⁸ and suggests that mitochondria undergo a permeability transition, losing the accumulated Ca²⁺ through the PTP. ^{3 4 5 6 39 40} To test the effect of Ru360 on PTP, we added it after Ca²⁺ accumulation had reached the maximal level and before PTP was activated; this prevented the release of Ca²⁺ (Fig. 5B) . We also tested the effect of Ru360 on mitochondrial swelling. As in liver mitochondria, ²³ swelling of retinal mitochondria was inhibited by CsA and prevented by EGTA. The effect of Ru360 was similar to that of RuR, in that it decreased swelling (Fig. 5C) . Inhibition of PTP opening by Ru360 is most likely not due to its interaction with the Ca²⁺ uniporter, because inhibition of Ca²⁺ uptake would increase, not decrease, Ca²⁺ release. Thus, both swelling and Ca²⁺ accumulation, which reflect PTP assembly or opening, are sensitive to Ru360, which closes VDAC. 

Localization of VDAC in Retina

In all species tested (rat, rabbit, bovine, mouse, guinea pig, cat, and monkey), immunostaining for VDAC was similar (Fig. 6) . Staining was most intense in the ellipsoid (the mitochondria-enriched region within the photoreceptor’s inner segment; except for bovine retina), and in both the inner and outer plexiform layers of the retina. Stain was also observed in the outer limiting membrane, formed by the distal processes of Müller cells (Fig. 6D , white arrow), and in somas of some horizontal, amacrine, and ganglion cells (Fig. 6E) . 

Localization of VDAC mRNAs

The antibody against VDAC recognizes all known isoforms of VDAC. ¹⁶ To determine which isoform is expressed in retina, we performed in situ hybridization using specific probes for mouse VDAC-1, -2, and -3. All probes hybridized with a similar pattern (shown here only for VDAC-1; Fig. 7 ). In mouse retina, VDAC-1 was expressed in the myoid of the photoreceptor’s inner segment (lightly), and in the inner nuclear layer (INL) and ganglion cell layer (GCL; strongly). Within the ganglion cell, clusters of silver grains were located over the cytoplasmic rim around individual ganglion cell nuclei. This is consistent with the localization of protein, as analyzed by immunostaining (Fig. 6) . Although the hybridization pattern was similar for different isoforms, hybridization intensity varied greatly. To quantify these differences, we estimated the density of silver grains in the INL and GCL, and divided it by the density in the outer nuclear layer (ONL; showed density as low as that of the sense background hybridization; Table 2 ). This shows that VDAC-1 is most abundant, followed by VDAC-2 and -3. For all isoforms, the relative density in INL was similar to that in GCL. 

Discussion

In this study we show that the retina expresses the three known isoforms of VDAC (VDAC-1, -2, and -3) with similar distribution, but at different levels. In retina, VDAC is localized, as expected, to mitochondria of most cell types, predominantly at the highly metabolically active mitochondria of photoreceptors. Although immunostaining for VDAC in the inner segment was very strong, we were surprised to see a weak response to in situ hybridization. This may suggest that photoreceptors express an unknown isoform of VDAC in addition to VDAC-1. Alternatively, VDAC mRNA in the photoreceptors may not be a good predictor of the amount of protein expressed. 

Channel properties of retinal VDAC (as reflected by channel activity after reconstitution into PLB) are similar to VDAC purified from other tissues. VDAC conductance is voltage-dependent with a maximal unitary conductance of 3.7 ± 0.l nS in 1 M NaCl and at 10 mV, it is permeable to anions and cations including Ca²⁺, its channel activity is inhibited by La³⁺ and Ru360, and it is modulated by glutamate (Figs. 2 3 and 4) . 

Until recently, the outer mitochondrial membrane was not considered a barrier for transport into and out of mitochondria, because of the assumption that VDAC is freely permeable to ions and uncharged molecules. However, this view was changed when VDAC was shown to regulate the outer mitochondrial membrane, suggesting that VDAC can control coupled respiration and cell survival. ⁴⁰ Our finding that Ru360 inhibits VDAC channel activity and PTP opening supports this concept. ^{23 24} Furthermore, the results suggest that VDAC is involved in the regulation of cytosolic and mitochondrial Ca²⁺ concentration and in PTP assembly and/or activation. 

In retina, various studies indicate that the mitochondria in the photoreceptor’s inner segments are involved in the process of retinal degeneration. ^{28 31 41 42 43} Isolated retina exposed to high external Ca²⁺ concentration has low ATP concentration and respiration level, and it undergoes rod degeneration. ^{31 41} In photoreceptors, mutations in phototransduction proteins (e.g., the rd mouse with a mutation in phosphodiesterase, the GCAP Y99C mutation, and deletion of cGMP-gated channel) cause a variety of retinal degeneration diseases. ⁴⁴ In other retinal neurons, especially ganglion cells, ischemia, diabetes, and mechanical damage trigger apoptotic events. ⁴⁵ In all, or most of these cases, apoptosis appears to be promoted by an abnormal concentration of intracellular Ca²⁺. ⁴⁴ The relationship between changes in Ca²⁺ concentration and mitochondria dysfunction is well established. ³ Ca²⁺ accumulated in the mitochondria, activates PTP; leading to mitochondrial swelling, and triggering an apoptotic event. ^{28 43} If, as suggested, the site of Ca²⁺ action is within the mitochondria, on the matrix side of the inner mitochondria membrane, ⁴³ then Ca²⁺ must first cross the outer mitochondrial membrane. VDAC is permeable to Ca²⁺, and it possesses Ca²⁺ binding sites. ²³ Thus, VDAC, as a component of the PTP, ^{3 5 23 46 47 48 49 50} may be involved in inducing mitochondrial swelling. This, together with its being highly localized to photoreceptors and some ganglion cell mitochondria, implies that it is most likely to participate in apoptotic events in these particular cells. In summary, VDAC may play a key role in processes of retinal degeneration that result from ATP depletion and Ca²⁺ overload. 

 Supported by a grant from the Israel Science Foundation (VS-B) and National Eye Institute Grant EY11105 (NV). VS-B is the incumbent of the Hyman Kreitman Chair in Bioenergetics at Ben-Gurion University of the Negev.

 Submitted for publication June 8, 2001; revised February 12, 2002; accepted March 15, 2002.

 Commercial relationships policy: N.

 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 1.

 

Purification of VDAC from retinal mitochondria. VDAC was purified from mitochondria of bovine retina by a two-step method: hydroxyapatite (HA) and reactive red-agarose (RRA) chromatography. Mitochondria before treatment (Mito) and after treatment with 0.2% Triton X-100 (lane A), 3% Triton X-100 extract (lane B), HA flow through (lane C), RRA flow through (lane D), and RRA fractions were subjected to SDS-PAGE (11% acrylamide) and Western blot analysis. (A) Coomassie blue stained gel; (B) corresponding immunoblot. The positions of molecular weight standards are indicated.

Figure 2.

 

Channel activity of purified retinal VDAC. Purified retinal VDAC (1 ng) was reconstituted into PLB. (A) Current traces were obtained in response to voltage steps from 0 mV to the voltage indicated below each current trace. Dashed line: zero current level (c), main (o), and subconductance (s) states of the channel. (B) Current response to a voltage ramp applied to a single VDAC; the membrane voltage was changed linearly between −60 mV and +60 mV at 62.5 mV/sec.

Figure 3.

 

VDAC channel activity is modulated by glutamate and La³⁺. VDAC was reconstituted into PLB, and single-channel currents were recorded. (A) Representative records in response to voltage step (0 to 10 mV) before and after the addition of 5 mM glutamate. (B) Representative records before and after addition of La³⁺ (50 μM) and after subsequent addition of EGTA (0.5 mM). The fully open, subconductance, and closed states of the channel are labeled o, s, and c, respectively.

Figure 4.

 

Ru360 inhibits VDAC channel activity. (A) VDAC reconstituted into PLB as in Figure 3 was exposed to 0.15 μM Ru360 in the presence of 1 M NaCl, and recordings were made before and 2 min after the addition of Ru360. (B) Average steady state conductance of VDAC (approximately 10 channels) as a function of voltage, before (•) and 10 minutes after the addition of 0.15 μM Ru360 (○). G, conductance; Gmax, maximal conductance.

Figure 6.

 

Expression of VDAC in retina. Cryostat sections (12 μm) from rabbit, rat, and bovine retina were immunolabeled for VDAC, using monoclonal anti-VDAC antibody (1:1000) and HRP-conjugated secondary antibody, and visualized with DAB reaction product. (A–C) Low magnification showing strong staining in rabbit (A) and rat (B) inner segments (IS), outer plexiform layer (OPL), and inner plexiform layer (IPL), but only weak staining in bovine retina (C). (D) Enlargement of rabbit outer retina. Stain was mainly concentrated in the ellipsoid (e) and in puncta on the outer limiting membrane (arrow). (E) Enlargement of rabbit inner retina. Stain was present in horizontal cell somas (H), amacrine cell soma (A), and ganglion cell somas (GC). OS, outer segments; m, myoid. Scale bars, 20 μm.

Table 1.

 

Ca²⁺ Accumulation and Inhibition in Mitochondria Isolated from Retina

Additions	Ca2+ Accumulation (% of Control)
		Energized with ATP + Mg2+	Energized with Succinate + Pi
None	100	100
FCCP (25 μM)	4.0 ± 0.6	50.8 ± 4.3
Oligomycin (20 μM)	24.5 ± 2.4	ND
RuR (10 μM)	5.9 ± 1.3	21.1 ± 0.4
Ru360 (0.5 μM)	7.5 ± 1.2	46.4 ± 6.3
La3+ (50 μM)	12.2 ± 2.6	ND

 Ca²⁺ uptake by retinal mitochondria (0.5 mg/mL) was energized by ATP or succinate. The uptake was assayed for 1 minute. Control activities (100%) for Ca²⁺ accumulation assayed in the presence of ATP + MgCl₂ or succinate + Pi were: 260 and 80.4 nmol/mg protein per minute, respectively. RuR and Ru360 were added to mitochondria 1 minute before addition of CaCl₂. The results are the means ± SD of three separate experiments. ND, not determined, because La³⁺ forms an insoluble complex with Pi.

Figure 5.

 

Ru360 inhibits the permeability transition pore in mitochondria. Mitochondria (at 0.5 mg/mL) were assayed for succinate-supported Ca²⁺ accumulation or (at 0.25 mg/mL) for mitochondrial swelling, and PTP opening was monitored. (A) Ca²⁺ content of the mitochondria was assayed in the absence and the presence of Pi (0.l mM). (B) Ru360 (0.15 μM) was added after Ca²⁺ accumulation reached a maximal level. (C) Mitochondrial swelling was assayed in the absence (control) and the presence of EGTA (1 mM), CsA (1 μM), or Ru360 (1 μM). The results represent three similar experiments.

Table 2.

 

Distribution of VDAC Isoforms in Retina

Isoform	Signal (Relative Amount)
		INL	GCL
VDAC-1	3.6	3.8
VDAC-2	2.5	2.7
VDAC-3	0.8	1.0

 Mouse retinas were fixed, and ³³P-labeled riboprobe complementary to mRNA of VDAC-1, VDAC-2, or VDAC-3 was hybridized. Silver grains in the ONL, INL, and GCL were counted from antisense and sense hybridization. Numbers represent grain density of antisense in the indicated layer divided by the density in the ONL. The level of the silver grains in the ONL was similar to that in the sense hybridization preparations; thus, it was used as an internal background.

Figure 7.

 

Distribution of VDAC mRNA in mouse retina. Mouse retina was fixed in buffered paraformaldehyde (4%). Twenty-micrometer-thick frozen sections were obtained and in situ hybridized to antisense ³³P-labeled riboprobe complementary to VDAC-1 cDNA, and then exposed to film. For the negative control, sections were hybridized with the sense probe. Silver grains concentrated in large and small somas in the GCL and throughout the INL. Weak stain was also observed in the myoid of the inner segment (brackets). DIC, differential interference contrast.

The authors thank Ran Zalk for supplying bovine eyes and Adrian Israelson for the synthesis of Ru360.