INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

A consistent finding in essentially all Na⁺-reabsorbing or Cl-secreting epithelia is an increase in basolateral membrane (BLM)¹ K conductance (G_K) concurrent with the activity of the Na⁺,K⁺-ATPase pump, and vice-versa (Schultz, 1992; Beck et al., 1994). Using rabbit kidney cortical tubule suspension, Balaban et al. (1980) directly demonstrated that inhibition of the Na⁺,K⁺-ATPase pump increased the ATP/ADP ratio, and concluded that [ATP]_i, [ADP]_i, or their ratio could be the link coupling cellular metabolism to transport. The demonstration of Type I K_ATP channels in cardiac muscle (Noma, 1983) opened the way to study the regulatory role of ATP at the single-channel level. Since this topic has been treated in several excellent reviews (e.g., see Beck et al., 1994; Lang and Rehwald, 1992; Edwards and Weston, 1993; Ashcroft and Ashcroft, 1990), as well as in Mauerer et al. (1998), the following is limited to renal K⁺ channels.

In the distal nephron, Hunter and Giebisch (1988) reported a Ca²⁺-activated K⁺ channel in the apical membrane of Amphiuma early distal tubule that was inhibited by ATP with a K_i of ~5 mM. This was the first ATP-sensitive K⁺ channel described in epithelia, and was classified as Type 3 (Ashcroft and Ashcroft, 1990) because of its Ca²⁺ activation and lower sensitivity to ATP. Moreover, this "maxi-K" channel was depolarization activated and did not exhibit rundown. Subsequently, Wang et al. (1990) demonstrated that the K⁺ channel in the apical membrane of rat cortical collecting duct, first identified by Frindt and Palmer (1989), was ATP sensitive. This 35-pS channel, which mediates K⁺ secretion by principal cells, is Type 1-like (Ashcroft and Ashcroft, 1990) in that it is insensitive to Ca²⁺, voltage independent, and exhibits rundown (Wang and Giebisch, 1991a; Wang et al., 1990). Thus, the single-channel findings in distal nephron segments were partially consistent with the notion that [ATP]_i could link cell metabolism and K⁺ conductance, although Schultz (1981) had specified that the relevant K conductance was on the BLM. Two K⁺ channels have been identified in the lateral membrane of rat cortical collecting duct (Wang et al., 1994), but they do not appear to be ATP sensitive.

The situation in the proximal tubule, however, is more compatible with homocellular regulation (Schultz, 1981). In their study of Necturus proximal tubule, Matsumura et al. (1984) showed that inhibition of Na⁺,K⁺-ATPase activity using ouabain, low bath K⁺, or low luminal Na⁺ perfusate brought about a fall in BLM G_K. They suggested that this effect might be a metabolic consequence of pump inhibition, but emphasized [Ca²⁺]_i as the proximate signal coupling pump activity to BLM G_K. Once it became evident that Type 1 K_ATP channels were not Ca²⁺ activated, the focus shifted to ATP itself as the link between pump activity and BLM G_K. Beck et al. (1991a) quantitated the inverse relationship between Na⁺ transport and [ATP]_i in perfused rabbit proximal tubule and confirmed a qualitative correlation between [ATP]_i and G_K. Shortly thereafter, Tsuchiya et al. (1992) observed that ATP reversibly inhibited BLM K_ATP channel activity in four of five inside-out patches pulled off the BLM of nonperfused rabbit tubules, but regulation on the single-channel level was not pursued in that study. Beck et al. (1993) looked at regulation of stretch-insensitive BLM K⁺ channels in perfused rabbit tubules and found that the addition of luminal substrates depolarized the cell and increased channel activity. In a follow-up study, the same group (Hurst et al., 1993) showed that this channel could be activated by diazoxide and inhibited by pump inhibition or direct application of ATP. These data, combined with our findings in nonperfused rabbit tubules and dissociated Ambystoma proximal tubule cells (see Mauerer et al., 1998) that maintain epithelial polarity (Segal et al., 1996), clearly show that a Type 1-like K_ATP channel exists on the BLM of the proximal tubule.

In the present study, we have used the dissociated Ambystoma proximal tubule cells (Segal et al., 1996) to investigate the regulation of this BLM K_ATP channel by protein kinases, intracellular nucleotides, pH (pH_i), Ca²⁺, and the cytoskeleton. We also show that regulation of the K_ATP channel is indirectly linked to transport dynamics in the proximal tubule through changes in intracellular [ATP], resulting from altered activity of the Na⁺,K⁺-ATPase pump as transport is modulated.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Solutions and Drugs

The composition of the solutions used is summarized in Table I. After titration to pH 7.5 (710A; Orion Research, Boston, MA), sucrose was added to adjust the osmolality of the solutions (3MO; Advanced Instruments Inc., Needham Heights, MA). KCl solutions containing low levels (50, 100, 200, 500, and 1,000 nM) of free Ca²⁺ were prepared by adding the appropriate amount of CaCl₂ (0.407, 0.579, 0.733, 0.873, and 0.933 mM, respectively) to solution d (Table I). Free Mg²⁺ was maintained at ~1 mM except in solution b (divalent-free NaCl). In solutions containing ATP, the nucleotide was added as the Mg-salt to maintain the free Mg²⁺ at ~1 mM (range 0.98-1.33 mM). Chemicals used were of the highest quality and obtained from Sigma Chemical Co. (St. Louis, MO), except ADP (Boehringer-Mannheim Biochemicals, Indianapolis, IN). Nucleotides were prepared fresh daily as 20- 50-mM stocks in bath solution. Stock solutions of glibenclamide (100 mM), PMA (10 mM), and 4-PMA (10 mM) were dissolved in DMSO. Stock solutions (10 mM) of forskolin, 1,9-dideoxyforskolin (Calbiochem Corp., La Jolla, CA), and nigericin were dissolved in ethanol. The catalytic subunit of the cAMP-dependent protein kinase A (Promega Corp., Madison, WI) or protein kinase C (0.5 U/ml; Promega Corp.) was pipetted directly into the bath.

Cell Preparation

Dissociated proximal tubule cells were isolated from amphibian kidneys as previously described (Segal et al., 1996). Briefly, kidneys from Ambystoma tigrinum were rapidly removed and placed in iced HEPES-buffered NaCl at pH 7.5 (solution a). After incubation in collagenase-dispase (0.2 U/ml collagenase; Boehringer-Mannheim Biochemicals), the enzyme reaction was stopped by washing with Ca²⁺- and Mg²⁺-free NaCl (solution b). Dissociated cells were resuspended in 2.5 ml of NaCl (solution a) in a 35-mm culture dish, and stored at 4°C until use. The dissociated proximal tubule cells retain their epithelial polarity for up to 14 d (Segal et al., 1996). Cells were used for experiments from 2 to 12 d after dissociation. A representative cell is shown in Fig. 1 of Mauerer et al. (1998).

View larger version (29K): [in this window] [in a new window]   Fig. 1.   cAMP agonists activate the BLM KATP channel. (A) Forskolin (10 µM) activates the BLM KATP channel in cell-attached patches. In the experiment depicted, the pipette contains KCl (solution d), the bath contains NaCl (solution c), and the command potential is 60 mV. A running average of current versus time (window width 16 ms) is shown. (B) Sample traces taken from within the regions marked , , and  in A. The corresponding amplitude histograms were taken for each region. After the addition of forskolin, channel activity increases ninefold (nPo increases from 0.17 to 1.46). Note the increase in the single channel current (isc) due to the hyperpolarization of the cell membrane as K+ channels open over the BLM. The all channels closed level (dashed line) and open channel levels (dotted lines) are indicated. (C) Forskolin increases both channel activity (NPo) and single channel current (isc). nPo () and isc () as measured in the experiment described in A are plotted versus time. isc increases as K+ channels on the BLM open. From the change in isc (isc = 0.9 pA), the hyperpolarization (Vm) is ~40 mV. Note that the increase in isc precedes the increase in nPo, consistent with other BLM KATP channels opening before those in the patch.

Electrophysiology, Data Acquisition, and Analysis

The methods are essentially as described in Mauerer et al. (1998). These include the mounting and selection of single proximal cells for electrophysiological study. The standard tight-seal patch-clamp configurations for single-channel recording were used (Hamill et al., 1981), and channel currents were acquired and analyzed as described in the accompanying paper. Briefly, channel activity (nP_o) was calculated over periods of 60-500 s, and open and closed dwell lifetimes were determined as reported elsewhere (Mauerer et al., 1998).

The number of observations or experiments is reported in the text, whereas n in the analysis denotes either the whole data set or the subset of total experiments in which precise quantitation could be reliably applied. In some figures, a running average (using a specified window width) of current versus time is displayed. Statistical values for the n elements are given as mean ± SEM. Student's t test was applied where appropriate.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

The regulation of the BLM K_ATP channel by PKA, PKC, [Ca²⁺]_i, and pH was studied in cell-attached (c/a) and inside-out (i/o) patches. Channel activity in response to perturbations of cell volume was examined in c/a patches, and the effect of membrane stretch and the role of the cytoskeleton was also tested. Finally, the coupling of channel behavior to changes in cellular energy levels and transport activity was investigated.

Forskolin activates the BLM K_ATP channel. The cAMP second messenger system was studied in c/a patches using forskolin (FK), which increases [cAMP]_i by activating adenylyl cyclase. In each experiment, a cell served as its own control. Fig. 1, A and B shows a representative experiment. Under control conditions with KCl in the pipette (solution d) in a NaCl bath (solution c), steady state channel activity was recorded for 3-5 min before the application of 10 µM FK at a command potential of 60 mV. Within 1-2 min, the single-channel current begins to increase, consistent with hyperpolarization of the cell membrane potential (V_m). This is most likely due to the opening of K⁺ channels on the cell membrane that precede those under the patch pipette itself. Within 5-10 min, FK increases the nP_o of the BLM K_ATP channels in the patch of membrane under study.

PKA directly activates the BLM K_ATP channel. To test the hypothesis that the channel is activated by the cAMP-dependent protein kinase, the effect of the catalytic subunit of PKA was tested in i/o patches. To prevent channel rundown, i/o patches were excised in a bath containing 0.2 mM ATP (see Mauerer et al., 1998). After a control period in 0.2 mM ATP, 50-100 U/ml PKA was directly added to the bath in the continued presence of ATP (n = 29). Channel activation occurred after a delay of 0.5-3 min (n = 26). This delay is most likely due to the time necessary for the reagents to diffuse to the patch, and then phosphorylate the channel(s). In some experiments, a low [Cl] bath was used to isolate the effect of PKA on the K⁺ channel since phosphorylation via PKA also activates a cystic fibrosis transmembrane regulator (CFTR)-like Cl channel present in the BLM (Segal et al., 1996). The mean nP_o increase was 5.1 ± 0.7-fold in patches containing K⁺ channels exclusively (n = 6).

View larger version (35K): [in this window] [in a new window]   Fig. 2.   The cAMP-dependent protein kinase directly activates the BLM KATP channel. In this inside-out experiment, the pipette contains KCl (solution d), the bath contains NaCl (solution c) plus 0.2 mM ATP, and the command potential is 60 mV. (A) The running average of current versus time (16-ms window) is shown. The catalytic subunit of PKA (csPKA, 100 U/ml) was added to the bath where indicated. (B) Original current traces representative of channel activity during the control period (top trace) and after stimulation by PKA (bottom three traces). In this experiment, PKA led to a fourfold increase in nPo. The all channels closed level (dashed line) and open channel levels (dotted lines) are indicated.

PKA phosphorylation is necessary but not sufficient for channel activation. ATP can serve as a substrate for (a) a kinase in a phosphorylation reaction, and/or (b) an ATPase in a hydrolysis reaction. It is possible that either or both are required for channel activation, but the use of ATP does not allow the two to be dissociated. To test the hypothesis that phosphorylated channels remain active after the removal of hydrolyzable nucleotides, the poorly hydrolyzable nucleotide ATP-S (0.2 mM) was substituted for 0.2 mM ATP in the presence of PKA. In this condition, the patch can be fully phosphorylated since ATP-S is a substrate for PKA. Indeed, thiophosphorylation may be even more resistant to phosphatases than phosphorylation (Eckstein, 1985). Fig. 3 shows that, after a period of steady channel activity in 0.2 mM ATP, activity begins to decrease upon substitution of ATP with 0.2 mM ATP-S. Subsequent addition of csPKA to thiophosphorylate the patch did not restore activity; in fact, the channels run down. Thus, because ATP-S and PKA were not sufficient to restore or even sustain channel activity (n = 7), it appears that phosphorylation per se is necessary but not sufficient to maximally activate the BLM K_ATP channel. That channel rundown occurs despite the presence of ATP-S and PKA confirms our previous finding that the BLM K_ATP channel also requires submillimolar levels of a hydrolyzable nucleotide (see Mauerer et al., 1998).

View larger version (30K): [in this window] [in a new window]   Fig. 3.   Phosphorylation itself does not prevent channel rundown. Activity of the BLM KATP channel is maintained in this representative inside-out patch excised in a bath containing 0.2 mM ATP, but rundown commences when the poorly hydrolyzable ATP-S is substituted for ATP. Rundown continues despite subsequent addition of PKA, suggesting that phosphorylation per se does not rescue rundown of the BLM KATP channel.

Kinetics of activated channels. Kinetic analysis was performed on the two patches that contained only one channel both before and after stimulation with FK (c/a) or PKA (i/o with 0.2 mM ATP in the bath). This analysis reveals that the major effect of PKA phosphorylation is to activate the channel by shortening, or possibly eliminating, the dwell time of the longest closed state. Under control conditions, the mean lifetime of the longest closed state is ~397 (c/a, Fig. 4 A, top) and 502 (i/o, Fig. 4 B, top) ms (see Mauerer et al., 1998). After activation by FK or the catalytic subunit of PKA, the mean dwell time of this state decreases to 7 (c/a, Fig. 4 A, bottom) and 100 (i/o, Fig. 4 B, bottom) ms, respectively. Note that there is more pronounced shortening occurring in the c/a mode (98%) compared with the i/o mode (80%). Activation by FK or PKA did not affect the open state dwell times.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   PKA phosphorylation shortens the longest closed time of the BLM KATP channel. (A) Closed-time histogram for a KATP channel in a cell-attached patch at 60 mV before (top) and after (bottom) forskolin. (top) Logarithmically binned (Sigworth and Sine, 1987) data before forskolin was fitted with two probability density functions (dashed lines) to give the overall fit (solid line), yielding time constants of c1 = 1.27 ms (74%) and c2 = 397 ms (26%). (bottom) Logarithmically binned data after forskolin was fitted with two probability density functions (dashed lines) to give the overall fit (solid line) yielding time constants of c1 = 0.81 ms (72%) and c2 = 7 ms (28%). Note that the main effect of forskolin is to reduce the longest closed lifetime from 397 to 7 ms. Based on the bandwidth of the recording system, the data and the fit were cutoff at 500 µs (dotted vertical line). (B) Closed-time histogram for a KATP channel in an inside-out patch at 60 mV before (top) and after (bottom) csPKA. (top) Logarithmically binned data before csPKA were fitted with two probability density functions (dashed lines) to give the overall fit (solid line), yielding time constants of c1 = 0.72 ms (48%) and c2 = 502 ms (52%). (bottom) Logarithmically binned data after csPKA was fitted with two probability density functions (dashed lines) to give the overall fit (solid line), yielding time constants of c1 = 1.5 ms (75%) and c2 = 100 ms (25%). As was the case for forskolin, the main effect of PKA phosphorylation was to reduce the longest closed lifetime from 502 to 100 ms. Based on the bandwidth of the recording system, the data and the fit were cutoff at 500 µs (dotted vertical line).

Stimulation of PKC inhibits the BLM K_ATP channel. To determine whether PKC is involved in the regulation of this K⁺ channel, c/a patches were treated with the phorbol ester PMA. As shown in Fig. 5, the application of 10 µM PMA markedly decreases channel activity concomitant with a reduction in i_sc, consistent with depolarization of V_m. Similar results were obtained in eight cells with an average reduction in nP_o of 63 ± 11% (n = 6). Based on an inward slope conductance of 22.2 pS under similar conditions (see Mauerer et al., 1998), the mean decrease of i_sc of 0.28 ± 0.11 pA (n = 6) reflects an average depolarization (V_m) of 12.0 ± 4.6 mV. A phorbol ester that does not activate PKC was used as a control. In five cell-attached patches, the addition of 10 µM 4-PMA did not affect nP_o (Fig. 5).

View larger version (33K): [in this window] [in a new window]   Fig. 5.   Phorbol ester inhibits the BLM KATP channel. (A) PMA (10 µM) inhibits the BLM KATP channel in cell-attached patches. In the experiment depicted, the pipette contains KCl (solution d), the bath contains NaCl (solution c), and the command potential is 60 mV. A running average of current versus time (16-ms window) is shown. (B) Sample traces taken from within the regions marked , , and , and in A. Addition of 4-PMA does not alter channel activity. After the addition of PMA, however, channel activity decreases by 85% (nPo decreases from 0.072 to 0.011). The all channels closed level (dashed line) and open channel levels (dotted lines) are indicated.

View larger version (41K): [in this window] [in a new window]   Fig. 6.   Protein kinase C inhibits the BLM KATP channel. PKC directly inhibits the BLM KATP channel. In this inside-out experiment, the pipette contains KCl (solution d), the bath contains NaCl (solution c) plus 0.2 mM ATP and 50 nM Ca2+, and the command potential is 60 mV. (A) The running average of current versus time (16-ms window) is shown. PKC (0.5 U/ml) was added to the bath where indicated. (B) Original current traces representative of channel activity during the control period (top trace) and after inhibition by PKC (bottom two traces). In this experiment, PKC produced an 89% decrease in nPo. The all channels closed level (dashed line) and open channel levels (dotted lines) are indicated.

pH Sensitivity

The activity of the BLM K_ATP channel is pH sensitive. Directly lowering the pH on the cytoplasmic side of an i/o patch from 7.5 to 6.8 (a fivefold increase in [H⁺]_i) decreases channel activity by 81 ± 2% (n = 3, Fig. 7 A). To assess whether this effect occurs under more physiological conditions, an intracellular acidosis was generated during cell-attached recording using the K⁺/H⁺ exchanger nigericin (Margolis et al., 1989) in a low K⁺ bath (solution c, Fig. 7 B). We have previously shown that under these conditions, pH_i decreases by ~0.4 pH units in the presence of 10 µM nigericin (Segal et al., 1992). This protocol decreases the nP_o of the BLM K_ATP channel by 61 ± 3% (n = 3).

View larger version (39K): [in this window] [in a new window]   Fig. 7.   Intracellular acidosis inhibits the BLM KATP channel. (A and B) Lowering pHi decreases channel activity in inside-out patches. In this inside-out experiment, the pipette contains KCl (solution d), the bath contains NaCl (solution c) plus 0.2 mM ATP, and the command potential is 60 mV. (A) The running average of current versus time (16-ms window) is shown. When the pH of the bath was lowered from 7.5 to 6.8, channel activity was inhibited by 79% (nPo changes from 0.67 to 0.14). This effect is reversible upon returning the bath pH to 7.5. (B) Original current traces representative of channel activity during the control period (pHi = 7.5, top), during acidosis (pHi = 6.8, middle), and after returning to pHi = 7.5 (bottom). The all channels closed level (dashed line) and open channel levels (dotted lines) are indicated. (C and D) Lowering pHi with nigericin decreases channel activity in cell-attached patches. In this cell-attached experiment, the pipette contains KCl (solution d), the bath contains NaCl with 2.5 mM KCl (solution c) plus 0.2 mM ATP, and the command potential is 80 mV. (C) The running average of current versus time (16-ms window) is shown. When nigericin (10 µM) was added to the bath, channel activity was progressively inhibited. nPo was reduced by 68% after 8 min in nigericin. (D) Original current traces and amplitude histograms representative of channel activity during the control period (top trace) and during nigericin (bottom two traces). The all channels closed level (dashed line) and open channel levels (dotted lines) are indicated.

Effect of Calcium

The BLM K_ATP channel does not require Ca²⁺ for activity since excising membrane patches into an EGTA-buffered Ca²⁺-free solution (solution c) does not alter channel current. However, increases in [Ca²⁺]_i inhibit the channel. The addition of the calcium ionophore ionomycin has no effect on cell-attached channel currents recorded from cells bathed in Ca²⁺-free solution (Fig. 8, A and B). However, in the continued presence of ionomycin, raising bath Ca²⁺ ([Ca²⁺]_o) to 1 µM (solution e) promptly and significantly reduces nP_o by 63 ± 9% (n = 4). This effect appears to be direct since exposing the cytoplasmic side of excised patches to varying [Ca²⁺]_i produces a dose-dependent decrease in channel activity with a K_i = 166 nM (Figs. 8, C and D). However, since the excised membrane may still contain elements of Ca²⁺-activated signal transduction pathways (e.g., Ca²⁺/calmodulin kinase, PKC), a possible indirect effect cannot be fully ruled out. Indeed, the Hill coefficient of 0.73 may suggest that Ca²⁺ has both direct and indirect effects. In any case, an increase in [Ca²⁺]_i leads to a decrease in channel activity.

View larger version (48K): [in this window] [in a new window]   Fig. 8.   Increases in [Ca2+]i inhibit the BLM KATP channel. (A and B) Elevation of [Ca2+]i decreases channel activity. (A) Summary of normalized channel activity from cell-attached patches in which bath [Ca2+] was increased in the presence of a calcium ionophore (ionomycin). The addition of 1 µM ionomycin in a Ca2+-free bath (solution c) does not significantly affect nPo (P = 0.72), but subsequent addition of 1 µM Ca2+ (solution e) leads to a 63 ± 9% (n = 4) reduction in nPo compared with control (P < 0.001). The inhibition tends to persist despite the removal of Ca2+ from a bath containing EGTA. (B) Representative single-channel traces and corresponding amplitude histograms for Ca2+-free bath (top), Ca2+-free bath plus ionomycin (middle), and 1 µM Ca2+ bath plus ionomycin (bottom). The all channels closed level (dashed line) and open channel levels (dotted lines) are indicated. (C and D) Increasing [Ca2+]i in inside-out patches directly inhibits channel activity. (C) Dose-response curve of channel activity versus [Ca2+]i. Fitting the data with the Hill equation (solid line) yields a Ki = 166 nM and nH = 0.73. (D) Representative single channel traces for [Ca2+]i = 0 (top) and 200 (bottom) nM at a command potential of 60 mV.

Membrane stretch. Several classes of cationic channels are known to be stretch-activated, including a K⁺ channel on the BLM of Necturus proximal tubule (Sackin, 1989). Application of negative pressure up to 60 mm Hg to the patch pipette using a calibrated electronic negative pressure generator (DPM-1B; Bio-Tek Instruments, Inc., Winooski, VT) had no detectable effect on the nP_o of this BLM K_ATP channel in either c/a or i/o patches (n > 100). Thus, the BLM K_ATP channel of Ambystoma proximal tubule is not stretch activated and is distinct from the SA-K found on the BLM of Necturus proximal tubule (Filipovic and Sackin, 1992).

Cell swelling. To assess whether the BLM K_ATP channel might be involved in cell volume regulation, nP_o in c/a patches was monitored as measured bath tonicity was reduced by 30% from 200 (solution g) to 140 (solution h) mosM/kg. Despite clearly observable cell swelling due to the hypotonic shock, K_ATP channel activity did not increase and actually tended to decrease. This slight inhibition may be mediated either by an increase in [Ca²⁺]_i and/or activation of PKC, since cell swelling is often associated with an increase in [Ca²⁺]_i (Robson and Hunter, 1994a, 1994b; Ubl et al., 1988). Nonetheless, measurements made in zero current clamp showed that V_m hyperpolarizes (by 15-20 mV) during hypotonicity, probably due to the opening of a swelling-activated K⁺ exit pathway. It does not appear that the BLM K_ATP channel mediates swelling-activated K⁺ efflux.

Cytoskeleton. The cytoskeleton is a major factor in maintaining epithelial polarity in the dissociated cells (Segal et al., 1996). It is also well established that cytoskeletal elements are important in anchoring some proteins to their membrane (Morrow et al., 1989). More recently, it has been demonstrated that Na⁺ (Prat et al., 1993), K⁺ (Wang et al., 1994), and CFTR Cl channels (Prat et al., 1995) can be regulated by the cytoskeleton.

View larger version (34K): [in this window] [in a new window]   Fig. 9.   Disrupting the actin cytoskeleton has a biphasic effect on the BLM KATP channel. Cytochalasin D (10 µM) has a biphasic effect on the BLM KATP channel. In this cell-attached patch, the pipette contains KCl (solution d), the bath contains NaCl (solution c), and the command potential is 60 mV. (A) A running average of current versus time (16-ms window) shows that channel activity increases within the first minute of treatment, but gradually declines thereafter. Irreversible inhibition of channel activity persists after washout of cytochalasin D from the bath. (B) Sample traces taken from within the regions marked , , , and  in A. Corresponding amplitude histograms are given for each region. After the addition of cytochalasin D, channel activity initially increases, but subsequently declines to 18% of control activity (region marked : nPo = 0.14, compared with 0.78 during region ). The inhibition persists after washout of cytochalasin D (region marked : nPo = 0.01, 1% of control). Note the decrease in the single channel current (isc) due to the depolarization of the cell membrane as K+ channels close over the BLM. The all channels closed level (dashed line) and open channel levels (dotted lines) are indicated.

Functional Coupling of the BLM K_ATP Channel to Na⁺,K⁺-ATPase Pump Activity

Sustained reabsorption of salt and water in the proximal tubule requires the action of the Na⁺,K⁺-ATPase pump on the BLM. As K⁺ ions are pumped into the cell, a basolateral exit pathway must exist to allow K⁺ to recycle, thus permitting continuous operation of the pump. The functional coupling of these two transport elements may be linked by the intracellular [ATP] ([ATP]_i) level. This was examined in c/a patches.

Inhibition of transport. Retarding pump activity should result in less ATP consumption and a rise in [ATP]_i. This was achieved either using either (a) cardiac glycosides (ouabain or strophanthidin), or (b) removal of extracellular K⁺. Both methods were used in an attempt to dissociate the rise in [ATP]_i from a voltage-dependent effect, since the former depolarizes V_m while the latter (at least initially) hyperpolarizes V_m. However, it must be noted that these maneuvers result in concomitant changes in V_m, [ATP]_i, and even [K]_i. Since nP_o of the K_ATP channel decreases with depolarization (see Fig. 3 C in Mauerer et al., 1998), it is not possible to be certain how much of the fall in nP_o is due to V_m versus [ATP]_i. Applying any of these maneuvers resulted in a gradual decrease of BLM K_ATP channel activity, as illustrated in Fig. 10.

View larger version (40K): [in this window] [in a new window]   Fig. 10.   Inhibition of transport decreases BLM KATP channel activity. (A) Inhibition of the Na+,K+-ATPase pump with ouabain (200 µM) results in a decrease in activity of the BLM KATP channel in cell-attached patches. A running average of current versus time (16-ms window) is shown. The application of ouabain caused nPo to decrease by 89% (region marked : nPo = 0.46, region marked : nPo = 0.05) presumably due to an increase in [ATP]i. In this experiment, the pipette contains KCl (solution d), the bath contains NaCl (solution c), and the command potential is 60 mV. (B) Sample traces taken from within the regions marked , , and  in A. The corresponding amplitude histograms were taken for each region. Note the decrease in the single channel current due to the depolarization of the cell membrane as K+ channels close over the BLM. The small channel appearing in the lower two traces is a CFTR-like Cl channel on the BLM previously described by us. An increase in [ATP]i may have led to the opening of the Cl channel. The all channels closed level (dashed line) and open channel levels (dotted lines) are indicated. (C) Inhibition of the Na+,K+-ATPase pump with a zero-K+ bath (solution c without KCl) also results in a decrease in activity of the BLM KATP channel in cell-attached patches. Sample traces taken at the times indicated show a progressive decrease in nPo (68%) and single channel current (control: nPo = 1.6, isc = 1.6. Zero K+, 6 min: nPo = 0.51, isc = 1.39) corresponding to duration in the zero-K+ bath. In this experiment, the pipette contains KCl (solution d) and the command potential is 80 mV. The all channels closed level (dashed line) and open channel levels (dotted lines) are indicated.

Stimulation of transport. A rise in intracellular [Na⁺] is the primary physiological stimulus for the Na⁺,K⁺-ATPase pump. Na⁺ entry via apical membrane Na-coupled cotransporters was enhanced by the addition of glucose (5 mM), alanine (5 mM), or both (2.5 mM each) to a NaCl bath while recording K_ATP channel activity from the BLM. Fig. 11 depicts a representative experiment in which 5 mM alanine was added to the bath. Within a minute of substrate addition, BLM K_ATP channel activity markedly increased despite an initial depolarization due to Na⁺ entry. Comparable effects were seen in 17 of 19 (89%) cells exposed to glucose and/or alanine, with an approximately twofold increase in nP_o (1.95 ± 0.20, n = 5).

View larger version (48K): [in this window] [in a new window]   Fig. 11.   Stimulation of transport increases BLM KATP channel activity. (A) Activation of the Na+,K+-ATPase pump with substrates, in this case 5 mM alanine, results in an increase in activity of the BLM KATP channel in cell-attached patches. A running average of current versus time (16-ms window) is shown. The addition of alanine caused nPo to increase nearly twofold (region marked : nPo = 2.53, region marked : nPo = 4.44). In this experiment, the pipette contains KCl (solution d), the bath contains NaCl (solution c), and the command potential is 60 mV. Sample traces taken from within the regions marked , , and  in A. Note that despite the opening of BLM KATP channels, the single channel current remains essentially constant due to a net balance of hyperpolarizing (K+ channel activation) and depolarizing (Na+ entry) forces. The all channels closed level (dashed line) and open channel levels (dotted lines) are indicated.

View larger version (63K): [in this window] [in a new window]   Fig. 12.   Stimulation of transport does not increase BLM KATP channel activity in ATP-loaded cells. Preloading the cell with ATP from the bath prevents the increase of BLM KATP channel activity due to alanine in cell-attached patches. A running average of current versus time (16-ms window) is shown in A, and sample traces are shown in B. Once inward K+ channel currents reached a steady state (region , up to 11 open channels), 2 mM ATP added to the bath is taken up by the cell, which maintains a constant [ATP]i (see text for details). The initial activation of current (region , up to 20 open channels) may represent the opening of KATP channels that had been quiescent due to a lack of ATP. Presumably, as [ATP]i continues to rise, KATP currents are inhibited (region , up to 9 open channels). In the presence of steady state currents and [ATP]i, the ability of alanine to increase channel activity is significantly diminished (region , up to 10 channels open). Similar results were observed in two other cells. In this experiment, the pipette contains KCl (solution d), the bath contains NaCl (solution c), and the command potential is 60 mV. The all channels closed level (dashed line) and open channel levels (dotted lines) are indicated.

ATP:ADP ratio. Stimulation of transport leads to activation of the Na⁺,K⁺-ATPase pump while the resultant fall in [ATP]_i disinhibits the BLM K_ATP channel. However, the increased turnover of the Na⁺,K⁺-ATPase pump produces a rise in [ADP]_i and/or a fall in the [ATP]_i:[ADP]_i ratio. It is not clear which of these signals is the primary determinant of BLM K_ATP channel activity in intact cells. This issue was examined in i/o patches in two ways: (a) varying the ratio of [ATP]: [ADP] while keeping the [ATP] constant, and (b) varying the ratio of [ATP]:[ADP] while keeping the sum of [ATP] + [ADP] constant. The latter was performed for sums of 0.5, 2.5, and 5 mM.

View larger version (31K): [in this window] [in a new window]   Fig. 13.   ADP does not relieve the block by ATP in excised patches. (A) Neither ADP nor a decrease in the ATP:ADP ratio antagonize the block by 5 mM ATP. In this inside-out patch, the pipette contains KCl (solution d), the bath contains K-aspartate (solution i) to isolate K+ current, and the command potential is 60 mV, and a running average of current versus time (16-ms window) is shown. Adding 5 mM ATP and 5 mM ADP blocked 96% of the inward current. After washout (with 0.2 ATP), 5 mM ATP alone produced nearly the same inhibition (92%), indicating that ADP enhances rather than attenuates the ATP block. This notion is confirmed after washout, when 10 mM ADP in the presence of 5 mM ATP again enhances the block (98%), despite an ATP:ADP ratio of 0.5. (B) Varying the ATP:ADP ratio while keeping the [ATP + ADP] constant does not antagonize the block by ATP. In this inside-out patch, the pipette contains KCl (solution d), the bath contains K-aspartate (solution i) to isolate K+ current, and the command potential is 60 mV, and a running average of current versus time (16-ms window) is shown. The extent of block exerted by [ATP + ADP] = 2.5 mM was not significantly affected, while the ATP:ADP ratio varied from 12.5 to 0.08.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The properties of the BLM K⁺ (K_ATP) channel were the subject of the companion paper (Mauerer et al., 1998). The regulation of basolateral K⁺ channels was not examined in most previous studies (Hunter, 1991; Sackin and Palmer, 1987; Kawahara et al., 1987; Parent et al., 1988; Gögelein and Greger, 1987). In the present study, we have investigated the regulation of this Type 1-like K_ATP channel in the BLM of dissociated Ambystoma proximal tubule cells that maintain epithelial polarity (Segal et al., 1996).

The BLM K_ATP Channel Is Regulated by PKA and PKC, but in Opposing Directions

The proximal tubule BLM K_ATP channel is activated by PKA, both in cell-attached patches via stimulation of adenylyl cyclase-elevating cAMP levels, or directly by PKA itself in inside-out patches. In contrast, this channel is inhibited by PKC, either by treating cell-attached patches with phorbol ester, or directly by PKC in excised patches. Opposite regulation by PKA and PKC has been reported for other K channels (Chen and Yu, 1994; Fakler et al., 1994) including K_ATP channels (Bonev and Nelson, 1993; Honore and Lazdunski, 1993; Zhang et al., 1994) in the kidney (Wang and Giebisch, 1991b).

Activation by PKA. In cell-attached experiments, stimulation of endogenous cAMP-dependent protein kinase (PKA) increases channel activity and single-channel current, the latter most likely due to hyperpolarization of the cell secondary to K⁺ channel opening. Activation of channel by direct exposure of cell-free membrane patches to the catalytic subunit of PKA strongly suggests that phosphorylation of a PKA site on the channel itself (or a closely associated protein) increases the mean open time. PKA increased channel activity by 5.1 ± 0.7-fold, compared with 7.50 ± 1.55-fold in the case of forskolin. This difference is at least in part due to the additional stimulatory effect of membrane hyperpolarization (see Mauerer et al., 1998) in the cell-attached experiments. Kinetic analysis limited to patches containing one channel reveals that PKA phosphorylation does not significantly affect the open states, rather the major effect appears to be shortening of the longest closed state.

Inhibition by PKC. In contrast to the effect of FK, treatment of cell-attached patches with PMA, a phorbol ester known to stimulate protein kinase C, results in a decrease in channel activity and depolarization as K⁺ channels close. Channel activity did not change in control experiments with 4-PMA, a phorbol ester that does not stimulate PKC. Subsequent removal of PMA did not restore channel activity, suggesting that phosphorylation by PKC is long lasting or that adequate specific phosphatase activity was not present. In inside-out patches from rabbit ventricular myocytes, Light et al. (1995) found PKC reversibly inhibited the K_ATP channel, and that a phosphatase inhibitor (okadaic acid) prevented channel recovery. These observations strongly suggested that the membrane patch contained an endogenous protein phosphatase. For the BLM K_ATP channel, the irreversible inhibitory effect of PKC in inside-out patches strongly suggests that such a phosphatase was not present in our membrane patches. Future experiments will be aimed at determining the role of phosphatases in the regulation of the BLM K_ATP channel.

Metabolic Regulation of the BLM K_ATP Channel

Schultz (1981) emphasized the implications of the Koefoed-Johnsen and Ussing model originally developed for frog skin (Koefoed-Johnsen and Ussing, 1958) as applied to Na-transporting epithelia operating over a wide range of transport rates, and reviewed the data then emerging that the K conductance of the basolateral membrane was correlated to pump activity, an example of what Schultz termed a "homocellular regulatory mechanism." Accordingly, Na⁺ entry across the apical membrane should be matched by basolateral Na⁺ efflux, thus maintaining a low intracellular [Na⁺] as Na⁺ is transported across the epithelium. This process is mediated by cross talk between opposing membranes. Schultz's group reported cross talk operating in leaky epithelia when apical Na⁺ entry was increased in amphibian small intestine via rheogenic Na-coupled cotransport (Grasset et al., 1983; Gunter-Smith et al., 1982). They found that addition of alanine to the mucosal solution promptly increased the Na conductance of the apical membrane (with depolarization), and secondarily brought about an increase in K conductance of the BLM (with hyperpolarization). In the Koefoed-Johnsen and Ussing model, maintenance of unidirectional Na transport requires that K moves in a closed circuit (recycling) across the BLM. Coordination of pump activity and basolateral K conductance is necessary to avoid large fluctuations in intracellular [K⁺] and volume, and to maintain a driving force for Na entry across the apical membrane.

It is important to recognize that pump activity and BLM K conductance can be coordinated in parallel or in series. For instance, if the initial perturbation is considered to be the incremental change in cell Na⁺ content due to increased apical Na⁺ entry, pump activity and BLM K conductance could be coupled via: (a) an increase in cell volume (Sackin, 1987, 1989; Cemerikic and Sackin, 1993; Kawahara, 1990), (b) increased [Ca²⁺]_i secondary to either a decrease in Na/Ca exchange (Yang et al., 1988) or swelling-induced opening of Ca²⁺ channels (McCarty and O'Neil, 1992), (c) an increase in pH_i dependent on the relative fractions of Na entering via the Na/H exchanger and Na-coupled cotransporters (Ohno-Shosaku et al., 1990; Beck et al., 1993), (d) increased [K⁺]_i due to pump activation (Messner et al., 1985b), (e) hyperpolarization of V_m due to pump activation (Lapointe and Duplain, 1991), and/or (f) decreased [ATP]_i due to pump activation. Note that the first set of three signals simultaneously mediate changes in both pump activity and BLM K conductance in parallel; the latter three signals are already the result of activation of the pump and mediate, in series, the increase in BLM K conductance (G_K). The behavior of the BLM K_ATP channel in regards to the first set of three will be considered separately, while the last set of three will be discussed together in the context of pump activity.

Cell volume, stretch, and the cytoskeleton. An increase in apical Na⁺ entry will increase steady state cell volume (Beck et al., 1991a), which has been shown to lead to a rise in BLM G_K (Beck et al., 1991b; Macri et al., 1993). It is likely that multiple factors are involved in this response, including signal transduction via the cytoskeleton and effects of membrane deformation (e.g., membrane stretch; for review, see Sackin, 1995). Swelling- (Sackin, 1989) and stretch-activated (Kawahara, 1990) K channels on the BLM of Xenopus, Rana (Cemerikic and Sackin, 1993), and Necturus (Sackin, 1989) proximal tubule cells exist, but they are not ATP sensitive. Although we have detected stretch-activated channels (Segal and Boulpaep, 1994) on the BLM of the Ambystoma cells, they are not K selective. We and others (Parent et al., 1988; Beck et al., 1993) have shown that rabbit BLM K⁺ channels are not stretch activated. The Ambystoma BLM K_ATP channel reported in the present study was not affected by stretch or hypotonic swelling and thus is analogous to the K_ATP channels described in mammalian preparations, which are also insensitive to membrane stretch (Beck et al., 1993; Tsuchiya et al., 1992).