Portal de Periódicos da CAPES

A Novel Monovalent Cation Channel Activated by Inositol Trisphosphate in the Plasma Membrane of Rat Megakaryocytes

1995; Elsevier BV; Volume: 270; Issue: 28 Linguagem: Inglês

10.1074/jbc.270.28.16638

1083-351X

B. Somasundaram, Martyn P. Mahaut‐Smith,

Ion Transport and Channel Regulation

The activation of a monovalent cation current was studied in rat megakaryocytes using patch clamp techniques combined with photometric measurements of intracellular concentrations of Ca2+ ([Ca2+]i) and Na+. ADP evoked a release of [Ca2+]i and transiently activated a monovalent cation-selective channel, which, at negative potentials and under physiological conditions, would be expected to carry an inward Na+ current. The single channel conductance, estimated by noise analysis from whole cell currents at −50 to −60 mV was 9 picosiemens. Thapsigargin-induced [Ca2+]i increases failed to stimulate the monovalent cation current, suggesting that neither [Ca2+]i nor the depletion of internal Ca2+ stores were activators of this conductance. However, buffering of [Ca2+]i changes with 1,2-bis-(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid showed that both activation and inactivation of the current were accelerated by a rise in [Ca2+]i. The monovalent cation conductance was activated by internal perfusion with inositol 1,4,5-trisphosphate, both in the presence and in the absence of a rise in [Ca2+]i. Internal perfusion with inositol 2,4,5-trisphosphate, the poorly metabolizable isomer of inositol trisphosphate, similarly activated the monovalent cation current, whereas 1,3,4,5-tetrakisphosphate neither activated a current nor modified the ADP-induced monovalent current. Heparin, added to the pipette, blocked activation of the channel by ADP. The intracellular concentration of Na+, monitored by sodium-binding benzofuran isopthalate, increased by 10-20 mM in response to ADP under pseudophysiological conditions. We conclude the existence of a novel nonselective cation channel in the plasma membrane of rat megakaryocytes, which is activated by IP3 and can lead to increases in cytosolic Na+ after stimulation by ADP. The activation of a monovalent cation current was studied in rat megakaryocytes using patch clamp techniques combined with photometric measurements of intracellular concentrations of Ca2+ ([Ca2+]i) and Na+. ADP evoked a release of [Ca2+]i and transiently activated a monovalent cation-selective channel, which, at negative potentials and under physiological conditions, would be expected to carry an inward Na+ current. The single channel conductance, estimated by noise analysis from whole cell currents at −50 to −60 mV was 9 picosiemens. Thapsigargin-induced [Ca2+]i increases failed to stimulate the monovalent cation current, suggesting that neither [Ca2+]i nor the depletion of internal Ca2+ stores were activators of this conductance. However, buffering of [Ca2+]i changes with 1,2-bis-(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid showed that both activation and inactivation of the current were accelerated by a rise in [Ca2+]i. The monovalent cation conductance was activated by internal perfusion with inositol 1,4,5-trisphosphate, both in the presence and in the absence of a rise in [Ca2+]i. Internal perfusion with inositol 2,4,5-trisphosphate, the poorly metabolizable isomer of inositol trisphosphate, similarly activated the monovalent cation current, whereas 1,3,4,5-tetrakisphosphate neither activated a current nor modified the ADP-induced monovalent current. Heparin, added to the pipette, blocked activation of the channel by ADP. The intracellular concentration of Na+, monitored by sodium-binding benzofuran isopthalate, increased by 10-20 mM in response to ADP under pseudophysiological conditions. We conclude the existence of a novel nonselective cation channel in the plasma membrane of rat megakaryocytes, which is activated by IP3 and can lead to increases in cytosolic Na+ after stimulation by ADP. Megakaryocytes are large cells located in the bone marrow that are responsible for producing blood platelets, yet little is known of the cellular mechanisms underlying their function. Uneyama and co-workers (1Uneyama C. Uneyama H. Akaike N. J. Physiol. 1993; 470: 731-749Crossref PubMed Scopus (33) Google Scholar, 2Akaike N. Uneyama H. News Physiol. Sci. 1994; 9: 49-53Google Scholar) have shown that rat megakaryocytes possess a novel purinergic receptor, which recognizes the ionized forms of ATP and ADP. Stimulation of this receptor leads to sustained oscillations of intracellular Ca2+ concentration ([Ca2+]i)1 1The abbreviations used are: [Ca2+]cytosolic Ca2+ concentrationSBFIsodium-binding benzofuran isopthalateIP3myo-inositol trisphosphate (with positional determinants of the phosphate groups as specified)IP4myo-inositol 1,3,4,5-tetrakisphosphateBAPTA1,2-bis-(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acidNMDGn-methyl-D-glucamine[Na+]cytosolic Na+ concentrationI-Vcurrent-voltage. and activation of Ca2+-dependent K+ channels. We have recently found that ATP also activates Na+ and Ca2+-permeable channels in rat megakaryocytes via two distinct classes of purinergic receptor(3Somasundaram B. Mahaut-Smith M.P. J. Physiol. 1994; 480: 225-231Crossref PubMed Scopus (47) Google Scholar). One receptor is activated by ATP but not noticeably by ADP and causes rapid, transient opening of a nonselective cation channel. A second purinoceptor is stimulated by ATP−4 and ADP−3 and activates both a monovalent cation-selective channel and a channel highly selective for Ca2+. Stimulation via this second receptor also causes a release of Ca2+ from intracellular stores and is most likely the same receptor as that responsible for the generation of Ca2+ oscillations in the experiments of Uneyama et al.(1Uneyama C. Uneyama H. Akaike N. J. Physiol. 1993; 470: 731-749Crossref PubMed Scopus (33) Google Scholar). The Ca2+-selective conductance is activated by depletion of internal Ca2+ stores via an as yet undetermined messenger and is indistinguishable from the store-regulated Ca2+ current found in other nonexcitable cell types(4Hoth M. Penner R. Nature. 1992; 355: 353-356Crossref PubMed Scopus (1506) Google Scholar, 5Fasolato C. Innocenti B. Pozzan T. Trends Pharmacol. Sci. 1994; 15: 77-83Abstract Full Text PDF PubMed Scopus (440) Google Scholar, 6Randriamampita C. Tsien R.Y. Nature. 1993; 364: 809-814Crossref PubMed Scopus (790) Google Scholar). The monovalent cation-selective current is also activated at the time of internal Ca2+ release via IP3 or another inositol lipid metabolite(3Somasundaram B. Mahaut-Smith M.P. J. Physiol. 1994; 480: 225-231Crossref PubMed Scopus (47) Google Scholar). This channel is particularly interesting because, at resting membrane potentials, the current will be mostly carried by Na+, and changes in [Na+]i have been proposed to play a role in the spreading reaction in megakaryocytes(7Leven R.M. Mullikin W.H. Nachmias V.T. J. Cell Biol. 1983; 96: 1234-1240Crossref PubMed Scopus (16) Google Scholar). The spreading reaction may represent the physiological mechanism whereby megakaryocytes invade the bone sinusoids to reach the blood vessels and release platelets(8Penington D.G. Gordon L. Platelets in Biology and Pathology. Elsevier Science Publishers B.V., Amsterdam1981: 20-41Google Scholar). In the present study we have used a combination of patch clamp and fluorescent indicators of [Ca2+] and [Na+] to study the monovalent cation-selective current activated by ADP and inositol phosphates in the plasma membrane of rat megakaryocytes. cytosolic Ca2+ concentration sodium-binding benzofuran isopthalate myo-inositol trisphosphate (with positional determinants of the phosphate groups as specified) myo-inositol 1,3,4,5-tetrakisphosphate 1,2-bis-(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid n-methyl-D-glucamine cytosolic Na+ concentration current-voltage. Adult male Wistar rats weighing 200-300 g were killed by cervical dislocation. Bone marrow from the femoral and tibial bones was removed by gentle lavage using a standard external solution containing 20 mg ml−1 apyrase and 0.1% bovine serum albumin. After filtration through a fine cotton mesh, the suspension was spun and washed twice before storage in the same standard solution. Megakaryocytes were distinguished from other bone marrow cells by their distinctive size (30-60 μm) and multilobular nucleus. Recordings were made at room temperature (20-23°C) within 3-24 h of isolation. The standard external solution contained 140 mM NaCl, 5 mM KCl, 1 mM MgCl2, 2 mM CaCl2, 10 mM glucose, 10 mM HEPES (pH 7.4 adjusted with Tris). K+-, Na+-, and Ca2+-free external media were obtained by replacing these ions with Cs+, NMDG+, and Mg2+, respectively. For low Cl− external solutions, all Cl−, except for that added with the divalent cation salts, was replaced by gluconate. In conventional whole cell patch recordings, Cs+ replaced K+ in order to block K+ currents and contained 140 mM cesium gluconate, 5 mM NaCl, 2 mM MgCl2, 10 mM HEPES, 0.2 mM Na2GTP, 0.05 mM K5·fura-2, (pH 7.4 adjusted with Tris). 0.1 mM (NH4)4·SBFI replaced the K5·fura-2 in experiments where [Na+]i and membrane current were measured simultaneously. For internal dialysis of inositol phosphates, the pipette tip was dipped in inositol phosphate-free pipette solution and then backfilled with pipette solution containing 10 μM inositol 1,4,5-trisphosphate (1,4,5-IP3), 50 μM 2,4,5-IP3, or 20 μM inositol 1,3,4,5-tetrakisphosphate (IP4). Highly calcium-buffered pipette solution was obtained by replacing 40 mM cesium gluconate with 10 mM Cs4·BAPTA. In nystatin-perforated patch recordings, the pipette contained 100 mM KCl, 40 mM K2SO4, 1 mM MgCl2, 10 mM HEPES (pH 7.4 adjusted with Tris), 150 μg ml−1 nystatin. In these perforated patch experiments, cells were loaded with SBFI by incubation with 10 μM SBFI-acetoxymethyl ester with 0.04% pluronic acid for 60 min at room temperature. Fura-2, SBFI, Cs4·BAPTA and pluronic F-127 were from Molecular Probes, Inc. (Eugene, OR). 1,4,5-IP3, 2,4,5-IP3, and IP4 were a gift of Dr. R. F. Irvine (Biotechnology and Biological Sciences Research Council, Babraham Institute, Cambridge, UK). All other reagents were from Sigma. A pressure injector was used to administer agonists from a patch pipette placed 150 μm from the cell. The time delay for arrival of agonists at the cell was measured and subtracted. Patch clamp experiments were performed in conventional whole cell (9Hamill O.P. Marty A. Neher E. Sakmann B. Sigworth F.J. Pfluegers Archiv. Eur. J. Physiol. 1981; 391: 85-100Crossref PubMed Scopus (15371) Google Scholar) or nystatin-perforated patch (10Horn R. Marty A. J. Gen. Physiol. 1988; 92: 145-159Crossref PubMed Scopus (1479) Google Scholar) configurations by means of an Axopatch 200A patch clamp amplifier (Axon Instruments, Inc., Foster City, CA). Pipettes were pulled from borosilicate glass tubing (Clark Electromedical Instruments) and had filled resistances of 2-3 megaohms. Series resistances were in the range of 7-30 megaohms, and 40-60% series resistance compensation was employed. Membrane currents during voltage ramps (0.6 mV/ms) were filtered at 2 kHz and sampled at 100 μs using Axon Instruments hardware (Digidata 1200) and software (pClamp). Where stated, voltage ramps prior to agonist application were used to subtract linear leak currents from ADP-evoked currents. A holding potential of −40 mV was used in many experiments because this was close to the membrane potential measured in perforated patch experiments under pseudophysiological conditions (see Fig. 6). A less negative potential permitted longer recordings, thus −20 mV was used in experiments where large currents were detected; however, holding potentials in the range of −70 to 40 mV were used in some experiments, depending upon the conditions, to check for measurable currents. For noise analysis, currents were filtered at 0.5 kHz and sampled at 2.5 kHz. Currents were also acquired at a rate of 60 Hz (filtered at approximately 30 Hz) by the Cairn spectrophotometer (see below) for simultaneous display alongside the fura-2 and SBFI fluoresence and signal-indicating agonist injection. Liquid junction potentials were measured by reference to a 3 M KCl agar bridge, and membrane potentials were adjusted accordingly in conventional but not nystatin whole cell recordings. Fura-2 and SBFI fluoresence measurements were made by single cell photometry using a Cairn spectrophotometer system (Cairn Research Ltd., Kent, UK) coupled to a Nikon Diaphot inverted microscope. Excitation light passed through a spinning filter wheel assembly containing four 340-nm and two 380-nm bandpass excitation filters. Emitted light (400-600 nm) was selected by two dichroic filters and further filtered by a 485-nm long pass filter. The combined output from all 340 and 380 nm excitation filters provided a 340/380 nm ratio for each revolution of the filter wheel. The signal was then averaged to obtain a ratio value every 67 ms. Background and cell autofluoresence was measured in the cell-attached recording mode and subtracted to give fura-2 or SBFI fluoresence. [Ca2+]i was calculated according to Grynkiewicz et al.(11Grynkiewicz G. Poenie M. Tsien R.Y. J. Biol. Chem. 1985; 260: 3440-3450Abstract Full Text PDF PubMed Scopus (80) Google Scholar) using a dissociation constant for fura-2 of 250 nM(12Williams D.A. Fay F.S. Cell Calcium. 1990; 11: 75-83Crossref PubMed Scopus (227) Google Scholar). Under conditions where SBFI was introduced into the cell through the recording pipette, background-corrected 340/380 ratios were used to provide an indication of [Na+]i changes. In experiments where SBFI was loaded from its acetoxymethyl ester, [Na+]i was clamped at different levels by perfusing solutions of known extracellular Na+ concentrations in the presence of a mixture of sodium ionophores (5 μM each of gramicidin, nigericin, and monensin; Ref. 13Harootunian A.T. Kao J.P.Y. Eckert B.K. Tsien R.Y. J. Biol. Chem. 1989; 264: 19458-19467Abstract Full Text PDF PubMed Google Scholar). The extracellular solutions were made from appropriate mixtures of high Na+ and K+ solutions. The former consisted of 110 mM sodium gluconate, 30 mM NaCl, 2 mM CaCl2, 1 mM MgCl2, 10 mM Na-HEPES (pH 7.4). The high K+ solution was identical except for substitution of all Na+ for K+. A plot of SBFI 340/380 nm intensity ratio versus [Na+] was linear in the range of 0-40 mM. The experimental 340/380 ratio values fell within this linear range, and therefore [Na+]ivalues were directly obtained from the calibration curve. At negative potentials and under conditions that blocked K+ currents (see "Materials and Methods"), 5 μM ADP activated a transient inward current and a concurrent large increase in [Ca2+]i (Fig. 1A). The initial [Ca2+]i increase reached a peak of 0.5-1.5 μM within 1-2 s and then returned to basal levels (approximately 50-100 nM) in the continued presence of ADP or was followed by further smaller increases in [Ca2+]i that sometimes fused to give a plateau, as shown in the cell of Fig. 1. In the absence of external calcium, an inward current and [Ca2+]iincrease were still activated by ADP(3Somasundaram B. Mahaut-Smith M.P. J. Physiol. 1994; 480: 225-231Crossref PubMed Scopus (47) Google Scholar). This suggests that the rise in [Ca2+]i is at least partly due to the release of internal stores and that the current is not selective for Ca2+. However, variability between cells did not allow us to quantify the extent to which Ca2+ influx contributed to the response. To further examine the conductance changes in reponse to ADP, membrane currents were recorded during 0.6 mV/ms voltage ramps within the range of potentials −100 to 90 mV. A ramp applied prior to agonist application (Fig. 1B, trace a) was used to subtract background currents from ramp currents obtained during the ADP-evoked transient (Fig. 1B, trace b). Fig. 1C shows the difference current (b-a) representing the I-V relationship for the ADP-evoked conductance. In 140 mM Na+ external solution and 140 mM cesium gluconate internal solution, the I-V relationship reversed at about −5 mV, was reasonably linear over the voltage range −90 to 40 mV, and, in most cells, displayed a distinct increase in slope at more positive potentials. The I-V relationship obtained by ramps at different times during the ADP-evoked current differed only in amplitude and not in reversal potential, suggesting that the response was due to activation of a single ionic conductance. The ionic selectivity of the ADP-evoked conductance was investigated by substitution of internal and external ions. In Fig. 2, ADP-evoked I-V relationships are shown for each of four ionic conditions, with the membrane currents and Ca2+ responses at a single negative holding potential in the insets. These results were obtained from the first ADP-evoked response in four different cells and were confirmed in at least five cells for each condition. Replacement of the majority of the internal and external Cl− had little effect on the current and Ca2+ response at −40 mV or on the difference I-V relationship (Fig. 2A) compared with corresponding Cl−-containing salines (see for example Fig. 1C). Therefore, the ADP-evoked conductance does not appear to be significantly permeant to anions. On the other hand, replacement of all internal and external monovalent cations with the impermeant cation NMDG abolished all current at −40 mV, in the presence of an ADP-evoked Ca2+ response (Fig. 2B, inset). No significant current developed within the voltage range −80 to 80 mV, as shown by the overlapping ramp currents before and during the Ca2+ response in Fig. 2B, indicating that the ADP-evoked current is carried by monovalent cations. 2 mM Ca2+ and 1 mM Mg2+ were present in the external media throughout, which indicates little or no permeability to divalent cations at these physiological concentrations. Increasing external Ca2+ decreases the level of ADP3−(1Uneyama C. Uneyama H. Akaike N. J. Physiol. 1993; 470: 731-749Crossref PubMed Scopus (33) Google Scholar) and abolishes the ADP response(3Somasundaram B. Mahaut-Smith M.P. J. Physiol. 1994; 480: 225-231Crossref PubMed Scopus (47) Google Scholar); therefore we were unable to increase the external Ca2+ concentration to test for any underlying permeability to Ca2+. The absence of membrane current in the experiment of Fig. 2B was not due to direct block by NMDG+ or the requirement of permeant ion on both sides of the membrane, because current was activated by ADP unidirectionally when either the external or internal NMDG was replaced by Cs+ (Fig. 2, C and D). In the presence of symmetrical 140 mM Cs+, the ADP-evoked I-V relationship was similar to that observed in Na+/Cs+ salines and reversed at about −5 mV (not shown), indicating similar Na+ and Cs+ permeabilities. Clear single channel events could not be clearly resolved during the off-phase of the ADP-evoked whole cell current. This was due to the noise generated by the high capacitance of the megakaryocyte (20-100 picofarads), although it also indicates that the ADP-evoked events are of relatively short duration(14Marty A. Neher E. Sakmann B. Neher E. Single Channel Recording. Plenum Publishing Corp., New York1983: 107-122Google Scholar). We therefore turned to noise analysis to obtain an estimate of single channel conductance. When the number of channels opening is small, the variance (σ2) of the current is linearly related to the mean current with a slope equal to the single channel current(15Anderson C.R. Stevens C.F. J. Physiol. 1973; 235: 655-691Crossref PubMed Scopus (573) Google Scholar, 16Sigworth F.J. J. Physiol. 1980; 307: 97-129Crossref PubMed Scopus (555) Google Scholar). For this analysis we used cells that displayed a small ADP-evoked current, as shown in Fig. 3A. This reduced response was most likely the result of receptor desensitization (e.g. by ADP and ATP from damaged cells during the cell preparation) rather than a low number of total channels. As expected, the variance of the mean of the whole cell current increased in response to ADP (Fig. 3A, lower record). A plot of variance against the mean current during the ADP response could be well fitted by a linear relationship with a slope of 0.49 pA. The holding potential was −60 mV, and the reversal potential under these conditions was −5 mV, giving a single channel conductance of approximately 9 picosiemens. Within the range of holding potentials −50 to −60 mV, the average conductance was 8.6 ± 0.4 picosiemens (n = 3). This analysis assumes the existence of a uniform single channel conductance and that all channels open independently. The estimate must be considered a lower estimate for the single channel conductance, and, in practice, direct measurements of channel conductance are higher(17Dart C. Standen N.B. J. Physiol. 1993; 471: 767-786Crossref PubMed Scopus (173) Google Scholar). The close association of the ADP-evoked current with the rise in [Ca2+]isuggested that this current may be modulated by Ca2+ or a Ca2+-dependent process. To test this, the current was activated by ADP when internal Ca2+ levels were strongly buffered by 10 mM BAPTA in the pipette saline. In order to eliminate the store-dependent inward current that is amplified under such conditions(3Somasundaram B. Mahaut-Smith M.P. J. Physiol. 1994; 480: 225-231Crossref PubMed Scopus (47) Google Scholar), Ca2+ was omitted from the external saline. Fig. 4A compares the [Ca2+]i changes and membrane currents activated by a 30-s application of 5 μM ADP in normal (Fig. 4Ai) and enhanced (Fig. 4Aii) calcium buffering in nominally Ca2+-free salines. With low buffering, the current was activated 0.6 ± 0.3 s (n = 10) after ADP application, peaked within 1-3 s and inactivated to 10% of peak current after 3.5 ± 2 s (n = 10). In most cells tested the current was inactivated well before [Ca2+]i returned to basal levels (Fig. 4Ai). In the absence of any increase in [Ca2+]i, ADP could still activate an inward current, although its kinetics were very different (Fig. 4Aii). The current was activated more slowly, reaching a peak after 5-10 s, and was inactivated more slowly (time to 10% of peak current was 18.5 ± 6 s; n = 6) compared with the currents activated in the unbuffered cells. This implies that Ca2+ or a Ca2+-dependent process, although not required for activation of the ADP-evoked current, accelerates the rate of both activation and inactivation. In order to test if this current could be activated by a rise in [Ca2+]i alone, [Ca2+]i was continuously elevated to micromolar levels using the endoplasmic Ca-ATPase inhibitor thapsigargin (Fig. 4B). This agent results in a permanent loss of Ca2+ from IP3-sensitive stores and store-dependent (capacitative) calcium entry(18Thastrup O. Dawson A.P. Scharff O. Foder B. Cullen P.J. Drobak B.K. Bjerrum P.J. Christensen S.B. Hanley M.R. Agents Actions. 1989; 27: 17-23Crossref PubMed Scopus (446) Google Scholar, 19Putney Jr., J.W. Cell Calcium. 1986; 7: 1-12Crossref PubMed Scopus (2153) Google Scholar). The absence of any current in the megakaryocyte after thapsigargin treatment in Ca2+-free saline suggests that the ADP-evoked current cannot be triggered by an increase in [Ca2+]i alone; neither is this current activated as a result of depletion of internal Ca2+ stores. Repeated exposures to ADP could reactivate the monovalent cation current, provided there was an interval of 1-2 min between successive applications. Under conditions of high internal Ca2+ buffering, thus removing Ca2+-dependent inhibition of the current, both activation and inactivation of the current became progressively slower with repeated ADP additions, as shown in the experiment of Fig. 4C. This suggested that dialysis of the cytoplasm removes factors responsible for activation and inactivation of the current. These factors may, for example, generate and metabolize the second messenger involved in channel activation. Release of internal Ca2+ in nonexcitable cells appears to ubiquitously involve an increase in cytoplasmic IP3 levels and IP3-dependent Ca2+ stores(20Berridge M.J. Nature. 1993; 361: 315-325Crossref PubMed Scopus (6205) Google Scholar). IP3 is therefore a candidate for the second messenger involved in the activation of the monovalent cation channel in the rat megakaryocyte. A previous study provided preliminary evidence for a role for IP3, because dialysis with 1,4,5-IP3 activated a monovalent cation current with similar characteristics to that activated by ADP(3Somasundaram B. Mahaut-Smith M.P. J. Physiol. 1994; 480: 225-231Crossref PubMed Scopus (47) Google Scholar). However, it was not shown whether IP3 was acting alone or in synergism with other second messengers. Fig. 5A compares the currents activated by dialysis of 1,4,5-IP3 with (5Ai) and without (5Aii) an increase in intracellular Ca2+ levels. For these dialysis experiments, 3 mM 4-aminopyridine was added to the external medium to accelerate the blockade of K+ currents. As shown by the current-voltage relationships acquired at three timepoints during each experiment, a current similar to that observed with ADP was activated by 1,4,5-IP3 and did not require an increase in [Ca2+]i. 1,4,5-IP3 is rapidly metabolized to other inositol lipid products, including IP4(20Berridge M.J. Nature. 1993; 361: 315-325Crossref PubMed Scopus (6205) Google Scholar). Its isomer, 2,4,5-IP3 is also active at IP3 receptors on the Ca2+ stores, although at higher concentrations(21Irvine R.F. Brown K.D. Berridge M.J. Biochem. J. 1984; 222: 269-272Crossref PubMed Scopus (124) Google Scholar), and is experimentally useful because it is a poor substrate for the 1,4,5-IP3-kinase. Therefore 2,4,5-IP3 can be used to stimulate IP3-dependent processes, whereas cytoplasmic levels of IP4 remain low. As shown in Fig. 5B (i, left panel), dialysis with 2,4,5-IP3 released internal Ca2+ and activated the monovalent cation current in a manner indistinguishable from that seen with 1,4,5-IP3. In the presence of high Ca2+ buffering power, 2,4,5-IP3 evoked the inward current as expected and application of ADP failed to activate further current (Fig. 5Bii). These results suggest that IP3 on its own is sufficient to activate the monovalent cation current without a need for other inositol lipid products such as IP4. Furthermore, following internal dialysis with 20 μM IP4, no inward current was observed and subsequent exposure to ADP produced a normal response (Fig. 5C). 1,4,5-IP3-activated channels located on the membrane of internal Ca2+ stores and those on the plasma membrane of olfactory receptor neurons are both blocked by heparin(22Supattapone S. Worley P.F. Barraban J.M. Snyder S.H. J. Biol. Chem. 1988; 263: 1530-1534Abstract Full Text PDF PubMed Google Scholar, 23Fadool D.A. Ache B.W. Neuron. 1992; 9: 907-918Abstract Full Text PDF PubMed Scopus (209) Google Scholar). Fig. 5D shows that internal perfusion of megakaryocytes with 10 mg ml−1 heparin for 5 min virtually abolished both the [Ca2+]i and membrane current response to ADP. Cd2+ has also been shown to block the olfactory neuron IP3-dependent plasma membrane channel(24Okada Y. Teeter J.H. Restrepo D. J. Neurophys. 1994; 71: 595-601Crossref PubMed Scopus (49) Google Scholar); however, neither Cd2+ nor Zn2+, added to the bath at 1 mM, affected the monovalent cation currents activated by ADP, 1,4,5-IP3, or 2,4,5-IP3 (not shown). Tetrodotoxin, a blocker of voltage-dependent Na+ channels, was also ineffective at concentrations up to 5 μM added to the bath saline (data not shown). In physiological salines, given a normal negative resting potential, the IP3 and ADP-evoked current would be inward and carried mainly by Na+. To detect whether this conductance can result in significant changes in [Na+]i, the Na+-sensitive indicator, SBFI(25Minta A. Tsien R.Y. J. Biol. Chem. 1989; 264: 19449-19457Abstract Full Text PDF PubMed Google Scholar), was added to the pipette saline in place of fura-2 and 1,4,5-IP3 dialyzed from the pipette (Fig. 6A). The inward current that developed at the holding potential of −40 mV, which we have shown above to be induced by 1,4,5-IP3, was associated with a gradual increase in [Na+]i. Depolarization to 0 or 20 mV prevented the [Na+]i increase, an effect that was fully reversible, although more negative potentials were required to produce similar rates of Na+ increase once a substantial increase in the 340/380 ratio signal had occurred. The above whole cell patch clamp experiments represent conditions that are far from physiological and dialyze important cytoplasmic factors. We therefore turned to the nystatin-perforated patch technique to further assess the magnitude of the ADP-evoked [Na+]i increase. In these experiments, represented by that in Fig. 6B, a K+-based pipette saline and current-clamp conditions were also used in an attempt to further mimic physiological conditions. SBFI was loaded prior to patch clamp by incubation with the acetoxymethyl ester (see "Materials and Methods"). Application of 5 μM ADP produced a regular oscillation in membrane potential from the resting level of −40 mV to −75 mV (approximately 6 times/min), known to result from oscillations of [Ca2+]i and activation of Ca2+-dependent K+ channels(1Uneyama C. Uneyama H. Akaike N. J. Physiol. 1993; 470: 731-749Crossref PubMed Scopus (33) Google Scholar, 2Akaike N. Uneyama H. News Physiol. Sci. 1994; 9: 49-53Google Scholar). In this cell, which is typical of 5 other experiments, [Na+]i increased gradually by 3-4 mM during the 3 min ADP application and then continued to increase after agonist removal. Further [Na+]i increases were observed in response to a second exposure to ADP. In 5 cells, after a 3-min application of 5 μM ADP, [Na+]i increased from a resting level of 15 ± 6 mM to 28 ± 13 mM, measured 2 min after removal of the agonist. The present study demonstrates that both extracellularly applied ADP and internally perfused 1,4,5-IP3 evoke a monovalent cation-selective current in rat megakaryocytes. The similarity between the I-V relationships, reversal potential, stimulation of Na+ influx, and failure of ADP to evoke a current on top of the IP3-induced response, suggests strongly that these two agents activate the same channel. Although there is no direct evidence for ADP-induced 1,4,5-IP3 production in megakaryocytes, ADP is known to stimulate 1,4,5-IP3 production in platelets (26Rink T.J. Sage S.O. Annu. Rev. Physiol. 1990; 52: 431-449Crossref PubMed Scopus (321) Google Scholar), and both ADP and IP3 induce a similar [Ca2+]i oscillation in the rat megakaryocyte (27Uneyama H. Uneyama C. Akaike N. J. Biol. Chem. 1993; 268: 168-174Abstract Full Text PDF PubMed Google Scholar). In addition, the lack of effect of ADP on [Ca2+]i after internal perfusion of heparin, a known blocker of IP3 receptors, suggests that this agonist acts via phospholipase C to increase intracellular IP3 and internal Ca2+ levels, as in many other nonexcitable cells (20Berridge M.J. Nature. 1993; 361: 315-325Crossref PubMed Scopus (6205) Google Scholar, 28Balla T. Catt K.J. Trends Endocrinol. Metab. 1994; 5: 250-255Abstract Full Text PDF PubMed Scopus (14) Google Scholar). The activation of the monovalent cation current by 2,4,5-IP3, a poor substrate for 1,4,5-IP3-kinase(21Irvine R.F. Brown K.D. Berridge M.J. Biochem. J. 1984; 222: 269-272Crossref PubMed Scopus (124) Google Scholar), and not by IP4 is strong evidence for direct stimulation by 1,4,5-IP3 rather than by any of its metabolites. The rate of inactivation of the current was reduced in the absence of a [Ca2+]i increase, which may be explained in part by slower hydrolysis of 1,4,5-IP3 because 1,4,5-IP3-kinase is calcium-dependent(29Biden T.J. Altin J.G. Karjalainen A. Bygrave F.L. Biochem J. 1988; 256: 697-701Crossref PubMed Scopus (22) Google Scholar). Direct inactivation of the current by calcium cannot be not ruled out, although this is unlikely because, in experiments where IP3 was introduced directly into the cells, fluctuations in [Ca2+]i had little or no effect on the IP3-dependent current(3Somasundaram B. Mahaut-Smith M.P. J. Physiol. 1994; 480: 225-231Crossref PubMed Scopus (47) Google Scholar). The reduced activation rate of the ADP-evoked current in the presence of BAPTA can be explained by the calcium dependence of phospholipase C activity because IP3 production is accelerated by an increase in [Ca2+]i(28Balla T. Catt K.J. Trends Endocrinol. Metab. 1994; 5: 250-255Abstract Full Text PDF PubMed Scopus (14) Google Scholar, 30Renard D. Poggioli J. Berthon B. Claret M. Biochem J. 1987; 243: 391-398Crossref PubMed Scopus (45) Google Scholar). The lack of effect of thapsigargin-induced rise in [Ca2+]i, which does not involve an increase in IP3(18Thastrup O. Dawson A.P. Scharff O. Foder B. Cullen P.J. Drobak B.K. Bjerrum P.J. Christensen S.B. Hanley M.R. Agents Actions. 1989; 27: 17-23Crossref PubMed Scopus (446) Google Scholar, 19Putney Jr., J.W. Cell Calcium. 1986; 7: 1-12Crossref PubMed Scopus (2153) Google Scholar), suggests that the monovalent cation current cannot be activated by [Ca2+]i alone. The observation that ADP-evoked currents inactivated more slowly and incompletely the longer a whole cell recording was made suggests loss by dialysis of a factor responsible for current inactivation. One likely candidate for this labile factor is 1,4,5-IP3 kinase. In many nonexcitable cells, including the megakaryocyte, IP3 stimulates a plasma membrane current indirectly by releasing Ca2+ from internal stores(3Somasundaram B. Mahaut-Smith M.P. J. Physiol. 1994; 480: 225-231Crossref PubMed Scopus (47) Google Scholar, 4Hoth M. Penner R. Nature. 1992; 355: 353-356Crossref PubMed Scopus (1506) Google Scholar). This pathway cannot account for the whole cell currents gated by internal perfusion of IP3 in this study, with low internal Ca2+ buffering, because thapsigargin, which releases internal Ca2+ without generation of IP3(18Thastrup O. Dawson A.P. Scharff O. Foder B. Cullen P.J. Drobak B.K. Bjerrum P.J. Christensen S.B. Hanley M.R. Agents Actions. 1989; 27: 17-23Crossref PubMed Scopus (446) Google Scholar), failed to elicit a significant current. Furthermore, the store-regulated Ca2+ currrent is highly selective for divalent cations, is blocked by Zn2+ and Cd2+, and is amplified by buffering of internal Ca2+ with BAPTA or EGTA(4Hoth M. Penner R. Nature. 1992; 355: 353-356Crossref PubMed Scopus (1506) Google Scholar, 31Hoth M. Penner R. J. Physiol. 1993; 465: 359-386Crossref PubMed Scopus (663) Google Scholar), none of which applied to the IP3-dependent response under our experimental conditions. 1,4,5-IP3 receptors have been found in the plasma membrane of T lymphocytes(32Khan A.A. Steiner J.P. Snyder S.H. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 2849-2853Crossref PubMed Scopus (94) Google Scholar), platelets (33Rengasamy A. Feinberg H. Biochem. Biophys. Res. Commun. 1988; 150: 1021-1026Crossref PubMed Scopus (22) Google Scholar), and olfactory cilia(34Kalinoski D.L. Aldinger S.B. Boyle A.G. Huque T. Marecek J.F. Prestwich G.D. Restrepo D. Biochem. J. 1992; 281: 449-456Crossref PubMed Scopus (68) Google Scholar), and channels activated by 1,4,5-IP3 have been identified in patch clamp recordings from Jurkat T cells(35Kuno M. Gardner P. Nature. 1987; 326: 301-304Crossref PubMed Scopus (503) Google Scholar), A431 cells(36Mozhayeva G.N. Naumov A.P. Kuryshev Y.A. J. Membr. Biol. 1991; 124: 113-126Crossref PubMed Scopus (30) Google Scholar), and olfactory neurons(23Fadool D.A. Ache B.W. Neuron. 1992; 9: 907-918Abstract Full Text PDF PubMed Scopus (209) Google Scholar, 24Okada Y. Teeter J.H. Restrepo D. J. Neurophys. 1994; 71: 595-601Crossref PubMed Scopus (49) Google Scholar, 37Restrepo D. Miyamoto T. Bryant B.P. Teeter J.H. Science. 1990; 249: 1166-1168Crossref PubMed Scopus (240) Google Scholar, 38Stengl M. J. Comp. Physiol. B Metab. Transp. Funct. 1994; 174: 187-194Google Scholar). In the T cell, the A431 cell, and insect or channel catfish olfactory neurons, only divalent cation currents were reported, implying a different selectivity from the IP3-gated channel in the megakaryocyte, which conducts little, if any Ca2+. Nonselective cation currents were activated by IP3 in rat olfactory neurons, although, unlike in the megakaryocyte, these were Cd2+-sensitive(24Okada Y. Teeter J.H. Restrepo D. J. Neurophys. 1994; 71: 595-601Crossref PubMed Scopus (49) Google Scholar). IP3 also gates ion channels in the membranes of internal organelles, including the sarcoplasmic reticulum (39Ferris C.D. Snyder S.H. Annu. Rev. Physiol. 1992; 54: 469-488Crossref PubMed Scopus (199) Google Scholar) and nucleus(40Mak D.-O.D. Foskett J.K. J. Biol. Chem. 1994; 269: 29375-29378Abstract Full Text PDF PubMed Google Scholar), which possess a higher permeability to calcium than monovalent cations. Thus, the IP3-dependent channel in the megakaryocyte may represent a new class of ion channel gated by this second messenger. The physiological function of the ADP-evoked monovalent cation channels in the rat megakaryocyte remains speculative. We could not detect any significant Ca2+ permeability, thus the conductance is unlikely to play a role in agonist-evoked Ca2+ signaling. Furthermore, the rat megakaryocyte has a store-dependent influx pathway that is highly selective for Ca2+(3Somasundaram B. Mahaut-Smith M.P. J. Physiol. 1994; 480: 225-231Crossref PubMed Scopus (47) Google Scholar, 4Hoth M. Penner R. Nature. 1992; 355: 353-356Crossref PubMed Scopus (1506) Google Scholar), and this is likely to account for most if not all of the Ca2+ influx that occurs during IP3-dependent Ca2+ release(3Somasundaram B. Mahaut-Smith M.P. J. Physiol. 1994; 480: 225-231Crossref PubMed Scopus (47) Google Scholar). Application of ADP caused a 10-20 mM increase in [Na+]i, which, considering the large volume of the megakaryocyte, amounts to a considerable Na+ influx. The continued, slow increase in [Na+]i, after removal of ADP, may be explained if IP3 levels remain elevated for some time. This certainly does appear to be the case because repetitive hyperpolarizations in the membrane potential, known to arise from IP3-induced Ca2+ release and activation of Ca2+-dependent K+ channels(27Uneyama H. Uneyama C. Akaike N. J. Biol. Chem. 1993; 268: 168-174Abstract Full Text PDF PubMed Google Scholar), continued for several minutes after ADP application. Agonist-evoked Na+ influx may outlast the Ca2+ responses if IP3 levels are higher at the plasma membrane, where this messenger is produced, than deeper in the cytoplasm near the Ca2+ stores or if the threshold for activation of the plasma membrane channel by IP3 is lower than for that of store Ca2+ channel. An alternative explanation is that the gradual increase in [Na+]i is due to slow equilibration of Na+ within the cytoplasm following IP3-evoked Na+ entry. This would imply a much greater increase in [Na+]i next to the plasma membrane and may be detectable by confocal ratiometric measurements of SBFI fluorescence. The monovalent cationic conductance that we report here is of particular interest since a previous study by Leven et al.(7Leven R.M. Mullikin W.H. Nachmias V.T. J. Cell Biol. 1983; 96: 1234-1240Crossref PubMed Scopus (16) Google Scholar) concluded that the ADP and thrombin-evoked spreading reaction in the rat megakaryocyte, which may be a functional response leading to platelet formation, depended upon an increased Na+ conductance. Further work, however, is needed to assess whether the IP3-activated Na+ influx is indeed involved in the cell spreading reaction because the present study did not assess the possible contribution of ADP-dependent stimulation of Na+/Ca2+ exchange or inhibition of Na+/K+ exchange to the observed [Na+]i increase. Platelets have little or no capacity to manufacture proteins, thus its progenitor, the megakaryocyte, must eventually express most, if not all, platelet ion channels. Therefore, the IP3-dependent cation current, in addition to playing a role in the megakaryocyte, may be important for platelet signaling. With a much larger surface area to volume ratio in the platelet, this current may result in large alterations of [Na+]i. In fact, in human platelets, thrombin induces a relatively greater production of IP3 than ADP (26Rink T.J. Sage S.O. Annu. Rev. Physiol. 1990; 52: 431-449Crossref PubMed Scopus (321) Google Scholar) and a greater increase in [Na+]i(41Borin M. Siffert W. J. Biol. Chem. 1990; 265: 19543-19550Abstract Full Text PDF PubMed Google Scholar, 42Sage S.O. Rink T.J. Mahaut-Smith M.P. J. Physiol. 1991; 441: 559-573Crossref PubMed Scopus (40) Google Scholar). The ADP-induced rise in [Na+]i has been shown to be mostly via ADP-activated receptor-operated channels (421); however, the mechanism of the thrombin-induced Na+ influx is not known and may involve the channel we report here in the rat megakaryocyte if this is also expressed in human megakaryocytes. In conclusion, we have demonstrated the existence of a plasma membrane conductance activated by 1,4,5-IP3 that carries Na+ into the cell at resting membrane potentials and may have a functional role in megakaryocyte signaling or be expressed for later use in platelet responses. We thank Andres Floto for helpful discussion, in particular on noise analysis methods, and thank Dr. Stewart Sage for comments on the manuscript.