The cell-attached configuration of the patch-clamp technique was used to investigate the effects of taurine on the basolateral potassium channels of rabbit proximal convoluted tubule. In the absence of taurine, the previously reported ATP-blockable channel, KATP, was observed in 51% of patches. It is characterized by an inwardly rectifying current-voltage curve with an inward slope conductance of 49 ± 5 pS (n = 15) and an outward slope conductance of 13 ± 6 pS (n = 15). The KATP channel open probability (P o) is low, 0.15 ± 0.06 (n = 15) at a −V p = −100 mV (V pis the pipette potential), and increases slightly with depolarization. The gating kinetics are characterized by one open time constant (τo = 5.0 ± 1.9 ms,n = 6) and two closed time constants (τC1 = 5.2 ± 1.5 ms, τC2 = 140 ± 40 ms;n = 6). In 34% of patches, a second type of potassium channel, sK, with distinct properties was recorded. Its current-voltage curve is characterized by a sigmoidal shape, with an inward slope conductance of 12 ± 2 pS (n = 4). ItsP o is voltage independent and averages 0.67 ± 0.03 (n = 4) at −V p = −80 mV. Both its open time and closed time distributions are described by a single time constant (τo = 96 ± 19 ms, τC = 10.5 ± 3.6 ms;n = 4). Extracellular perfusion of 40 mM taurine fails to affect sK channels, whereas KATP channelP o decreases by 75% (from 0.17 ± 0.06 to 0.04 ± 0.02,n = 7,P < 0.05). In conclusion, the absolute basolateral potassium conductance of rabbit proximal tubules is the resulting combination of, at least, two types of potassium channels of roughly equal importance: a high-conductance low-open probability KATP channel and a low-conductance high-open probability sK channel. The previously described decrease in the basolateral absolute potassium conductance by taurine is, however, mediated by a single type of K channel: the ATP-blockable K channel.
- patch clamp
- regulatory volume decrease
- potassium conductance
- adenosine 5′-triphosphate-dependent potassium channel
- low-conductance potassium channel
basolateral potassium channels have a very important role to play in the transepithelial reabsorption mediated by proximal convoluted tubules (PCT). Indeed, the sodium-dependent translocation of luminal substrates requires the maintenance of a cellular Na+ electrochemical gradient, which is controlled by the basolateral Na-K-ATPase, coupled to the basolateral potassium conductance (3, 14, 17, 29). Sodium ions, cotransported with the luminal substrates, are extruded by the basolateral pump, while potassium ions pumped into the cell by the Na-K-ATPase are recycled by the basolateral potassium conductance and create the negative intracellular potential.
In rabbit proximal tubules, potassium diffusion across the basolateral membrane is known to be mainly mediated by an inwardly rectifying channel with an inward unitary conductance of ∼50 pS (11, 16, 19). This potassium channel is inhibited by raising intracellular ATP (14, 24) and by lowering intracellular pH (2). In addition to these regulators, we have recently found that an increase in intracellular taurine concentration inhibits the basolateral potassium conductance (G K) by 50% (4). Very interestingly, the basolateral taurine permeability was shown to be activated by cell swelling (4), which suggests that taurine could play a significant role in maintaining cell volume and membrane potential during stimulation of transepithelial sodium transport. First, the swelling-activated basolateral taurine efflux would directly contribute to the volume regulatory decrease. Then, the resulting decrease of intracellular taurine concentration would increase the potassium conductance and facilitate further volume and membrane potential restoration by enhancing the potassium efflux.
In this study, we investigated the effect of taurine on the basolateral potassium conductance, at the single-channel level, using the patch clamp technique in the cell-attached configuration. This allowed the identification of a previously unrecognized potassium channel in the basolateral membrane of PCT, which is responsible for ∼40% of the membrane conductance. At the single- channel level, only the previously identified 50-pS channel was proved to be taurine sensitive.
MATERIALS AND METHODS
Tubule preparation. Female New Zealand White rabbits were decapitated, and the left kidney was perfused with a cold (4°C) preservation solution containing (in mM) 56 Na2HPO4, 13 NaH2PO4, and 140 sucrose (5). Thin kidney slices were obtained and placed immediately in the cold preservation fluid. Segments of S1 and S2 cortical PCT were dissected from the midcortical region under ×40 magnification at 4°C in the same solution. Once dissected, PCTs were transferred to the perfusion chamber and bathed, at room temperature, in the control solution, buffered at pH 7.4 and containing (in mM) 30 NaCl, 5 KCl, 1.2 MgSO4, 1.8 CaCl2, 1 NaH2PO4, 3 Na2HPO4, 4 sodium acetate, 1 trisodium citrate, 25 NaHCO3, 10 glucamine chloride, 128 mannitol, 5.5 glucose, and 6 alanine. Tubules were held by micropipettes and placed in the flux of a perfusion system as previously described (18). To gain access to the basolateral membrane, the basal membrane was removed by a 20-min incubation with collagenase A (Boehringer Mannheim) added to the control solution (final enzymatic activity of 0.5 U/ml) (2). All experiments were conducted on collapsed tubules.
Electrophysiology. The cell-attached configuration of the patch-clamp technique was used to record currents from basolateral membrane channels. Patch pipettes were fabricated from hematocrit capillary tubes (Fisher) using a Narishige PP-83 two-stage patch pipette puller. Pipettes filled with a solution containing (in mM) 150 KCl, 2 CaCl2, 1 MgCl2, and 10 HEPES (titrated to pH 7.4 with 2.4 NaOH) had resistances between 5 and 10 MΩ when placed in the control solution. Ionic selectivity was studied in inside-out configuration. Micropipette solution contained (in mM) 150 KCl, 2 CaCl2, 1 MgCl2, 10 HEPES, 2.4 NaOH, and 20 mannitol. In symmetrical potassium gradients, bath solution contained (in mM) 140 KCl, 2 MgCl2, 5 EGTA, 10 HEPES, 18 NaOH, 5 glutathione, and 35 mannitol. In asymmetrical potassium gradients, 70 mM NaCl was substituted for 70 mM KCl of the bath solution.
The pipettes were advanced onto the basolateral membrane of tubules by means of a hydraulic micromanipulator (model MX 630R; Newport, Irvine, CA). The contact between microelectrode and membrane was monitored by detecting change in the membrane shape before applying suction to obtain a gigaohm seal. Channel currents amplified with a patch-clamp amplifier (Axopatch-1D; Axon Instruments, Foster City, CA) were recorded onto videotape through a digital data recorder (model VR-10B CRC; Instrutech, Great Neck, NY). To analyze data, records were low-pass filtered with an eight-pole Butterworth filter (model 901; Frequency Devices, Haverhill, MA) at 500 Hz, digitized at 2.5 kHz with a TL-1 DMA Labmaster interface (Axon Instruments), and stored in a 486 PC by means of a specific acquisition software (Axotape 1.2.01, Axon Instruments). Digitized records with very low-amplitude events were additionally filtered at 300 Hz (with a digital Gaussian filter) to improve peak discrimination in amplitude histograms. When such a cutoff frequency was needed, no kinetic analysis was performed. Channel currents were analyzed with the pClamp 5.5.1 software (Axon Instruments). Channel conductance was estimated from linear regression of single-channel current-voltage curve between −40 mV and +20 mV for inward currents and between +60 mV and +100 mV for outward currents. Reversal potentials were obtained by interpolation using a third-order polynomial fit of the current-voltage curves. Neither a linear, a quadratic, nor a Goldman-Hodgkin-Katz equation could provide an acceptable fit of these curves. The open probability (P o) was calculated from amplitude histograms according to the equation Equation 1where i is the number of channels observed at the same time,Pi is the probability that i channels are simultaneously open, and N is the total number of channels in the patch. In some experiments, open probability at a given potential was plotted as a function of time. In this case, recordings at constant potential were divided into 30-s segments from which oneP o was calculated. In the text and legends to Figs. 1-8, potentials are given as −V p, whereV p is the micropipette potential.
Experimental values are expressed as means ± SE. Student’st-test for unpaired observations is used for statistical comparisons.
With control solution bathing the tubule, cell-attached experiments confirmed the presence of the ATP-dependent potassium channel, KATP, which was previously identified at the basolateral membrane of the PCT (2, 14, 19, 24). Figure1 A shows the typical KATP channel activity, which was observed in 51% of the patches showing some channel activity. At a potential of −40 mV, the amplitude of inward currents was −3.2 ± 0.5 pA (n = 14). Interestingly, 34% of recordings performed in the same experimental conditions demonstrated the presence of a previously unrecognized potassium channel (sK). As seen in Fig. 1 B, this channel is characterized by smaller inward currents at −40 mV (−1.4 ± 0.2 pA, n = 5,P < 0.05) and by a more frequent open state. In the remaining 15% of experiments, the two channel types were observed together in the same patch (Fig.1 C). The fact that one type of channel could be recorded independently of the other in some patches and that independent gating of the two channels was observed when they were present in the same patch strongly suggest the presence of distinct channels.
Characteristics of sK vs. KATP.
The current-voltage curves were plotted for the two types of channel in cell-attached configuration when the tubule is bathed in the control solution (Fig. 2). The KATP channel demonstrated an inward rectification, with an inward slope conductance of 49 ± 5 pS (n = 15) and an outward slope conductance of 13 ± 6 pS (n = 15). Concerning the sK channel, currents were characterized by a sigmoidal current-voltage relation with an inward slope conductance (measured between −V p= −40 mV and −V p = 20 mV) equal to 12 ± 2 pS (n = 4), which is significantly lower than the KATP channel inward conductance. Reversal potentials (E rev) for KATP and sK channels were equal to +53 ± 14 mV (n = 15) and +48 ± 6 mV (n = 4), respectively. As we previously demonstrated that the potassium-to-sodium permeability ratio (P K/P Na) of the KATP channel was greater than 7 (2), we assume that the reversal potential of KATP currents corresponds to the −V p value for which potassium ions are at their equilibrium. This already suggests that sK channels display a good selectivity for potassium ions.
A better evidence of the ionic selectivity was brought by experiments made in inside-out patch that allows the control of both membrane potential and ionic gradients (Fig. 3). With nearly symmetrical potassium concentrations in the pipette and in the bath (140 mM bath/150 mM pipette,E K = +1.7 mV), reversal potentials for KATPcurrents and sK currents were equal to 1.1 ± 0.6 mV (n = 12, Fig.3 A) and 1.7 ± 1.5 mV (n = 6, Fig.3 B), respectively. When equilibrium potential for K+ ions was displaced by reducing bath concentration to 70 mM (70 mM bath/150 mM pipette, E K = +19.3 mV), the reversal potential of KATP channels shifted to 19.3 ± 0.9 mV (n = 6, Fig.3 A), indicating a channel almost perfectly selective for potassium. The smallest observed value forE rev led to a minimum permeability ratioP K/P Naof 26. In similar conditions, reversal potential of sK currents shifted to 18.0 ± 0.7 mV (n = 13, Fig.3 B), corresponding to a minimumP K/P Naof 13.
The KATP and sK channels can also be distinguished by considering their open probability (P o), as illustrated in Fig. 4. The KATP channel was characterized by a low P o slightly dependent on membrane potential, going from 0.15 ± 0.06 (n = 15) at −V p = −100 mV to 0.35 ± 0.09 at −V p = +80 mV. In contrast, the sK channel open probability was significantly higher (P < 0.001) over the whole potential range. It ranged from 0.67 ± 0.03 (n = 4) at −V p = −80 mV to 0.53 ± 0.08 at −V p = 0 mV and did not vary significantly with membrane potential.
The kinetic parameters of the two types of potassium channels were also investigated by means of an analysis of the open and closed dwell-time distributions (see Fig. 5 for a representative example). For −V p = −40 mV, the gating of the KATP channel was characterized by brief openings between short or long closed periods (Fig.1 A). Open dwell-time histograms were well fitted by a single exponential function with an average time constant, τo, of 5.0 ± 1.9 ms (n = 6). On the other hand, adjustment of the closed dwell-time histogram led to determination of a fast (τC1) and a slow (τC2) time constant equal to 5.2 ± 1.5 and 140 ± 40 ms (n = 6), respectively. Contrary to the KATP channel, recordings of the small potassium channel (sK) demonstrated that the open state was the most often observed event (Fig. 1 B). Analysis of the open and closed dwell-time histograms allowed us to calculate a single time constant for each of the states (Fig. 5). Moreover, the open time constant, τo = 96 ± 19 ms (n = 4), was slower than the closed time constant, τC = 10.5 ± 3.6 ms (n = 4), which explains the high open probability of the sK channel.
Effects of taurine. The preceding results demonstrate that two potassium channels, with very distinct properties, are present in the basolateral membrane of the PCT. As previous data on the inhibitory effect of taurine were obtained at the macroscopic level, it became even more interesting to investigate the inhibitory effect of taurine at the single channel level.
After a 4-min period in control solution, the taurine solution (40 mM replacing mannitol in the control solution) was perfused on the basolateral side of the tubule during another 4-min period. Then, taurine was removed by perfusing the control solution. During the entire experimental period, −V p was held to −40 mV and currents were continuously recorded.
We first tested the effect of taurine on the small potassium channel. Figure 6 Aillustrates typical recording of the sK channel under the three experimental periods. It appeared from recordings that neither the channel gating nor the inward current amplitude were affected by the peritubular taurine perfusion. The sK channel open probability as averaged for four experiments was plotted as a function of time in Fig.6 B. It confirms that taurine does not have any effect on the sK channel open probability, as fluctuations ranging from 0.75 to 1 are very likely related to the relatively short time interval (30 s) used to calculateP o.
The same type of experiment was performed on the KATP channel. Figure7 Aillustrates the effect of taurine on the channel activity. When taurine was perfused, a progressive decrease of channel openings was observed within the first minute of perfusion. The maximum effect was reached after 4 min of perfusion. Replacing taurine solution by control did not restore the KATP channel activity. To quantify the effects of taurine, KATP channel open probability was plotted as a function of time for seven experiments (Fig.7 B). In presence of taurine, KATP channelP o decreased from 0.17 ± 0.06 (n = 7) at the beginning of the perfusion (t = 0 min) to 0.04 ± 0.02 (n = 7,P < 0.05) att = 4 min. This decrease inP o was poorly reversible after returning to control conditions, as one would expect if taurine was to accumulate inside the cell. The analysis of gating kinetics demonstrated that taurine acts by increasing the slow component of the closed time kinetics. Whereas open time (τo) and fast closed time (τC1) constants remained unchanged, slow closed time constant (τC2) increased significantly from 140 ± 40 to 610 ± 190 ms (n = 6,P < 0.05).
Estimation of taurine affinity for KATP.
In a previous report, we have shown that taurine, under identical experimental conditions, induced a cell volume increase and that this volume increase was related to the entry of taurine into the cell by way of a sodium-dependent taurine transporter (6). Therefore, known changes in cellular volume induced by taurine perfusion allow an estimation of the amount of taurine flowing into the cell as a function of time. The cellular osmolarity, ΠO, is defined as Equation 2whereN is the mole number of osmotically active solutes, and V is the cell volume. The intracellular osmolarity is assumed to be in equilibrium with the osmolarity of the extracellular solution (300 mosM). Thus, as the cell volume increases in response to the osmolyte influx to keep ΠO constant, the variation of intracellular osmolyte concentration (Δ[Osmolytes]in) with time (see Fig.8 A) is calculated as Equation 3where ΔV(t) is the change in cell volume with time. Then, the variation of intracellular taurine concentration, ΔT, is estimated by Equation 4wheres is the net number of molecules entering the cell per taurine molecule. We assumed that the cotransporter stoichiometry for sodium vs. taurine was 2:1, as demonstrated in rabbit and rat brush-border membranes (30, 31) and in the apical and basolateral membranes of LLC-PK1 and MDCK cell lines (15). Moreover, we assumed that electroneutrality was maintained by the exit of one potassium ion and the net entry of one bicarbonate anion for each taurine transport cycle. Thus, according to these hypotheses, the net number of molecules entering the cell for each taurine molecule is 3. In general, at the time the patch-clamp experiment is performed, tubules have been bathed in taurine-free solution for more than 1 h. If we assumed that the initial intracellular taurine concentration is negligible, Eq. 4 and the time course of taurine-induced cell swelling predict that intracellular taurine concentration progressively increases from 0 to 30 mM during the first 4 min of the experiment. From the curve illustrated in Fig. 8 Aand Eq. 4, the time course ofP o (Fig.7 B) was converted into a relation between P o and the estimated taurine concentration (Fig.8 B). This relation was fitted according to the equation Equation 5whereP Max is the open probability in the absence of taurine, andK 0.5 is the taurine concentration for half inhibition. TheK 0.5 value could be estimated to 8.7 mM, and theP Max to 0.26, which is close to the measured value (0.23 ± 0.06,n = 7) in the absence of external taurine. Please note that this rough calculation would not be much affected by a reasonable underestimation ofs; if the number of molecules accompanying taurine were underestimated by 1, then the calculatedK 0.5 would decrease by only 25%.
Two types of potassium channels. Passive potassium transport across the basolateral membrane of mammalian proximal tubules was thought to be mediated by a unique potassium channel regulated by intracellular ATP (14, 24), intracellular pH (2), and, potentially, by cell volume (2, 7, 14, 16,19, 24). Although other types of potassium channels, such as stretch-activated K channels, have been shown inNecturus proximal cells, no evidence for existence of such channels was provided for the mammalian proximal tubule (20), and the relationship between cell volume and potassium channel remained to be understood. In the present study, single-channel recordings revealed the presence of a second and previously unrecognized type of potassium channel, sK, with lower current amplitude than the KATP channel. As it is extremely difficult to establish a gigaohm seal on the basolateral membrane of the proximal tubule, our knowledge of the basolateral K channel(s) characteristics is limited. This may explain why the presence of this second type of basolateral K channel was generally overlooked but for one mention in a 1993 abstract (1) from our laboratory [see also figure 7 in Parent et al. (19)]. Several lines of evidence allow us to conclude that we are in the presence of two distinct types of potassium channels. First, when the two channels are recorded together, independent gating can be observed. Second, the probability of observing both channel types in the same patch is equal to 0.15. This measured value is consistent with the estimated probability of 0.17, obtained from the probabilities of finding the KATP (0.51) and the sK (0.34) channels individually (0.51 × 0.34 = 0.17), assuming that these two types of K+ channels are independently distributed. Lastly, after excising the membrane patch to reach inside-out configuration, we usually observed a rapid rundown (within 30 s, data not shown) of the KATP channel, whereas low-amplitude activity remained steady all along the experiments, suggesting the existence of two differently regulated channels. Taken together, these data strongly suggest that two distinct potassium channels coexist at the basolateral membrane of PCT cells.
In cell-attached configuration, the sK potassium channel is characterized by a sigmoidal current-voltage curve with a unitary conductance of 12 pS as measured between −40 mV and +20 mV. Potassium channels with low conductance (10–30 pS) have been described at the apical membrane of rabbit thick ascending limb of Henle’s loop (25, 28), at both apical and basolateral membrane of rat cortical collecting tubule (10, 13) and at the apical membrane of inner medullary collecting duct (21). These low-conductance channels all display a high open probability in cell-attached configuration and, in the case of the inner medullary collecting duct, the cortical collecting tubule, the thick ascending limb of Henle’s loop, and the proximal sK channel,P o is not dependent on the membrane potential and is not altered upon membrane excision (8, 10, 21, 27, 28). Finally, the gating mechanisms of the proximal sK channel and the cortical collecting tubule low-conductance channel are both described by one open dwell-time constant and one closed dwell-time constant (13, 26). Although these channels appear to share some common properties, conclusions about the precise identity of proximal sK channel would be premature.
The finding of a second type of potassium channel in the basolateral membrane of PCTs raises the question of its role and its importance with respect to the KATP channel. Concerning the sK channel contribution to the absolute basolateral potassium conductance, its high open probability appears to fully compensate its low conductance. Therefore, the relative conductance for each type of channel strongly depends on the number of active channels. The mean potassium current, I, carried by a given type of channel, is given by Equation 6whereI 0 is the unitary current, P o is the channel open probability,N C is the total number of channels. At the resting membrane potential, corresponding to a micropipette potential of 0 mV, our experimental data allow the estimation of both sK and KATP mean currents, which are equal to I sK = −0.66 × 0.53 × N sK, andI KATP = −1.67 × 0.22 ×N KATP, respectively. To determine theI sK/I KATPratio, we measured aN KATP/N sKratio of 1.43 from patches where the two types of channel coexist. From these data, we estimated that the sK conductance would represent 40% of the total potassium conductance of the PCT basolateral membrane.
Effects of taurine on the basolateral potassium channels. Our results demonstrate that neither the open probability nor the number of channels nor the unitary conductance of the sK channels is modified by taurine. By contrast, the KATP channelP o decreased by 76% (from 0.17 ± 0.06 to 0.04 ± 0.02;n = 7,P < 0.05) within 4 min. Since current recordings are performed in cell-attached configuration and taurine is perfused in the bath solution, theP o decrease must be related to an intracellular effect. In ventricular myocytes, it has been shown that taurine could inhibit ATP-dependent potassium channels in inside-out configuration, suggesting a direct interaction process (12, 22). In agreement with this hypothesis, reversibility of the inhibitory effect could be observed when taurine was removed (12, 22). In comparison with inside-out membrane patches, the use of cell-attached configuration, particularly on a whole proximal tubule, is more respectful of the cell physiology but does not allow an accurate control over the intracellular taurine concentration. It is possible that the removal of taurine under the membrane patch studied would require more than a few minutes. It is interesting to note that the effect of taurine on the basolateral conductance was fully reversible when we were using microelectrodes (4). The possible factors that differ between the two experimental conditions are the temperature (25°C in patch-clamp vs. 38°C with microelectrodes) and the actual geometry of the membrane studied (a microscopic membrane patch vs. the whole basolateral membrane). It is possible that, at 25°C, more than a few minutes are needed to lower taurine concentration under the membrane patch studied.
In proximal tubule cells, as in myocytes, the decrease of open probability is related to the increase of the long closed time. We estimated that taurine concentration should be equal to 8.7 mM for the KATP channel half inhibition. ThisK 0.5 value is quite similar to those found (13.5 mM) for the KATP channels of ventricular myocytes (12). Such aK 0.5 value in the low millimolar range suggests that in the proximal tubule cell in vivo, KATP channels are under the constant inhibitory effect of intracellular taurine. In vitro, intracellular taurine concentrations are expected to reach 30 mM in 4 min upon exposure to 40 mM basolateral taurine. This should inhibit most of the KATPchannels, leaving the basolateral K conductance solely dependent on sK. The observed decrease of the basolateral potassium conductance by 50% (4) is quite consistent with estimation that sK channels are taurine insensitive and responsible for ∼40% of the macroscopic K conductance.
Thus the present results bring a microscopic explanation for the inhibitory effects of taurine on the macroscopic basolateral potassium conductance, previously demonstrated with the intracellular microelectrodes technique (4). Although the mechanisms remain to be determined, our data allow us to definitively conclude that intracellular taurine acts on basolateral ATP-dependent potassium channels by decreasing their open probability. On the other hand, we also demonstrated the presence of a second type of potassium channel that is taurine insensitive and responsible for ∼40% of the basolateral potassium conductance. The estimated sensitivity of KATP to intracellular taurine versus its physiological concentration and the fact that cell swelling modulates taurine permeability indicate that taurine could play a very significant role in maintaining a constant cell volume and membrane potential in the face of varying rates of apical Na-coupled solute transport.
We acknowledge the excellent technical assistance of Mireille Marsolais.
Address for reprint requests and other correspondence: J.-F. Noulin, Université de Montréal, GRTM, C.P. 6128, Succursale Centre-ville, Montreal, Quebec, Canada H3C 3J7 (E-mail:).
This work was supported by the Medical Research Council of Canada Grant MT-10900 to J.-Y. Lapointe and R. Laprade.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 1999 the American Physiological Society