INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The M conductance (g_M) is a noninactivating K⁺ conductance that is suppressed by receptors that couple to the Gq/G₁₁ family of G proteins (Caulfield et al. 1994; Haley et al. 1998; Marrion 1997). Muscarinic agonists suppress g_M in bullfrog (BFSG) and in mammalian sympathetic neurons (Adams et al. 1982a; Brown and Selyanko 1985) in hippocampal neurons (Halliwell 1990) and in neuroblastoma/glioma cell lines (Robbins et al. 1992; Schafer et al. 1991). This effect has been reconstituted in tsA-201 cells expressing M₁ muscarinic receptors and the KCNQ2/KCNQ3 K⁺ channels that are thought to underlie neuronal g_M (Shapiro et al. 2000).

The effect of bradykinin on g_M in mouse sympathetic ganglia is transduced wholly via G₁₁, whereas M₁ muscarinic effects are transduced partly by G_q, more substantially by G₁₁, and partly by pertussis toxin-sensitive G proteins (Haley et al. 2000b). These differences may account for the participation of different downstream effectors in the actions of each agonist. Thus bradykinin acts via phospholipase C-4 (PLC-4), generation of inositol trisphosphate (InsP₃), and the release of intracellular Ca²⁺ that directly inhibits g_M in mammalian sympathetic neurons (Cruzblanca et al. 1998; Haley et al. 2000a). By contrast, the participation of InsP₃ and intracellular Ca²⁺ in suppression of g_M by muscarinic agonists in both mammalian and frog ganglia has been excluded (Bofill-Cardona et al. 2000; Chen et al. 1993; Cruzblanca et al. 1998; del Rio et al. 1999; Pfaffinger et al. 1988; Selyanko et al. 1990) as has protein kinase C (PKC) (Bosma and Hille 1989; Marrion 1994; Selyanko et al. 1990) and phospholipase C (PLC) (del Rio et al. 1999; Haley et al. 2000b). The details of the muscarinic mechanism therefore remain to be elucidated (Cruzblanca et al. 1998).

ATP and nucleotides such as UTP represent a third class of agonists that suppress g_M by interaction with P_2Y receptors in both amphibian and mammalian sympathetic neurons (Bofill-Cardona et al. 2000; Bosma and Hille 1989; Brown et al. 2000; Groul et al. 1981; Jones 1985; Jones et al. 1984; Tokimasa et al. 1993). Although UTP appears to act in a similar fashion to bradykinin to suppress g_M via PLC, InsP₃, and mobilization of intracellular Ca²⁺ in rat superior cervical ganglion neurons (Bofill-Cardona et al. 2000), a staurosporin-sensitive protein kinase has been implicated in ATP-induced suppression of g_M in frog sensory neurons (Tokimasa and Akasu 1990). In the present study, we report a novel mechanism for P_2Y receptor-induced g_M suppression in BFSG neurons. Although PLC is involved, downstream signaling involves neither InsP₃, Ca²⁺, nor PKC. The transduction mechanism used by ATP in BFSG neurons is therefore distinct from the PLC-independent mechanism used by muscarinic agonists, from the Ca²⁺-dependent mechanism used by UTP and bradykinin in mammalian neurons (Bofill-Cardona et al. 2000; Cruzblanca et al. 1998), and from the kinase-dependent mechanism that may operate in frog sensory neurons (Tokimasa and Akasu 1990). Because protein kinase C, InsP₃, and Ca²⁺ were excluded, we tested the involvement of the neuronal Ras-guanyl nucleotide releasing protein, RasGRP (Ebinu et al. 1998). Diacylglycerol (DAG), formed as a result of the action of PLC, is normally assumed to exert its downstream effects via PKC. DAG also, however, binds to RasGRP. This in turn, activates the monomeric G protein, Ras. Because activated Ras can exert rapid, transcription-independent modulation of ion channel function via activation of mitogen-activated protein kinases (MAPK) (Fitzgerald 2000; Fitzgerald and Dolphin 1997), RasGRP is an attractive candidate for implication in the ATP-induced modulation of g_M. We found, however, that ATP-induced g_M suppression in BFSG neurons persisted under conditions where Ras function was impaired. Although it remains to be determined how ATP stimulation of PLC leads to g_M suppression in BFSG neurons, one possibility is the involvement of "upstream" rather than "downstream" signaling. In other words, g_M suppression may result from PLC-induced depletion of phosphatidylinositol 4,5,biphosphate (PIP₂). This type of mechanism has been implicated in agonist control of inwardly rectifying K⁺ channels (Huang et al. 1998; Kobrinsky et al. 2000; Lei et al. 2001a; Xie et al. 1999) but its possible role in regulation of g_M is yet be tested. A preliminary report of this work has appeared (Smith et al. 2000).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals were cared for in accordance with the principles and guidelines of the Canadian Council on Animal Care and experimental protocols were approved by the Health Sciences Animal Welfare Committee of the University of Alberta.

Cell isolation

Neurons in the VIth to Xth paravertebral sympathetic ganglia of male or female bullfrogs (Rana catesbeiana; body length, 10-12 cm) were dissociated using trypsin and collagenase as described previously (Selyanko et al. 1990). Experiments were done on freshly dissociated neurons or on neurons that were maintained for 1-2 days in tissue culture (see following text).

Tissue culture

The effects of Ras isoprenylation inhibitors were examined over periods of days. This was done in cultured cells. Freshly dissociated neurons were maintained for up to 2 wk in serum-free, low-density, defined-medium, neuron-enriched culture as previously described (Lei et al. 1997). Neuron-enriched cultures were prepared by preplating the total yield of ganglion cells from each frog into two or three 35-mm culture dishes. After 1-2 h, most of the nonneuronal cells adhered to the bottom of the dishes and the nonadherent cells, which were primarily neurons, were harvested, redistributed to 30 culture dishes (35 mm) and cultured in fresh medium (3 ml/dish). The culture medium consisted of diluted L-15 medium (73%) supplemented with 10 mM glucose, 1 mM CaCl₂, 100 units/ml penicillin, 100 µg/ml streptomycin, and 10 µM cytosine arabinoside.

Electrophysiology

All experiments were carried out at room temperature (20°C). Standard whole cell recording methods using an Axopatch 1B amplifier were used to record M currents from acutely isolated or cultured BFSG neurons. Details of methods are already published (Selyanko et al. 1990). In experiments where the perforated technique was used, 4.5 mg of nystatin (Sigma) and 30 mg of pluronic F127 were first dissolved in 300 µL DMSO. The mixture was made 10 ml in "internal" solution. The first 1 mm of the tip of each pipette was filled with regular internal solution and the nystatin/pluronic solution admitted to the back of the pipette. Within 25 min of forming a giga seal, access resistance through the perforated patch fell to 20.2 ± 2.0 (SE) M (n = 24). Although this is somewhat higher than that seen with conventional whole cell recording (6.0 ± 0.3M; n = 24), we cannot rule out the possibility that complete rupture of the patch occurred in some of the cells studied with the perforated-patch technique (Chung and Schlichter 1993).

Resting membrane potential was 50 to 55 mV, and cells were normally held at 30 mV. An estimate of cell size was obtained from the input capacitance (C_in). Experiments were carried out on B cells that were identified by their "large" size (C_in >30pF) and by their response to muscarine, which reduced steady-state outward current at 30 mV (Kurenny et al. 1994). Because the currents to be recorded were usually <0.5 nA, no corrections were made for the voltage drop across the series resistance, which varied between 4 and 20 M. Although this means that the maximum possible voltage error due to series resistance could be as much as 10 mV, errors of this magnitude would only have been encountered in the few cells that exhibited exceptionally large currents while being studied with relatively high-resistance pipettes. Current-voltage relationships were examined using a 4.5-s ramp command from the holding potential of 30 to 110 mV (~18 mV/s). For consistency with conventional current-voltage plots, the direction of the voltage axis has been reversed for presentation. Currents are displayed for the 90- to 30-mV range only. The high conductance part of the plot (i.e., more positive than 75 mV) is dominated by current through M channels plus a small leak current (Selyanko et al. 1990). Leak current (I_L) at 30 mV was estimated by extrapolation of the I-V plot obtained at voltages between 75 and 90 mV. The percentage of agonist-induced g_M suppression was calculated from the following formula.

For studying I_M, extracellular solution contained (in mM) 113 NaCl, 6 KCl, 2 MgCl₂, 2 CaCl₂, 5 HEPES/NaOH (pH 7.2), and 10 D-glucose. Patch pipettes had DC resistances of 3-10 M. Pipette solution contained (in mM), 110 KCl, 10 NaCl, 2 MgCl₂, 0.4 CaCl₂, 4.4 EGTA, 5 HEPES/KOH(pH 6.7), 10 D-glucose, and 2 Na₂ATP. (pCa = 7) (Selyanko et al. 1990). Whole cell recordings of Ca²⁺ channel currents (I_Ca) were obtained by means of discontinuous single-electrode voltage-clamp (Axoclamp 2A) using previously published methods (Lei et al. 1997). Ba²⁺ was used as the charge carrier (I_Ba) and external solution contained (in mM) 117.5 N-methyl-D-glucamine (NMG) chloride, 2.5 NMG-HEPES, and 2.0 BaCl₂ (pH7.2). The internal solution in the patch pipette consisted of (in mM) 76.5 NMG-Cl, 2.5 HEPES, 10 BAPTA, 5 Tris-ATP, and 4 MgCl₂ (pH7.2).

Intracellular Ca²⁺ measurements

Cells were loaded with fura-2 (Molecular Probes) via the AM method. For each experimental day, fura-2 AM was first dissolved in fresh DMSO (supplied in sealed ampoules, Sigma) as a 10 mM stock solution, and diluted to 5 µM with the extracellular solution immediately before 15 min of incubation in individual dishes of cells at room temperature. The cells were then superfused with extracellular recording solutions for 10-15 min to remove untrapped fura-2 before the beginning of each experiment. Groups of cells (typically 5) were imaged via a [mult]40, 1.3 numerical aperture, oil-immersion objective with an imaging system (TILL Photonics, Eugene, OR), which included a monochrometer capable of switching the excitation wavelength in <1 ms and a cooled CCD camera synchronized to integrate the emitted light at each excitation wavelength. Ratiometric imaging of fura-2 fluorescence was performed at 0.2 Hz with sequential 340/380-nm excitation. Each fluorescence image was collected at 640 × 480 pixels resolution via a long-pass 510 nM dichroic mirror/emission filter and integrated for 10 ms. After subtraction of background fluorescence at each wavelength of excitation, the fluorescence ratio (R) of fura-2 at 340-nm excitation/380-nm excitation, was displayed as a continuous record showing the time course of changes of R from an individual region of interest.

Conversion of R into [Ca²⁺]_i was performed in separate calibration experiments, in which individual cells were dialysed in whole cell recording with intracellular solutions of known [Ca²⁺] and fura-2 (0.1 mM, potassium salt from Molecular Probes). The three solutions for this calibration are identical to those previously used for calibrating indo-1 (Tse and Tse 2000) and have [Ca²⁺] of <0.1 nM, 212 nM, and 15 µM, respectively. R values were converted to [Ca²⁺]_i using the following equation (Grynkiewicz et al. 1985)
In this study, the values of R_min was 0.132, R_max was 3.4, and K was 2.7 µM.

Drugs and chemicals

For electrophysiological studies, drugs were applied using a rapid superfusion system that was constructed from a set of five 0.8-mm-diam polyimide tubes (Cole Parmer, Vernon Hills, IL). These were connected via a tap system to a series of reservoirs, and their tips fed into a small-volume mixing chamber. The mixing chamber fed into another 0.8-mm polyimide tube that was placed within 100 µm of the neuron under study. Control voltage ramps were applied to measure I_M while extracellular solution was applied from the superfusion system, the flow was then switched to allow superfusion of drugs, and a second voltage command applied. All drugs except for xestospongin C were applied by bath perfusion for the Ca²⁺ imaging or for the electrophysiological experiments. Xestospongin C was applied following interruption of superfusion and introduction of the drug from a pipette to attain a concentration of 1.8 µM in a stationary bath. After 5- or 10-min exposure to xestospongin C, superfusion was resumed to retest ATP responses

Special care was taken to prepare solutions of U73122 and U73343 according to the manufacturer's instructions. Aliquots of these substances were first dissolved in chloroform which was then allowed to evaporate under a stream of nitrogen to yield a residue of U73122 or U73343. This residue was stored at 20°C until the day of the experiment when it was dissolved to make a stock solution in fresh DMSO. Serial dilutions were arranged so that the final concentration of DMSO in solutions applied to cells was <0.1%. Acute application of 0.1% DMSO or culturing cells for 7 days in a medium containing 0.1% DMSO had no noticeable effect on their electrophysiological properties. Due to their lipophilic nature, -hydroxyfarnesyl-phosphonic acid (-HFA), perillic acid, xestospongin C and PD 98059 were also initially dissolved in fresh DMSO and applied to cells in solutions that contained <0.1% of this solvent. In the case of xestospongin C, an appropriate concentration of DMSO in extracellular medium was introduced from a pipette to attain a 0.1% DMSO solution in a stationary bath. Cells were exposed to this solution for 10 min prior to resuming superfusion and retesting the effect of ATP. Drugs and chemicals were purchased from Sigma (Oakville, ON, Canada) except for U73122, U73343, chelerythrine, perillic acid, -HFA, okadaic acid, and PD 98059 which were from Biomol (Plymouth Meeting, PA), xestospongin that was from Calbiochem (San Diego, CA) and pluronic F127 (pluronic acid), which was from BASF Wyandotte, Canada. Data are expressed as means ± SE and significance of difference estimated using Student's two-tailed unpaired t-test, Data were considered significantly different when P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

U-73122 inhibits the effects of ATP on g_M

A typical response to ATP and its inhibition by the phospholipase inhibitor, U73122 is illustrated in Fig. 1A. The downward deflections on the record reflect current responses to voltage ramps to 110 mV evoked before, during, and after ATP application. The current-voltage plots derived from some of these ramps are shown Fig. 1, B-D. ATP suppresses current only in the activation range for g_M (i.e., more positive than 75 mV). Because no ATP-induced inward current is seen at potentials more negative than 75 mV, it is unlikely that the cells express P_2X receptors. Because the reversal potential for P_2X -induced currents is typically 0 mV (Khakh et al. 1995), a robust inward current would be predicted at negative potentials if the receptors were present. Figure 1, A and B, also shows that steady-state g_M is increased following ATP removal. This represents a well-documented over-recovery phenomenon (Chen et al. 1993; Marrion et al. 1991; Pfaffinger 1988).

View larger version (24K): [in this window] [in a new window]   Fig. 1. Antagonism of the effects of ATP on M current by U-73122. A: chart recording of steady-state outward current at 30 mV. Downward deflections are 4.5-s ramp commands to 110 mV to measure M-channel conductance (gM; voltage command trace omitted for clarity). ATP (250 µM) decreased membrane conductance and suppressed steady-state outward current at 30 mV. Application on 10 µM U73122 produced an outward current that abated within ~2 min, and application of ATP immediately after U73122 failed to exert an effect. B-C: I-V plots derived from the 4.5-s ramp commands as labeled in A and as explained in the methods section. B: superimposed plots show that ATP suppressed current at voltages positive to 75 mV (i.e., in the gM activation range), washout of ATP was associated with gM over-recovery. C: superimposed plots to show direct potentiation of outward current in the gM activation range by U73122. D: small effect of ATP after U73122, apparent slight depression of the current reflects the fact that the control response was acquired during U73122-induced gM potentiation. E: graph to show effectiveness of U73122 in antagonizing ATP responses, (n = 5 for all points, line fit by eye to a sigmoidal curve), 50% antagonism was achieved with 0.14 µM U73122. F: log-concentration effect curve for U73122-induced potentiation of gM (n > 5 for all points; line fit by eye to a sigmoidal curve). G and H: effects of 250 µM ATP on [Ca2+]i before and after superfusion of U73122 for 7 min.

Application of 10 µM U73122 also caused a direct and transient increase in g_M (Fig. 1, A and C), but subsequent application of ATP produced no change in steady-state outward current at 30 mV (Fig. 1A). Modest suppression of g_M appears in the I-V plots shown in Fig. 1D, but this probably reflects the fact that the control ramp was taken during transient potentiation of g_M by U73122. Thus control applications of 250 µM ATP suppressed g_M by 93.9 ± 2.0% (n = 5), whereas in the presence of 10 µM U73122, ATP produced only 21.6 ± 3.6% suppression (n = 5, P < 0.001; Table 1). The antagonism of ATP responses and the direct, transient enhancement of g_M produced by U73122 were concentration dependent. Figure 1E shows that the half-maximal effect for attenuation of ATP was seen with 0.14 µM U73122. Because the maximal concentration for potentiation of g_M by U73122 was not determined, the concentration for a half-maximal effect could not be estimated (Fig. 1F). Unlike U73122, the inactive isomer, U73343 (10 µM) failed to attenuate ATP-induced g_M suppression (P > 0.9; Table 1). In four experiments, we demonstrated the lack of effect of U73343 and antagonism of ATP responses by U73122 in the same cell. U73343 (10 µM) also failed to increase steady-state g_M in a similar fashion to the active U73122 in any of four neurons tested. It remains to be determined whether the direct effect of U73122 on g_M reflects its ability to inhibit PLC.

The aforementioned results were obtained with whole cell recording. Essentially the same results were seen in five cells studied with perforated patches; 10 µM U73122 blocked ATP responses and itself produced a transient increase in g_M.

U 73122 inhibits PLC-dependent release of intracellular Ca²⁺

Because U73122 antagonized ATP-induced g_M suppression, it was necessary to demonstrate its effectiveness in inhibiting PLC when applied extracellularly to BFSG B cells. This was done by testing whether U73122 would antagonize ATP-induced elevation of intracellular Ca²⁺, a response that requires PLC activation. A typical experiment is shown in Fig. 1, G and H.

Figure 1G shows the increase in intracellular Ca²⁺ concentration ([Ca²⁺]_i) produced by superfusion of 250 µM ATP. Figure 1H shows that after 7 min in the presence 10 µM U73122, the ATP-induced elevation of Ca²⁺ is blocked. Similar results were obtained in three experiments. The lack of effect of ATP in the presence of U73122 cannot be attributed to lack of reproducibility as up to three successive Ca²⁺ responses to ATP could be evoked in cells studied in the presence of the inactive isomer U73343 (n = 3).

It could be argued that the presence of fura-2-AM in the Ca²⁺ measurement experiments increased the Ca²⁺-buffering capacity so that data from these experiments is not directly comparable with those in which ATP-induced g_M suppression was studied. We found however that ATP also suppressed the current by 74.6 ± 4.6% in 5/5 neurons that had been exposed to 5 µM fura-2-AM for 30 min.

Chelerythrine fails to affect ATP-induced g_M suppression

Prior to testing the actions of the PKC inhibitor, chelerythrine (5 µM extracellular + 10 µM intracellular) on ATP-induced g_M suppression, positive control experiments were done to demonstrate its effectiveness in altering a protein kinase-dependent effect in BFSG neurons. This was done by examining the ability of chelerythrine to antagonize okadaic acid-induced enhancement of Ca²⁺ channel inactivation (Werz et al. 1993). Inclusion of this phosphatase inhibitor in the recording pipette increases the inactivation of Ca²⁺ channel conductance (g_Ca) in BFSG neurons. This effect is attenuated by the kinase inhibitor, staurosporin (Werz et al. 1993). The effects of the PKC inhibitor, chelerythrine are shown in Fig. 2A. Ca²⁺ channel currents (Fig. 2B) were evoked once every 60 s by 300-ms commands to 10 mV from a holding potential of 80 mV using Ba²⁺ as the charge carrier (Jassar et al. 1993). Inactivation was defined as the ratio of the peak to end-of-pulse current (Fig. 2B). Inactivation initially appeared large yet rapidly decreased during the first few minutes of establishing whole cell recording conditions. This was probably due to dialysis of the cells with the internal solution and blockade of outward K⁺ currents. In control cells, inactivation again very slowly increased from ~1.4 to 1.8 over the course of a 70-min recording session. More inactivation was seen when 5 µM okadaic acid was included in the recording pipette. Inactivation increased from ~1.3 to 2.3 during 70-min recording. This increase in inactivation was prevented when the effects of okadaic acid were examined with pipettes that also contained 10 µM chelerythrine. Chelerythrine (5 µM) was also applied in the extracellular solution. Interestingly, chelerythrine alone attenuated the development of g_Ba inactivation. This suggests that the development of increased inactivation over the 70-min time course of the experiment may be phosphorylation dependent. The importance of these results to the present study is that they show that inclusion of 10 µM chelerythrine within the recording pipette plus extracellular application of 5 µM chelerythrine, effectively reduces kinase activity, presumably PKC activity, in BFSG neurons. This effect is clear even after 20-min intracellular dialysis with chelerythrine (Fig. 2A). We therefore used this protocol to examine the role of PKC in ATP-induced g_M suppression.

View larger version (39K): [in this window] [in a new window]   Fig. 2. Positive control experiments for chelerythrine and its lack of effect on ATP-induced gM suppression. A: effects of chelerythrine (10 µM intracellular + 5µM extracellular) on okadaic acid-induced enhancement of IBa inactivation. IBa was evoked once per minute using a 300-ms depolarizing voltage command to 10 mV from a holding potential of 80 mV. Inactivation, defined as the ratio of the peak to end of pulse current, increased over the 70-min time course of the experiment. Inclusion of 5 µM okadaic acid in the pipette facilitated the development of inactivation, but this effect was prevented in cells where chelerythrine was included in the extracellular medium (5 µM) and within the pipette (10 µM) with 5 µM okadaic acid. Treatment with chelerythrine alone, in the absence of okadaic acid, also attenuated the accumulation of inactivation. B: sample record of IBa illustrating points of measurement for peak and end-of-pulse current. C: lack of effect of chelerythrine on ATP-induced gM suppression. Top: chart recording of steady-state outward current at 30 mV, interrupted by responses to hyperpolarizing ramp commands to measure gM and leak conductance. Pipette contained 10 µM chelerythrine, and 5 µM chelerythrine was applied extracellularly as indicated by the bar. Note similar effects of ATP (250 µM) on gM 7, 14, and 20 min after initiating recording and progressive increase in leak conductance throughout the experiment. Middle: the estimated concentration of intracellular chelerythrine [C]i. It is a graph of the line [C]i = [C]p (1  et/) where [C]p is the concentration of chelerythrine in the pipette (10 µM) and  is the time constant for equilibration of the chelerythrine in the pipette with the intracellular fluid. In this case,  = 20.2 min. This was calculated for the 4.7 M access resistance (RA) using the empirical equation o = 0.6 RA M W.1/3, to calculate the time constant for diffusion into a chromaffin cell where Cin = 6pF and M. W for chelerythrine =384. The value of  for chelerythrine equilibration with the illustrated BFSG B-neuron (Cin = 90pF) is calculated from the equation /o = Cin/6 (Pusch and Neher 1988). Bottom: voltage-ramp commands. D-F: I-V plots derived as in methods from the voltage-ramp commands indicated during 3 ATP responses. Note that ATP induces the same percentage suppression of outward current at 30 mV (i.e., it produces the same amount of gM suppression), but the leak conductance, measured between 90 and 75 mV is substantially increased.

Chelerythrine failed to affect ATP-induced g_M suppression in any of five cells tested (P > 0.1). A typical experiment is shown in Fig. 2, C-F, and numerical data are summarized in Table 1. The top trace in Fig. 2C shows suppression of steady-state outward current at 30 mV (I_M) by three successive applications of 250 µM ATP recorded 7, 14, and 20 min after starting intracellular dialysis with 10 µM chelerythrine. Although 5 µM chelerythrine was also applied extracellularly, the ATP-induced suppression of outward current was unaffected. The middle trace in Fig. 2C is an estimate of the theoretical time course of change of intracellular chelerythrine concentration that occurred during this experiment. The line is derived from the empirical equations of Pusch and Neher (1988; see figure legend for details). By the time of the third application of ATP, the estimated intracellular chelerythrine concentration approached 6 µM. The extracellularly applied chelerythrine (5 µM) would also have access to the intracellular fluid. The estimated intracellular concentration of chelerythrine was almost 10 times the IC₅₀ for inhibition of rat brain PKC (0.66 µM) (Herbert et al. 1990). The bottom trace in Fig. 2C shows the voltage-ramp commands to 110 mV used to obtain I-V relationships shown in shown in Fig. 2, D-F. ATP suppresses outward current at 30 mV by 80, 87, and 80% in the three records. The top record in Fig. 2C, as well as those in D-F, also show that chelerythrine-increased leak conductance measured between 75 and 90 mV. This effect was seen in all cells tested. Leak conductance thus increased from 5.0 ± 1.6 to 12.0 ± 2.2 nS (n = 4) after ~15 min with chelerythrine in the pipette.

Thapsigargin blocks ATP-induced Ca²⁺ response but not ATP-induced g_M suppression

Thapsigargin is a Ca²⁺-ATPase inhibitor that impairs the refilling of Ca²⁺ stores following their depletion by Ca²⁺ -mobilizing agonists (Thastrup et al. 1990), it impedes the release of intracellular Ca²⁺ produced by muscarinic agonists in rat sympathetic neurons but does not prevent g_M suppression (del Rio et al. 1999). In fura-2 experiments, we found that application of thapsigargin in the presence of our usual concentration of extracellular Ca²⁺ (2 mM) produced a transient elevation of intracellular Ca²⁺ but did not block ATP-induced Ca²⁺ release (data not shown). The inability of thapsigargin to block the ATP-induced Ca²⁺ signal may reflect refilling of Ca²⁺ stores from the extracellular medium. It was therefore necessary to examine the effects of thapsigargin using an extracellular solution containing 1 mM EGTA, 0 Ca²⁺, and 10 mM MgCl₂. Under these conditions, thapsigargin (5 µM) prevented the Ca²⁺ response to ATP. The time course of changes of intracellular Ca²⁺ concentration ([Ca²⁺]_i) in two different cells are illustrated in Fig. 3A. There was considerable cell-to-cell variability in the magnitude of the changes in [Ca²⁺]_i induced by ATP and thapsigagin. In one cell, initial application of ATP increases ambient [Ca²⁺]_i by 33% (from 61 to 81 nM), whereas in the other, ATP increases [Ca²⁺]_i by only 6% (from 168 to 178 nM). Application of thapsigargin is associated with a large increase in [Ca²⁺]_i signal in the first cell but with only a modest change in the second. Application of ATP after thapsigargin treatment failed to produce a measurable increase in [Ca²⁺]_i in either cell.

View larger version (28K): [in this window] [in a new window]   Fig. 3. Effects of ATP and thapsigargin on gM and intracellular Ca2+ in bullfrog sympathetic ganglion (BFSG) B neurons. A: time course of changes in intracellular Ca2+ ([Ca2+]i) induced by ATP (250 µM) and thapsigargin (5 µM). Recordings from 2 different cells obtained in extracellular medium containing 0Ca2+, 10 mM Mg2+, and 1 mM EGTA. Note transient increases in Ca2+ produced by ATP and thapsigargin, variation in the size of the response in the 2 different cells, and lack of effect of 2nd application of ATP on [Ca2+]i. B, top: chart recording of changes in steady-state outward current at 30 mV induced by ATP and thapsigargin in another cell, displayed on the same time scale as the [Ca2+]i measurements in A. Note decrease in steady-state current (gM) in presence of 0Ca2+, 10 mM Mg2+, and 1 mM EGTA. Application of ATP prior to thapsigargin induces robust gM suppression, but steady-state outward current declines when thapsigargin is introduced. Second application of ATP (after thapsigargin) produces suppression of the outward current that remains. Bottom: 4.5-s voltage ramp commands to 110 mV. C: I-V plots obtained as in methods from voltage ramps as indicated in B note 92% suppression of gM by 1st (control) ATP application. D: I-V plot for effect of thapsigargin; steady-state gM at 30 mV is decreased and leak conductance is increased. E: I-V plot for ATP response in the presence of thapsigargin as indicated from B. Although steady-state gM is reduced, ATP suppresses available conductance by 89%.

We used the same experimental protocol to examine the effect of thapsigargin/0Ca²⁺/1 mM EGTA/10 mM Mg²⁺ on ATP-induced g_M suppression. This solution produced a large and poorly reversible increase in leak conductance from 6.7 ± 1.9 to 18.6 ± 6.9 nS (n = 4). It also reduced steady-state g_M at 30 mV to 54.8 ± 12.4% of control (n = 4). Despite this, the percentage suppression of the available g_M induced by ATP was unchanged when thapsigargin was included in a 0Ca²⁺/1 mM EGTA/10 mM Mg²⁺ solution (P > 0.1, Table 1). A typical experiment is illustrated in Fig. 3B. The top trace is a chart recording of steady-state current set to the same time scale as the Ca²⁺ measurements in Fig. 3A. Voltage commands are shown in the bottom trace. Figure 3C illustrates I-V plots derived from the voltage ramps in 0Ca²⁺/1 mM EGTA/10 mM Mg²⁺ to show that the first (control) application of ATP induced 92% g_M suppression. Figure 3D shows that thapsigargin increased leak conductance, measured between 90 and 75 mV from 4.1 to 59.5 nS and decreased steady-state g_M to 27% of its initial value. Figure 3E illustrates that ATP still invokes 89% suppression of the available g_M when 5 µM thapsigargin is added to the 0Ca²⁺/1 mM EGTA/10 mM Mg²⁺ extracellular solution.

Lack of effect of InsP₃ receptor antagonists on ATP-induced gM inhibition

Using the equations of Pusch and Neher (1988), we estimated that inclusion of 500 µM heparin in our patch pipettes would achieve an intracellular concentration of 300 µM after 20-25 min. This is 10⁵ times the IC₅₀ for its antagonistic effect at the InsP₃ receptor (Ghosh et al. 1988) and inclusion of a lower concentration (200 µM) heparin for 14 min antagonizes bradykinin-induced g_M suppression in mammalian neurons (Cruzblanca et al. 1998). Despite this, inclusion of 500 µM heparin failed to antagonize ATP-induced g_M suppression in any of 4 BFSG cells tested after 20-25 min of recording (data not shown).

We confirmed the effectiveness of the membrane-permeant InsP₃ antagonist, xestospongin C by demonstrating its ability to prevent ATP-induced elevation of [Ca²⁺]. This response was completely blocked in six of seven cells exposed to xestospongin C (1.8 µM) for 5-10 min. In the remaining cell, the ATP-induced elevation of [Ca²⁺]_i was reduced by ~50%. As mentioned in METHODS, the flow of extracellular solution through the bath was interrupted and a small volume of xestospongin C was added to a stationary bath to attain a final concentration of 1.8 µM. In control experiments where vehicle was pipetted into the bath, ATP-induced g_M suppression was slightly attenuated. Thus before vehicle, ATP suppressed g_M by 62.6 ± 4.0% (n = 7), whereas after interruption of superfusion and introduction of vehicle (containing DMSO), ATP suppressed g_M by 50.4 ± 3.5% (n = 7; P < 0.05, Table 1). Prior to testing with xestospongin, ATP suppressed g_M by 76.1 ± 5.3% (n = 7) and by 61.2 ± 2.9 (n = 7) after 10 min of drug exposure (P < 0.05, Table 1). Although there is an apparent reduction in the effectiveness of ATP under these circumstances, there is no difference (P > 0.05) between the percentage g_M suppression produced following exposure of cells to xestospongin compared with vehicle (Table 1). It was therefore concluded that 1.8 µM xestospongin C had no effect on ATP-induced g_M suppression. These experiments were carried out using perforated-patch recording.

Figure 4A illustrates a typical experiment in which ATP produced the same amount of g_M suppression before and after a 10-min exposure to 1.8 µM xestospongin C. However, in another cell (Fig. 4B), a shorter (5 min) exposure to xestospongin abrogates the Ca²⁺ response.

View larger version (20K): [in this window] [in a new window]   Fig. 4. Lack of effect of xestospongin C on ATP-induced gM suppression and blockade of ATP-induced Ca2+ response. A: chart recordings of steady-state outward current recorded at 30 mV (IM) and its suppression by 250 µM ATP. Responses recorded before and after interruption of flow and exposure of cell to 1.8 µM xestospongin C for 10 min. Approximately the same effect is seen before and after exposure to drug. Deflections are current responses to voltage ramps to 110 mV used to generate I-V plots; these and voltage command traces have been omitted for clarity. B1: increase in [Ca2+]i induced in another cell by 250 µM ATP. B2: blockade of ATP-induced Ca2+ response following interruption of flow and exposure to 1.8 µM xestospongin C for 5 min.

Inhibition of Ras function fails to inhibit ATP-induced g_M suppression

Because none of the obvious downstream effectors of PLC activation seemed to be involved in transducing the effects of ATP, we investigated the possible involvement of Ras, which can be activated following DAG interaction with the Ras-guanyl nucleotide releasing protein, RasGRP (Ebinu et al. 1998). Because active Ras requires isoprenylation to function, isoprenylation inhibitors such as -HFA or perillic acid may be used to perturb Ras function in experimental situations. We have previously shown that these substances attenuate nerve-growth factor-induced enhancement of Ca²⁺ channel current in BFSG B cells (Lei et al. 1998). Because -HFA or perillic acid would not be expected to acutely inhibit Ras function, their effects were assessed following long-term exposure in tissue culture (Lei et al. 1997). Cells were cultured for 3 days in the presence or absence of perillic acid or  -HFA, and their response to ATP were examined. Both inhibitors were ineffective. Overall, ATP suppressed g_M by 77.4 ± 7.6% in five control cells cultured with 0.1% DMSO and by 80.4 ± 4.3% in five cells cultured with -HFA (10 µM, P > 0.7; Table 1). Similarly 3 days culture in the presence of perillic acid (0.1 mM) had no effect on ATP-induced g_M suppression (P > 0.7; Table 1).

There are several downstream effectors of low-molecular-weight G proteins such as Ras (p21 ras). The best-characterized pathway involves a cascade of serine-threonine-tyrosine kinases that eventually lead to the activation of mitogen-activated protein kinases (MAP kinases) (Grewal et al. 1999; Jaiswal et al. 1996). P21 ras activates various MAP kinase kinase kinases (such as raf) and these phosphorylate and activate MAP kinase kinases, including MEK1/2. MEK-like kinases in turn, phosphorylate and activate MAP kinases, including the extracellular receptor-activated kinases (Erk1 and Erk2). This pathway can be perturbed with the MEK inhibitor, PD 98059. We have previously shown that 10 µM PD 98059 can attenuate nerve growth factor-induced enhancement of I_Ca in BFSG neurons (Lei et al. 1998). However, 10-min application of PD 98059 (10 µM) was ineffective in attenuating ATP-induced g_M suppression (P > 0.95; Table 1).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The main finding of this paper is that ATP-induced suppression of g_M in BFSG B cells proceeds via PLC but not by any of the usual downstream messengers; InsP₃, Ca²⁺, or chelerythrine-sensitive protein kinases.

Role of PLC in ATP-induced g_M suppression

The data implicating PLC in ATP-induced g_M are clear. The ability of U73122 to antagonize ATP responses is not shared by the inactive isomer, U73343. Moreover, extracellularly applied U73122 effectively inhibits PLC in BFSG as it attenuates mobilization of intracellular Ca²⁺ by ATP (Fig. 1). P_2Y receptors usually exert their effects via PLC (Burnstock 2001), so this isoform is likely involved in the effects of ATP on g_M.

Exclusion of InsP₃ and Ca²⁺ as downstream effectors

The lack of effect of xestospongin C, heparin or the combination of 5 µM thapsigargin/0Ca²⁺/10 mM Mg²⁺/1 mM EGTA on the percentage g_M suppression induced by ATP argue against a role for InsP₃-induced Ca²⁺ release in ATP-induced g_M suppression. Both xestospongin C and thapsigargin/0Ca²⁺/10 mM Mg²⁺/1 mM EGTA appeared to block ATP-induced Ca²⁺ release (Figs. 3A and 4B). It may be argued, however, that xestospongin C and thapsigargin failed to completely abrogate the Ca²⁺ response but simply reduced the amplitude of the response to a level undetectable by our fura-2 methodology. Moreover a small, possibly localized change in [Ca²⁺]_i may have been sufficient to promote M-channel closure. Although we cannot completely exclude this possibility, it has been reported that even in the presence of normal extracellular [Ca²⁺], 5 µM thapsigargin blocks bradykinin-induced g_M suppression in rat sympathetic neurons. This observation was used to implicate Ca²⁺ mobilization in the action of bradykinin (Cruzblanca et al. 1998). Because 5 µM thapsigargin fails to attenuate ATP-induced g_M suppression in BFSG, this argues against a role for Ca²⁺ in the generation of this effect. The poorly reversible attenuation of steady-state g_M by thapsigargin may reflect the suppression of the conductance by high levels of intracellular Ca²⁺ (Marrion et al. 1991).

Additional support for the hypothesis that ATP-induced g_M suppression in BFSG neurons is independent of ATP-induced elevation of [Ca²⁺]_i comes from a study of changes in activation and deactivation kinetics associated with ATP-induced g_M suppression. Although Ca²⁺-induced g_M suppression is associated with increases in activation/deactivation rate (Yu et al. 1994), no such changes were associated with ATP (or muscarine)-induced g_M suppression (Chen et al. 2001).

Our findings differ from those seen with another P_2Y agonist, UTP in mammalian sympathetic neurons (Bofill-Cardona et al. 2000). In that system, g_M suppression appears to be mediated by PLC and by InsP₃-induced increases in [Ca²⁺]_i. There is no obvious explanation, other than species difference, to account for the disparity between our findings and those of Bofill-Cardona et al.

Is DAG involved in ATP-induced g_M suppression?

Because the effect of ATP in BFSG neurons proceeds via PLC and not via Ca²⁺ or InsP₃, the production of DAG may be necessary for g_M suppression. However, DAG produced as a result of agonist-induced activation of PLC cannot act in a conventional fashion to suppress g_M via PKC activation. This is because all PKC inhibitors hitherto tested fail to prevent agonist-induced g_M suppression in amphibian or mammalian sympathetic ganglia, regardless of the agonist used (Bosma and Hille 1989; Marrion 1994, 1997; Selyanko et al. 1990). The lack of effect of chelerythrine on ATP-induced g_M suppression in the present experiments is therefore not altogether unexpected. This result does, however, differ from that of Tokimasa and Akasu (1990), who implicated a kinase-dependent mechanism in ATP effects on putative g_M in frog sensory neurons.

Because diacylglycerol analogs suppress g_M in frog and mammalian sympathetic neurons (Bosma and Hille 1989; Selyanko et al. 1990), we examined the role of a novel DAG target, the Ras GTP exchange factor Ras-GRP (Ebinu et al. 1998) in transducing the effect of DAG on M channels. This is a viable target for inclusion in the g_M transduction process as it is selectively and strongly expressed in neurons. It was not feasible to examine the role of Ras-GRP directly, but the role of Ras could be readily examined in tissue culture experiments. Under similar culture conditions, we have shown that both -HFA and perillic acid effectively prevent nerve growth factor-induced upregulation of Ca²⁺ channel current in BFSG neurons (Lei et al. 1998). Because these Ras isoprenylation inhibitors failed to antagonize ATP-induced g_M suppression, Ras is unlikely to be involved in the transduction mechanism. The lack of effect of the MEK inhibitor, PD98059, on ATP-induced g_M was consistent with this observation. We have also previously reported that long-term exposure of cultured BFSG neurons to nerve growth factor, which is a potent activator of the Ras-MAPK pathway (Kaplan and Stephens 1994), fails to attenuate g_M (Lei et al. 2001b).

It has been suggested that DAG can interact directly with M channels (Chen et al. 1994; Clapp et al. 1992). It may therefore directly transduce the effects of agonists such as ATP. Although we have found that DAG-induced g_M suppression is insensitive to the PKC inhibitors, H-7 and staurosporin (Chen et al. 1994; Selyanko et al. 1990), different results have been reported by other laboratories (Bosma and Hille 1989). We therefore examined the effects of chelerythrine on 1,2-dioctanyl-sn-glycerol (DOG)-induced g_M suppression using the same treatment protocol as was for established PKC inhibition in the present study (Fig. 2) (P. L. Stemkowski, unpublished observations). Unfortunately, the results were difficult to analyze. Changes in g_M produced by DOG were small (<20% suppression by 40 µM DOG), so that measures of response amplitude were unreliable, especially in the presence of the large increase in leak conductance that was produced by chelerythrine (Fig. 2C-F). We could not say with any certainty whether chelerythrine antagonized DOG responses and could neither refute nor confirm the hypothesis that ATP effects involve a direct action of DAG on M channels. Although there is evidence that phorbol-ester induced g_M suppression is sensitive to PKC inhibitors (Marrion 1994), these observations may not be relevant to the present discussion as it is now recognized that the actions of diacylglycerols and phorbol-esters are not identical (Schreur and Liu 1996; Thomas et al. 1991).

Because a role for downstream signaling from PLC cannot be established, it is possible that ATP-induced g_M suppression may be a consequence of "upstream" signaling. It may result from the depletion of PIP₂ rather than from actions of products of PLC. This would imply that M channels would tend to be open in the presence of PIP₂ and would tend to close when PIP₂ is depleted following the action of PLC. This type of mechanism has been implicated in agonist control of inwardly rectifying K⁺ channels (Lei et al. 2001a; Xie et al. 1999) and cardiac K⁺ channels derived from the human ether-a-go-go related gene (Bian et al. 2001; Huang et al. 1998; Kobrinsky et al. 2000). The observation that U73122 potentiates g_M (Fig. 1) is consistent with this possibility. If inhibition of PLC by U73122 preserves the concentration of PIP₂ in the vicinity of the channels, our hypothesis would predict an increase in g_M.

M channels were first defined in BFSG on the basis of their sensitivity to muscarinic agonists (Adams et al. 1982a,b). Although considerable progress has now been made regarding the transduction mechanisms activated by various agonists that suppress g_M in a variety of cell types (Bofill-Cardona et al. 2000; Bosma and Hille 1989; Cruzblanca et al. 1998; Haley et al. 1998, 2000a,b; Hildebrandt et al. 1997; Kirkwood et al. 1991; Pfaffinger 1988; Pfaffinger et al. 1988; Selyanko et al. 1990; Shapiro et al. 2000), the mechanism that underlies the defining, classical effect of muscarine on these channels still remains to be elucidated. This will be difficult, as our preliminary data suggest that muscarine-induced g_M suppression in BFSG is antagonized both by the PLC inhibitor U73122 and by the inactive isomer, U73343. It therefore remains to be determined whether PLC participates in the transduction mechanism for muscarine.