The maxi-anion channel with a large single-channel conductance of >300 pS, and unknown molecular identity, is functionally expressed in a large variety of cell types. The channel is activated by a number of experimental maneuvers such as exposing cells to hypotonic or ischemic stress. The most effective and consistent method of activating it is patch membrane excision. However, the activation mechanism of the maxi-anion channel remains poorly understood at present. In the present study, involvement of phosphorylation/dephosphorylation in excision-induced activation was examined. In mouse mammary fibroblastic C127 cells, activity of the channel was suppressed by intracellular application of Mg-ATP, but not Mg-5′-adenylylimidodiphosphate (AMP-PNP), in a concentration-dependent manner. When a cocktail of broad-spectrum tyrosine phosphatase inhibitors was applied, channel activation was completely abolished, whereas inhibitors of serine/threonine protein phosphatases had no effect. On the other hand, protein tyrosine kinase inhibitors brought the channel out of an inactivated state. In mouse adult skin fibroblasts (MAFs) in primary culture, similar maxi-anion channels were found to be activated on membrane excision, in a manner sensitive to tyrosine phosphatase inhibitors. In MAFs isolated from animals deficient in receptor protein tyrosine phosphatase (RPTP)ζ, activation of the maxi-anion channel was significantly slower and less prominent compared with that observed in wild-type MAFs; however, channel activation was restored by transfection of the RPTPζ gene. Thus it is concluded that activation of the maxi-anion channel involves protein dephosphorylation mediated by protein tyrosine phosphatases that include RPTPζ in mouse fibroblasts, but not in C127 cells.
- protein tyrosine phosphatase
the maxi-anion channel is a voltage-dependent, large-conductance anion-selective channel that has been observed in a wide variety of cell types (45). Maxi-anion channels are usually silent in unperturbed cells but are generally activated by excision of patch membranes. They can also be activated in cell-attached patches by hypotonic or ischemic stress (13, 14, 16, 23, 31, 32, 42, 54) and by stimulation of some receptors (19, 25, 52, 58, 59). This channel has fairly uniform biophysical properties: the single-channel currents flicker between the fully open and fully closed states, through transitional current levels; the channel has a large single-channel conductance of ∼300–400 pS with linear current-to-voltage (I-V) characteristics; and the channel has voltage-dependent gating, exhibiting a bell-shaped activation curve with a maximal open probability (Po) at ∼0 mV. The channel shows selectivity for anions over cations with a permeability ratio of PCl/PNa = 10–30 and an anion permeability sequence of I− > Br− > Cl− > F−. The maxi-anion channel can also pass large organic anions, such as aspartate− and glutamate− (6, 32, 39, 50), and even ATP4− (4, 14, 31, 42).
Roles of the maxi-anion channel in cell volume regulation (43, 57), cell-to-cell signaling (4, 43, 44), and programmed cell death (1, 15) have been hypothesized. Our recent studies (45) have shown that the maxi-anion channel serves as a conductive pathway for regulated release of ATP in a number of cell types. The released ATP was shown to upregulate the regulatory volume decrease in C127 cells (20). In kidney macula densa cells, salt stress activates maxi-anion channels and induces massive release of ATP, which is involved in tubuloglomerular feedback (4). Despite the wide distribution and physiological importance of the maxi-anion channel, its molecular identity is not known.
The activity of the maxi-anion channel has been shown to be modulated by G proteins (36, 37, 53, 58, 59), cAMP (11, 63), cytoskeletal actins (54), protein kinase C (PKC) (19, 26, 47, 52), and okadaic acid (OA)-sensitive serine/threonine protein phosphatases (10). However, the precise activation mechanism of the maxi-anion channel is still poorly understood. Since the most consistent and powerful stimulus for activating this channel is patch excision, in the experiments here we studied the mechanism of excision-induced activation of maxi-anion channels in murine mammary fibroblastic C127 cells and mouse adult skin fibroblasts (MAFs) in primary culture. We demonstrate that tyrosine dephosphorylation mediated by protein tyrosine phosphatases (PTPs) governs excision-induced activation of the maxi-anion channels.
MATERIALS AND METHODS
Adenosine 5′-triphosphate disodium salt (ATP), 5′-adenylylimidodiphosphate lithium salt (AMP-PNP), okadaic acid potassium salt (OA), cyclosporin A (CsA), ascomycin (FK520), sodium orthovanadate, isoquinoline-5-sulfonic 2-methyl-1-piperazide (H-7), 2,3-bis[(2-hydroxyethyl)thio]-1,4-naphthoquinone (NSC 95397), and sodium tartrate dibasic dehydrate were purchased from Sigma (St. Louis, MO). N-(9,10-dioxo-9,10-dihydro-phenanthren-2-yl)-2,2-dimethyl-propionamide [CD45 inhibitor (CD45I)], 4-methoxyphenacyl bromide [PTP inhibitor II (PTPI-II)], bis(4-trifluoromethylsulfonamidophenyl)-1,4-diisopropylbenzene [PTP inhibitor IV (PTPI-IV)], potassium bisperoxo(bipyridine)oxovanadate V (bpV), 3,4-dihydroxy-N-methyl-N-nitrosoaniline (3,4-dephostatin), and tyrphostin A23 (AG18) were from Calbiochem (Darmstadt, Germany). Genistein, daidzein, sodium molybdate, and other salts were purchased from WAKO (Osaka, Japan). (−)-p-Bromotetramisole oxalate (pBTM) was from Katayama Chemical (Osaka, Japan), and 1,3-diazacyclopenta-2,4-diene (imidazole) was from Nacalai Tesque (Kyoto, Japan).
Normal Ringer solution was composed of (in mM) 135 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, 5 Na-HEPES, 6 HEPES, and 5 glucose (pH 7.4, 290 mosmol/kgH2O). The hypotonic solution was composed of (in mM) 100 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, 5 Na-HEPES, 6 HEPES, and 5 glucose (pH 7.4, 210 mosmol/kgH2O). All artificial intracellular solutions (AISs) were composed of (in mM) 140 KCl, 5 EGTA, and 5 HEPES (pH 7.4, adjusted with KOH; 294 mosmol/kgH2O). Depending on the experimental conditions used, different amounts of CaCl2 and MgCl2 were added to AISs to maintain the free Ca2+ concentration at 100 nM and the free Mg2+ concentration at varying levels, respectively (Table 1). For experiments testing the effect of MgATP on the channel activity, varying concentrations of Na2ATP were added into AISs (Table 1). All ATP-containing solutions were kept on ice and warmed to room temperature immediately before experiments. The concentrations of free Ca2+ and Mg2+ were calculated with CaBuf software (kindly provided by Dr. G. Droogmans, Katholieke Universiteit Leuven, Belgium) and are listed in Table 1.
Patch-clamp experiments started with normal Ringer solution in the bath and pipette. In the inside-out experiments the bath solution was replaced with AIS before patch excision. For measurements of glutamate permeability of maxi-anion channels, the low-Cl− AIS containing 15 mM Cl− was prepared by replacing 140 mM KCl with 140 mM K-glutamate.
In whole cell experiments, the pipette solution was composed of (in mM) 135 CsCl, 5 HEPES, 10 EGTA, 1 MgCl2, and 2 CaCl2 (pH 7.4 adjusted with CsOH; 280 mosmol/kgH2O); the bath was normal Ringer solution.
In experiments with inhibitors of phosphatases and kinases, inhibitor stock solutions were prepared in DMSO and were diluted to their final concentrations just before the experiments. The final DMSO concentration did not exceed 0.1%, and DMSO did not have any effects when added alone.
The osmolality of solutions was measured with a freezing-point depression osmometer (OM802, Vogel).
The experimental protocol was approved in advance by the Ethics Review Committee for Animal Experimentations of the National Institute for Physiological Sciences. A fibroblastic cell line of mouse mammary tissue origin, C127, was obtained from the American Type Culture Collection and cultured in high-glucose Dulbecco's modified Eagle's medium (DMEM) supplemented with 200 mg/l penicillin (Meiji Seika, Tokyo, Japan), 100 mg/l streptomycin (Meiji Seika) and 10% fetal bovine serum (FBS) (MP Biomedicals, Irvine, CA). Primary cultures of MAFs were prepared from skin tissues of ears of wild-type (WT) mice and mice deficient in receptor protein tyrosine phosphatase (RPTP)ζ (56). The cultured cells were kept at 37°C with 5% CO2 and split every 3–4 days.
Primary skin fibroblasts (MAFs) were isolated from 110-day-old mice as described elsewhere (48) with some modifications. Briefly, the mouse was anesthetized with diethyl ether by inhalation before use. A piece of tissue of ∼3-mm width and ∼5-mm length was then cut from each ear, which had been cleaned with 70% ethanol before the operation. After the sample tissue was taken, the animal was killed by decapitation. The tissue was then minced and suspended in 0.5 ml of MAF growth medium (MGM), which contained high-glucose DMEM, 20% FBS, 200 mg/l penicillin, 100 mg/l streptomycin, and 1% fungizone, supplemented with 1 mg/ml collagenase type II (Gibco-Invitrogen, Carlsbad, CA). A 12-well culture dish with the tissue biopsies was incubated overnight inside a 37°C humidified incubator with 5% CO2 in air. The cell-tissue mixtures were suspended by repeated pipetting (5-ml pipette) and passed through sterile nylon netting into sterile 14-ml centrifuge tubes. The cells were collected by a 5-min centrifugation at 1,000 g, and then the collagenase solution was drawn off from the cell pellet. Cells were resuspended in 3 ml of MGM in each well of a six-well dish. After 2 days, approximately two-thirds of the total volume of the medium was removed and replaced with fresh MGM. Six or seven days after seeding the initial cultures usually became confluent, and cells were replated for the next passage. When MAF cultures were established, MGM was changed to the regular culture medium: high-glucose DMEM supplemented with 10% FBS, 200 mg/l penicillin, and 100 mg/l streptomycin. For patch-clamp experiments the cells were grown on glass coverslips.
Expression vectors and transfection.
Genes for RPTPζ and its dominant-negative form (24) as well as for RPTPγ (55) were inserted into the SmaI site of a bicistronic vector, pIRES2-EGFP (Clontech), to produce the expression vectors for cell transfections. Expression vectors for RPTPζ and its dominant-negative form were transfected into MAFs derived from RPTPζ-knockout (RPTPζ-KO) mice (56). Effectene transfection reagent (Qiagen, Chatsworth, CA) was used for transfections according to the manufacturer's instructions. Electrophysiological experiments were performed within a period of 24–48 h after transfection. The transfection was monitored by a fluorescence microscope with a green fluorescent protein filter set.
RT-PCR analysis of mRNA expression.
Total RNA was isolated from C127 cells and MAFs with Sepasol RNA I reagent (Nacalai Tesque) according to the manufacturer's instructions. Total RNA was treated with DNase I (Takara Bio., Otsu, Japan) to remove genomic DNA. First-strand cDNA was synthesized from the isolated RNA with reverse transcriptase (Roche Diagnostics, Mannheim, Germany) and oligo(dT) primers (Invitrogen, Carlsbad, CA). Synthesis of cDNA was performed according to the manufacturer's protocol. Gene-specific primers used for PCR were designed with Primer3 software (www.genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi). The sequences of the forward and reverse primers for RPTPζ were 5′-CACATGGGGACTAAATACAATGAA-3′ and 5′-GGATCCTTTAGTGATTCCTCTGAA-3′ (product size 557 bp), respectively. PCR was carried out with AmpliTaq Gold (Applied Biosystems, Tokyo, Japan) and iCycler (Bio-Rad Laboratories, Tokyo, Japan). Cycling conditions were 10 min at 95°C, followed by 30 cycles of 1 min at 95°C, 2 min at 50°C (annealing temperature), and 2 min at 72°C, and finally 10 min at 72°C. As a positive control, we amplified a β-actin sequence with a specific set of primers with sequences 5′-AACACCCCAGCCATGTAGGTAG-3′ and 5′-TGTCAAAGAAAGGGTGTAGGTAG-3′ (product size 789 bp). In experiments for the negative control, reverse transcriptase was omitted. PCR products were analyzed on a 2% agarose gel to confirm the sizes expected from the known cDNA sequences.
All patch-clamp single-channel recordings were performed at room temperature (23–25°C). Pipettes were pulled from borosilicate glass capillaries (outer diameter 1.4 mm, inner diameter 1.0 mm) with a micropipette puller (model P-97; Sutter Instruments, Novato, CA) and had a tip resistance of ∼2 MΩ when filled with the pipette solution. Membrane currents were measured with an Axopatch 200A patch-clamp amplifier coupled to a DigiData 1320 interface (Axon Instruments, Union City, CA). The time course of current change was monitored by repetitively applying (every 5 s) alternating step pulses (500-ms duration) to ±25 mV from a holding potential of 0 mV. To observe voltage dependence of the current profile, step pulses were applied with command potentials up to ±50 mV in 10-mV increments. Data acquisition and analysis were done with pCLAMP software (version 9.0.2, Axon Instruments) and WinASCD software (kindly provided by Dr. G. Droogmans). Current signals were filtered at 2 kHz and digitized at 5 kHz. When the bath Cl− concentration was altered, a salt bridge containing 3 M KCl in 2% agarose was used to minimize the bath electrode potential variation. The liquid junction potentials were calculated with pCLAMP software and corrected as necessary. In inside-out and cell-attached experiments, the liquid junction potentials were ∼0 mV in our experimental conditions, and therefore no correction was made.
Single-channel amplitudes were measured manually by placing a cursor at the open and closed channel levels. Background currents were subtracted, and the mean patch currents were measured at the beginning (first 25–30 ms) of current responses to voltage steps in order to minimize the contribution of voltage-dependent current inactivation and the channel occupancy in the subconductance states. The mean number of open channels, NPo (where Po and N represent the open channel probability and the number of active channels, respectively), was calculated by dividing the mean macro-patch current by the single-channel amplitude. The reversal potentials were calculated by fitting I-V curves to a second-order polynomial. Permeability ratios were calculated with the Goldman-Hodgkin-Katz equation: where ΔErev is the reversal potential in the presence of glutamate−; R is the gas constant; T is temperature; F is the Faraday constant; [Glu]o and [Glu]i are glutamate− concentrations in the extracellular and intracellular solutions, respectively; and [Cl]o and [Cl]i are Cl− concentrations in the extracellular and intracellular solutions, respectively. PCl and PGlu are the permeabilities to Cl− and glutamate−, respectively.
Concentration-response data for the MgATP effect on macro-patch currents were fitted to the equation: where A0 and A are the NPo values measured 18 min after patch excision in the absence and presence of MgATP, respectively, at a given MgATP concentration ([MgATP]), and IC50 is the apparent inhibition constant.
Data were analyzed in OriginPro 7 (OriginLab, Northampton, MA). Plotted data are given as means ± SE of n observations. Statistical differences of the data were evaluated by ANOVA and the paired or unpaired Student's t-test where appropriate and considered significant at P < 0.05.
Biophysical properties of maxi-anion channel activated by patch excision from C127 cells in low-Ca2+, cytoplasm-approximating electrolyte solution.
The strongest stimulus known to open the maxi-anion channel is patch excision (45), which ordinarily exposes the intracellular side of the patch membrane to bath solution containing a high concentration of Ca2+. To expose the intracellular side to a low-Ca2+, cytoplasm-approximating electrolyte solution, the patch membrane was excised from C127 cells after replacement of normal Ringer bath solution with AIS containing 100 nM free Ca2+ and 0.804 mM free Mg2+ (see Table 1). This maneuver led to a gradual increase in the patch membrane current. Examples of time-dependent changes in the mean patch current recorded during repetitive application (every 5 s) of alternating test pulses (±25 mV for 0.5 s) from the holding potential (0 mV) immediately after patch excision are shown in Fig. 1A, where for simplicity only data recorded at a positive potential are shown. Immediately after excision, membrane patches usually had no channel activity, as shown in Fig. 1B (t = 0 min). Gradually, the number of detectable single-channel events increased, and after 15–25 min events of 10–15 channels could be observed (Fig. 1B). The observed single-channel current amplitude was 10.0 ± 0.2 pA at +25 mV and −10.1 ± 0.3 pA at −25 mV. Single maxi-anion channels were continuously active at near 0 mV but closed in a time-dependent manner at voltages greater than ±20 mV (Fig. 1C). The unitary I-V relationship was linear under symmetrical Cl− conditions and reversed at ∼0 mV, as shown in Fig. 1D. The unitary slope conductance was 367.4 ± 4.6 pS (n = 7). When the intracellular Cl− concentration was reduced from 155 to 15 mM by replacement of KCl with K-glutamate, the reversal potential shifted to a value of −36.3 ± 0.4 mV (Fig. 1D). This result indicates that the channel is anion selective with a permeability ratio of glutamate− to Cl− of 0.146 ± 0.004 (n = 6). These biophysical properties of the maxi-anion channel recorded in the low-Ca2+, cytoplasm-approximating electrolyte solution were essentially the same as those observed when patch membranes were excised in high-Ca2+ bath solution from C127 cells (42), cardiomyocytes (12, 14), primary cultured astrocytes (32), and mouse embryonic or adult fibroblasts derived from WT and voltage-dependent anion channel (VDAC)-deficient mice in primary culture (46) in the absence of ATP.
MgATP sensitivity of excision-induced activation of maxi-anion channel in C127 cells.
Since no maxi-anion channel activity was observed in the patch before and immediately after excision, there is a possibility that intracellular MgATP plays an inhibitory role in the channel activation. Indeed, addition of Mg2+ and ATP at 1 mM each ([MgATP] = 0.636 mM; see Table 1) to AIS abolished excision-induced activation of the maxi-anion channel (Fig. 2, A and B). The inhibitory effect of MgATP could be observed also on preactivated maxi-anion channels in a fully reversible manner (Fig. 2A, inset). In contrast, when ATP was replaced with its nonhydrolyzable analog, AMP-PNP, excision induced activation of the maxi-anion channel in the same way that excision induced activation in the control ATP-free conditions (Fig. 2, A and B). Thus it appears that ATP hydrolysis is a necessary step for keeping the maxi-anion channel in the closed state. The effect of MgATP was concentration dependent, and excision-induced channel activation was suppressed by increasing [MgATP] while maintaining the total Mg2+ concentration at 1 mM (Fig. 2C). The NPo value (a measure of the mean open channel activity) plotted as a function of [MgATP] could be fitted well to Eq. 2 with IC50 = 29.2 ± 8.9 μM (Fig. 2D). These results suggest that the excised membrane patches retained auxiliary proteins including kinases and phosphatases and that under ATP-free conditions phosphatase activity dominated and caused channel opening, whereas in the presence of MgATP the channel protein remained phosphorylated and stayed in the inactive closed state.
Pharmacological evidence for sensitivity of excision-induced activation of maxi-anion channel in C127 cells to a tyrosine, but not serine/threonine, protein kinase.
To test a possible involvement of protein kinases in the maxi-anion channel activity, we next performed pharmacological experiments using a number of protein kinase inhibitors. As noted above, when the membrane patches were excised in the presence of 0.636 mM MgATP, no channel activation could be observed (Fig. 3). Under this condition, a broad-spectrum serine/threonine protein kinase inhibitor, H-7, had no significant effect on excision-induced activation of the maxi-anion channel at a concentration of 20 μM (Fig. 3). In contrast, a tyrosine kinase inhibitor, AG18 (100 μM) or genistein (100 μM), caused robust activation of the maxi-anion channel even in the presence of 0.636 mM MgATP as demonstrated in Fig. 3. These results suggest that tyrosine, but not serine/threonine, phosphorylation negatively controls the maxi-anion channel activity in C127 cells.
Pharmacological evidence for involvement of a tyrosine, but not serine/threonine, protein phosphatase in excision-induced activation of maxi-anion channel in C127 cells.
We next examined a possible involvement of protein phosphatases in excision-induced activation of the maxi-anion channel by using a variety of protein phosphatase blockers in an intracellular solution with no ATP and 5 mM total Mg2+ (Table 1). As shown in Fig. 4, OA, which is the best-known inhibitor of serine/threonine protein phosphatases and blocks PP1, PP2A, PP4, PP5, and PP6 phosphatases, had no significant effect on maxi-anion channel activation at 100 nM. Similarly, maxi-anion channel activation was not affected by 1 μM CsA or ascomycin (FK520), which are blockers of PP2B (Fig. 4).
In contrast, broad-spectrum inhibitors of PTPs had drastic effects on excision-induced activation of the maxi-anion channel. First, a tyrosine phosphatase inhibitor cocktail (TPhIC), which was prepared by mixing (in mM) 100 sodium orthovanadate, 117 sodium molybdate, 484 sodium tartrate, and 200 imidazole, was tested. This cocktail, at a final concentration of 1% (a 100× dilution), abolished the channel activity in excised membrane patches (Fig. 5, A and B). A similar but slightly weaker effect was produced by sodium orthovanadate alone at 1 mM and by two other broad-spectrum PTP inhibitors, 3,4-dephostatin and pBTM, at 0.1 mM (Fig. 5, A and B). Together, these results strongly suggest an involvement of PTP(s) rather than serine/threonine protein phosphatases in the mechanism of excision-induced activation of the maxi-anion channel in C127 cells.
To specify the type of PTP responsible for activation of the maxi-anion channel, the following five inhibitors were tested: 10 μM NSC 95397, an irreversible inhibitor of Cdc25 that is a dual-specificity phosphatase capable of acting on both tyrosine and serine/threonine sites (29); 10 μM PTPI-IV, a reversible PTP inhibitor that inhibits PTPs broadly (IC50 = 1.8, 2.5, 6.4, 6.7, 8.4, 13 and 20 μM for SHP-2, PTP1B, RPTPζ, RPTPμ, RPTPε, PTP-Meg2, and RPTPδ, respectively) (21); 1 μM CD45I, a potent, selective, and reversible inhibitor of the CD45 receptor tyrosine phosphatase (62); 50 μM PTPI-II, a potent cell-permeant PTP inhibitor of the SHP family (3); and 10 μM bpV, an inhibitor of the PTEN family of PTPs, which also inhibits PTP1B and RPTPβ (51). As shown in Fig. 5, C and D, none of these drugs had a statistically significant effect on excision-induced activation of the maxi-anion channel in C127 cells. These results pharmacologically exclude a large number of known PTPs from consideration as candidates with a role in activation of the channel in C127 cells. It should be noted, however, that the inhibitors listed above do not cover the whole range of known PTPs.
Pharmacological evidence for involvement of protein tyrosine in activation of maxi-anion channels in primary skin fibroblasts.
The activity of the maxi-anion channel that was observed after patch membrane excision from MAFs in low-Ca2+ AIS was essentially the same as that previously observed for excision in high-Ca2+ Ringer solution (46). Single channels were most active near 0 mV and exhibited inactivation kinetics at voltages larger than ±20 mV (Fig. 6A). The unitary I-V relationship was linear under symmetrical Cl− conditions with a slope conductance of 415.5 ± 2.4 pS (Fig. 6B). When intracellular Cl− was reduced from 155 to 15 mM, the reversal potential shifted from ∼0 mV to −37.1 ± 1.2 mV (Fig. 6B), indicating a permeability ratio of glutamate− to Cl− of 0.138 ± 0.011 (n = 8).
As observed in C127 cells (Fig. 5, A and B), excision-induced activation of the maxi-anion channel in MAFs was prominently inhibited by broad-spectrum PTP inhibitors TPhIC (1%), orthovanadate (1 mM), and pBTM (0.1 mM) as shown in Fig. 6, C and D. Although activity of the channel was insensitive to PTPI-IV in C127 cells (Fig. 5, C and D), it was significantly, although not completely, suppressed by this PTP inhibitor in MAFsFig. 5, C and D), it was significantly, (Fig. 6, C and D). This result may suggest that the expression of some PTPI-IV-sensitive PTPs (SHP-2, PTP1B, RPTPζ, RPTPμ, RPTPε, PTP-Meg2, RPTPδ, etc.) in MAFs is different from that in C127 cells.
In the cell-attached configuration, primary skin fibroblasts (MAFs) exhibited no maxi-anion channel activity (Fig. 7A). However, after 10-min preincubation of MAFs with a broad-spectrum tyrosine kinase inhibitor, genistein (100 μM) or AG18 (100 μM), we consistently observed the activation of maxi-anion channels even in the on-cell mode (Fig. 7A). AG18 was significantly more efficient in activating maxi-anion channels compared with genistein (Fig. 7B). As shown in Fig. 7C, dialysis of MAFs with ATP-free pipette solution supplemented with 100 μM AG18 elicited whole cell currents with kinetics reminiscent of those previously reported for maxi-anion channels. The whole cell I-V relationship was linear with a reversal potential of −3.8 ± 0.6 mV. The reversal potential shifted to a value of +31.5 ± 0.8 mV on replacement of 135 NaCl with 135 mM Na-glutamate in the bath, indicating anionic selectivity of the AG18-activated whole cell current with a permeability ratio of glutamate− to Cl− of 0.22 ± 0.01 (n = 4).
Differential molecular expression of RPTPζ in primary skin fibroblasts and C127 cells.
We hypothesized that certain RPTPs would be better candidate regulators of the channel because they are membrane proteins and likely retained in the membrane patches even after excision. Of the four RPTPs examined so far (21), RPTPζ is known to be the most sensitive to PTPI-IV (IC50 = 6.4 μM), so we next examined molecular expression of RPTPζ in MAFs and C127 cells by RT-PCR. As shown in Fig. 8A, expression of RPTPζ mRNA was observed in MAFs isolated from WT mice but not in MAFs from RPTPζ-KO mice. The band corresponding to RPTPζ mRNA was sequenced and confirmed to be identical to its known sequence. However, RT-PCR studies failed to detect RPTPζ mRNA in C127 cells (Fig. 8B).
Involvement of RPTPζ in excision-induced activation of maxi-anion channel in primary skin fibroblasts.
We then tested a possible involvement of RPTPζ in the mechanism of excision-induced activation of the maxi-anion channel in MAFs. Membrane patches derived from WT MAFs responded to patch excision in ATP-free solution containing 5 mM total Mg2+ with robust activation of maxi-anion channels (Fig. 9A). When the excision-induced activation of the maxi-anion channel in MAFs from RPTPζ-KO mice was examined, it was found that the channel activation rate was much slower (Fig. 9A) and the NPo values were significantly lower (Fig. 9B) compared with those in MAFs from WT mice. To confirm that the effect was specific to the RPTPζ knockout, the WT RPTPζ construct was transfected into MAFs derived from RPTPζ-KO mice. Transfection of WT RPTPζ, but not a dominant-negative mutant, into the MAFs derived from RPTPζ-KO mice largely restored the maxi-anion channel activation rate (Fig. 9A) and the NPo values (Fig. 9B) to levels comparable to those in the WT MAFs. The transfection of another closely related phosphatase, RPTPγ, did not significantly alter activation of the maxi-anion channel in MAFs derived from RPTPζ-KO mice (Fig. 9). These results strongly suggest that RPTPζ represents an important component of the mechanism of excision-induced activation of the maxi-anion channel in MAFs. It should be noted, however, that the maxi-anion channel current was not completely eliminated in RPTPζ-KO MAFs. It is thus possible that some other type(s) of PTP could also be involved in dephosphorylation of the maxi-anion channel protein or its accessory protein on membrane patch excision from primary skin fibroblasts.
Role of tyrosine dephosphorylation in excision-induced activation of maxi-anion channel.
Maxi-anion channel activity has been observed very rarely in the on-cell mode but has been frequently observed after membrane patch excision. This fact suggests the existence of some regulatory mechanism(s) that keeps the channel in the inactive closed state under resting nonstimulated conditions. Since a Ca2+ ionophore, A-23187, was reported to activate the maxi-anion channel in cell-attached patches in several cell types (25, 27, 30), we first hypothesized that a low (submicromolar) concentration of cytosolic Ca2+ was the factor keeping the channel closed. However, excision-induced activation of the maxi-anion channel could be observed even when the intracellular side of patch membranes was exposed to a low-Ca2+, cytoplasm-approximating electrolyte solution containing 100 nM free Ca2+ in C127 cells (Fig. 1) and primary skin fibroblasts (Fig. 6). Next, we hypothesized that cytosolic MgATP prevents the maxi-anion channel from being activated. Indeed, supplementing the intracellular solution with MgATP at a physiological level (0.636 mM) completely abolished excision-induced activation of the maxi-anion channel in inside-out patches (Fig. 2). In contrast, the nonhydrolyzable analog of MgATP, Mg-AMP-PNP, failed to suppress activation of the maxi-anion channel (Fig. 2), suggesting that phosphorylation of the maxi-anion channel or its regulatory subunit keeps the channel in the inactive closed state.
An involvement of phosphorylation in regulation of the maxi-anion channel was first suggested by Saigusa and Kokubun (47) in cell-attached patches on cultured vascular smooth muscle cells. They showed that inhibiting the activity of PKC with a serine/threonine kinase inhibitor, H-7, activated maxi-anion channels. Similarly, another PKC inhibitor, polymyxin B, was found to activate the maxi-anion channel in cell-attached patches on pig aortic endothelial cells (19). However, a PKC activator, phorbol 12,13-didecanoate (PDD) or 1-oleoyl 2-acetyl glycerol (OAG), was shown to activate the maxi-anion channel in cell-attached patches on rabbit RCCT-28 cells (52). In contrast, in the present study we found that H-7 did not significantly affect the excision-induced activation of the channel under conditions favoring the phosphorylated state (Fig. 3). On the other hand, application of broad-spectrum PTP inhibitors, such as AG18 and genistein, was found to rescue the channel from inactivation induced by MgATP (Fig. 3).
If phosphorylation of the channel, or its regulatory protein, keeps the channel closed, activation of the maxi-anion channel might be linked to a dephosphorylation reaction. Indeed, Diaz et al. (10) showed that antiestrogen-activated maxi-anion channels in C1300 mouse neuroblastoma cells were inhibited by OA, which is known to inhibit most serine/threonine protein phosphatases, including PP1, PP2, PP3, PP4, PP5, and PP6, at 10–100 nM. In the present study, however, excision-induced activation of this channel in C127 cells was totally insensitive to 100 nM OA (Fig. 4). Furthermore, PP2B-specific inhibitors CsA (1 μM) and ascomycin (FK520 1 μM) were also found to be ineffective in blocking channel activation (Fig. 4). Also, involvement of the remaining serine/threonine protein phosphatases, Mg2+-dependent PP2C and Mg2+- and Ca2+-dependent PP7, could be ruled out, because our preliminary experiments showed that removal of both Mg2+ and Ca2+ from an intracellular solution with 1 mM EDTA failed to prevent excision-induced activation of the maxi-anion channel (A. H. Toychiev and Y. Okada, unpublished data; n = 5). In contrast, excision-induced activation of the maxi-anion channel in both C127 cells and primary skin fibroblasts was abolished by a TPhIC, which is a mixture of orthovanadate, molybdate, tartrate, and imidazole, and was markedly inhibited by orthovanadate alone (1 mM) or another broad-spectrum PTP inhibitor, pBTM (0.1 mM) (Figs. 5 and 6). These pharmacological studies suggest an involvement of a tyrosine, but not serine/threonine, protein phosphatase in excision-induced activation of the maxi-anion channel in mouse fibroblasts.
Possible involvement of receptor-type PTPs in excision-induced activation of maxi-anion channel.
Protein tyrosine phosphorylation and dephosphorylation by protein tyrosine kinases (PTKs) and PTPs play a major role in cellular signaling, affecting protein-protein interactions and enzymatic activities (2, 22, 64). There are 107 known human PTPs (2). However, 11 of these are catalytically inactive, 2 dephosphorylate mRNAs, and 13 dephosphorylate inositol phospholipids. Thus only 81 PTPs are active protein phosphatases with the ability to dephosphorylate phosphotyrosine (2). The PTP family is subdivided into low-molecular-weight cytosolic PTPs and high-molecular-weight transmembrane RPTPs (17, 60). Since we used the inside-out mode with excised patches in our patch-clamp experiments, we hypothesized that RPTPs would be better candidates for dephosphorylating the channel or its regulatory protein, because they are transmembrane proteins and would likely be retained in the membrane patches even after excision. The present pharmacological studies failed to identify the RPTP responsible for the dephosphorylation involved in excision-induced activation of the maxi-anion channel in C127 cells. Channel activation was not affected by bpV, the CD45 inhibitor that inhibits RPTPβ, and PTPI-IV, which inhibits RPTPζ, RPTPμ, RPTPε, and RPTPδ (Fig. 5). However, in mouse primary skin fibroblasts, channel activation was significantly suppressed by PTPI-IV (Fig. 6). PTPI-IV is known to be most effective in inhibiting RPTPζ (21), which is a RPTP expressed as a chondroitin sulfate proteoglycan (28, 38). RPTPζ is known to regulate ion channels, interacting with the β1-subunit of the voltage-gated Na+ channel (41), and is known to be mainly distributed in both glia and neurons of the peripheral and central nervous systems (7, 34, 56). Significant expression of RPTPζ was also found in human colorectal mucosae (65) and mouse gastric gland (18) as well as in mouse lung and bone, especially in differentiated osteoblasts (49). The RT-PCR study presented here demonstrated that RPTPζ mRNA is expressed in MAFs in primary culture but not in mouse mammary C127 cells (Fig. 8). Excision-induced activation of the maxi-anion channel in MAFs was then found to be reduced by ∼50% by knockout of the RPTPζ gene; furthermore, channel activation was restored by transfection of WT RPTPζ, but not a dominant-negative mutant of RPTPζ or WT RPTPγ (Fig. 9). Thus, in mouse skin fibroblasts, RPTPζ is likely to participate, at least in part, in the dephosphorylation-mediated activation of the maxi-anion channel. However, further investigation is required to precisely identify all the PTP types that are responsible for the dephosphorylation that takes place in excision-induced activation of the maxi-anion channel in mouse fibroblasts, especially in C127 cells. Recent studies on voltage-gated cation-selective channels provided firm evidence that kinases and phosphatases are assembled into supramolecular signaling complexes conjugated with ion channel subunits (9). Our finding that the kinase and phosphatase activities are retained in the excised inside-out patches would suggest that the maxi-anion channel may form a similar signaling complex that contains a receptor tyrosine phosphatase(s).
Relevance to activation mechanism of maxi-anion channel in intact cell membranes.
The present study showed that the maxi-anion channel was activated even in the cell-attached patches on skin fibroblasts pretreated with a broad-spectrum tyrosine kinase inhibitor, AG18 or genistein (Fig. 7, A and B), or in the whole cell mode in MAFs dialyzed with ATP-free solution supplemented with AG18 (Fig. 7, C and D). Previous studies also showed that the maxi-anion channel is activated not only in cell-free excised membranes but also in intact cell membranes on application of a hypotonic challenge (5, 8, 12–14, 16, 23, 32, 42, 54), ischemic stress (14, 31, 32), and G protein-coupled receptor stimulation (19, 25, 52, 58, 59). Such activation of the maxi-anion channel in intact cell membranes has been shown to play multiple roles in many important cell functions including cell volume regulation (43, 57), cell-to-cell signaling mediated by regulated release of ATP (4, 12, 14, 31, 33, 42) and glutamate (32), as well as programmed cell death (1, 15). Osmotic swelling induced by hypotonic stimulation may cause dilution of intracellular ATP in the immediate vicinity of the plasma membrane due to massive water influx, thereby leading to situations in which dephosphorylation processes may predominate over phosphorylation processes mediated by membrane-associated protein phosphatases and kinases. Ischemia is associated with ATP depletion, which should impair the phosphorylation-dephosphorylation balance by reducing the activity of protein kinases. Stimulation of some G protein-coupled receptors has been shown to result in PTP-mediated dephosphorylation of proteins in a number of cell types (35, 40, 61). It is therefore plausible that PTP-mediated dephosphorylation of the maxi-anion channel or its regulatory protein is also involved in activation of the channel in the intact membrane of cells subjected to hypotonic stimulation, ischemic stress, or receptor stimulation. In fact, our preliminary study showed that hypotonicity-induced activation of the maxi-anion channel in cell-attached patches on C127 cells was significantly suppressed by pretreatment with the broad-spectrum PTP inhibitor orthovanadate (1 mM for 10 min) (see Supplemental Fig. S1).1
In conclusion, we have demonstrated in this study that the mechanism of excision-induced activation of the maxi-anion channel involves PTP-mediated dephosphorylation of the channel or its auxiliary regulatory subunit in mouse fibroblastic cells.
This work was supported by Grants-in-Aid for Scientific Research (A) and (C) to Y. Okada and R. Z. Sabirov from the Ministry of Education, Culture, Sports, Science and Technology of Japan and by a grant for Bio-Molecular-Sensor Project from the National Institute of Natural Sciences to Y. Okada and M. Noda.
We thank Dr. Tatiana Sheiko (Baylor College of Medicine, Houston, TX) for advice on MAF preparation. We thank Dr. Elbert Lee for manuscript preparation and T. Okayasu for secretarial help as well as M. Ohara and K. Shigemoto for technical assistance.
↵1 The online version of this article contains supplemental material.
- Copyright © 2009 the American Physiological Society