We studied the effects of acidic pH on the currents through voltage-gated K+ channels, Kv1.2, Kv1.4, and their tandem construct (Kv1.4-Kv1.2). Kv1.4 currents were inhibited considerably under acidic pH in a voltage-independent manner, whereas Kv1.2 currents were less inhibited in an apparently voltage-dependent manner. The apparent voltage-dependent block of Kv1.2 currents was mostly ascribed to the shift of activation voltage, which is probably due to surface charge effects of H+ ions. Mutagenesis analysis identified the histidine residue at 508 (H508) in the S5-H5 linker as a molecular determinant of pH sensitivity of Kv1.4. Currents through the tandem channel showed intermediate characteristics between the two parent channels in both sensitivity and voltage dependence of pH effects. Our results suggest that 1) the H508 plays a critical role in determining pH sensitivity of Kv1.4; and 2) the two parent channels, Kv1.2 and Kv1.4, are involved in determining pH sensitivity and apparent voltage dependence of the tandem channel.
Because voltage-gated K+ channels contribute to termination of action potential, inhibition of K+channels could result in prolongation of the action potential duration (APD) leading to abnormal excitability of the cell. Suppression of K+ currents could be a cause of cardiac arrhythmias and increased neuronal excitability. In fact, the voltage-gated K+ channel was first cloned based on the leg-shaking (hyper-reactivity) phenotype ofDrosophila Shaker mutant, which lacks A-type K+ channel (Kamb et al., 1987; Papazian et al., 1987; Pongs et al., 1988). Kv1.2 and Kv1.4 are two of the voltage-gated K+ channels that belong to Shakersubfamily. Both channels have been cloned from the heart and brain (McKinnon, 1989; Stühmer et al., 1989; Tseng-Crank et al., 1990;Roberds and Tamkun, 1991). Although their contribution to native currents has not been completely understood, they are considered to be involved in the generation of transient outward current (Ito1) in the heart and presynaptic A-type K+ current in the neuron. Coexpression of Kv1.4 with Kv1.2 has been reported to generate transient outward currents with similar characteristics to Ito1 (Po et al., 1993). Colocalization of Kv1.2 and Kv1.4 has also been observed in the axons and nerve terminals by immunocytochemical study (Sheng et al., 1993).
It is well recognized that acidosis affects a variety of ion channel activities (Hille, 1968; Shrager, 1974; Kaibara and Kameyama, 1988; Ito et al., 1992). One of the most extensively studied cases is cardiac ischemia, which causes marked acidosis (Garlick et al., 1979; Ichihara et al., 1984; Yan and Kleber, 1992). In cardiac muscle, acidosis reduces the height of action potential plateau and changes APD. However, the changes in the duration are variable (Chesnais et al., 1975; Fry and Poole-Wilson, 1981; Kurachi, 1982). Shortening of APD probably reflects the inhibition of Ca2+currents, and prolongation of APD is probably due to the inhibition of K+ currents. Prolongation of cardiac APD by acidic pH can be arrhythmogenic (Orchard and Cingolani, 1994). The present study was aimed to investigate 1) how acidic pH affects the currents through Kv1.2 or Kv1.4 homomeric channel and Kv1.4-Kv1.2 heteromeric channel, and 2) which amino acid residue(s) is responsible for the sensitivity if acidic pH affects the currents.
Materials and Methods
In Vitro Mutagenesis.
A scheme of Kv1.4-Kv1.2 tandem construct is shown in Fig. 1A and amino acid residues in the S5-S6 linker region of Kv1.2 and Kv1.4 are shown in Fig. 1B. The method of construction of the tandem channel is reported in a previous article (Nunoki et al., 1994). A chimera channel was constructed by replacing the first half of the S5-S6 linker of Kv1.4 with that of Kv1.2. Kv1.4 cDNA was digested with SphI at nucleotide 1428 and BstEII at nucleotide 1560 to remove the region containing S5 and the first 21 amino acid residues of the S5-S6 linker. The corresponding region of Kv1.2 was generated by polymerase chain reaction (PCR) using a sense primer having anSphI site at the 5′ end and an antisense primer having aBstEII site at the 5′ end. The PCR product was digested withSphI and BstEII, and ligated into the chimera. The resulting chimera has Kv1.2 sequences only in the S5-H5 linker, since the amino acid sequences of S5 and most of H5 in Kv1.2 and Kv1.4 are identical. Five point mutations at the residues 505 through 508 and 510 in the S5-H5 linker of Kv1.4 were generated by overlap extension PCR. Each residue of Kv1.4 was substituted with the corresponding residue of Kv1.2. In the first round PCR, a pair of complementary primers containing the desired mutation and a sense primer upstream of the mutation and an antisense primer downstream of it were used to generate a set of PCR products. Each set of the PCR products was then denatured and annealed for the second round PCR, which was carried out with the same sense and antisense primers as the first round. The final PCR products generated as described above were digested withSphI and BstEII and ligated into the mutants. For all the mutations, the nucleotide sequences of the fragments generated by PCR were verified by the dideoxy chain termination method using an A.L.F. DNA sequencer II (Pharmacia Biotech Inc., Uppsala, Sweden).
Expression and Current Recording.
Expression of the K+ channels was carried out as described previously (Ishii et al., 1992). The pBluescript II vectors containing the wild type and mutant K+ channels were linearized with EcoRI, and capped cRNAs were prepared from these templates with T7 RNA polymerase (Stratagene, La Jolla, CA). Transcribed RNAs were dissolved in water at a final concentration of 0.2 μg/μl for oocyte injection. The integrity of the cRNAs was checked by running the samples on formaldehyde-containing agarose gels. Defolliculated Xenopus oocytes (stage V-VI) were injected with 40 to 50 nl (8–10 ng) of cRNA. The injected oocytes were incubated in Barth's medium supplemented with penicillin G (71.5 units/ml) and streptomycin (35.9 μg/ml) at 18°C for 2 to 5 days before electrophysiological measurements. Oocytes expressing K+ channels were continuously perfused with the bath solutions at 1 ml/min. Whole-cell currents were first measured in the control condition (pH 7.5) and then in the solution with different pH to evaluate its effects. Depolarizing pulses were applied to the oocytes from a holding potential of −80 mV at 30-s interval. The K+ currents were recorded by a conventional two-microelectrode voltage-clamp method with 3 M KCl-filled electrodes. The bath recording solution was composed of 96 mM NaCl, 2 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, and 5 mM HEPES (pH 8.5 and 7.5), 5 mM ACES (pH 6.5), 5 mM MES (pH 5.5) or 5 mM acetate (pH 4.5). The pH of each solution was freshly adjusted with NaOH before each experiment. All electrophysiological measurements were carried out at room temperature (21 ± 1°C). Current records were low pass-filtered at 3 kHz.
For construction of normalized conductance-voltage relationship curves, conductance (G) was determined by the formulaG = I/(V m− V rev), where I is the peak current at voltage V m, assuming a reversal potential (V rev) of −90 mV, and data were fit with a Boltzmann function;G/G max = 1/(1 + exp(−(V m −V a)/a n)), where G max is the maximal conductance,V m is the membrane voltage of depolarization pulse, V a is the voltage for half-activation, and a n is the slope factor. For construction of the corrected current-voltage relationships, the voltage shifts were offset by subtracting the difference of V a value (between at pH 7.5 and others) from the test potentials and the currents were plotted against the corrected potentials. All data are expressed as the mean ± S.E. The statistical significance was evaluated by Student's paired or unpaired t test. A P value smaller than 0.05 was considered to be significant.
Effects of External Acidic pH on Wild-Type Channels.
Figure2, A and B, shows Kv1.2 and Kv1.4 currents elicited by voltage step to +40 mV from a holding potential of −80 mV under different external pH. Relative currents to those at pH 7.5 are plotted against external pH in Fig. 2, D and E. The extent of inhibition by external acidic pH was markedly different between Kv1.4 and Kv1.2 currents. The two channels also showed apparent differences in voltage-dependence of the inhibition by acidic pH. The effects of acidic pH on Kv1.4 currents were not voltage-dependent, whereas the effects on Kv1.2 currents showed apparent voltage dependence; the more positive the voltage, the less the inhibition (Fig. 2, D and E). Accordingly, the difference in effects of acidic pH between Kv1.2 and Kv1.4 becomes less marked when evaluated at lower voltage. At +40 mV, lowering external pH from 7.5 to 5.5 reduced Kv1.4 and Kv1.2 currents by 88.8 ± 1.1% (n = 10) and by 26.3 ± 2.6% (n = 11), respectively. However, when the effects of pH 5.5 were measured at 0 mV, Kv1.4 current was decreased by 89.2 ± 1.5% (n = 10), which is similar value at +40 mV, whereas Kv1.2 current was decreased by 66.1 ± 5.5% (n = 11). In Fig. 3, A and B, activation curves for Kv1.2 and Kv1.4 are shown. There are marked differences in the effects of acidic pH between the two channels. Voltage dependence of activation for Kv1.2 channel was shifted to positive membrane potential to a large extent by lowering external pH, whereas the shift of the curve for Kv1.4 channel was small. In Kv1.2, this voltage shift of activation curve seems to account for most of the observed inhibition of the currents, since current-voltage relationships under different external pH down to 5.5 became almost identical if the voltage shifts were simply offset (Fig.3, D and G). In contrast, there was not a marked difference between the original and the corrected current-voltage relationships for Kv1.4 (Fig. 3, E and H). Effects of acidic pH on the time course of N-type inactivation of Kv1.4 currents were also evaluated by fitting the current decay with a single exponential function. When the currents elicited by depolarizing pulse to +20 mV were subjected to analysis, time constants of 55.4 ± 5.7 ms (pH 7.5) (n = 14) and 47.5 ± 4.3 ms (pH 6.5) (n = 14) were obtained. Acidic pH tended to accelerate the decay of Kv1.4 currents, but there was not significant difference between the two values (P > 0.1).
Chimera Channel Shows Reduced Sensitivity to Acidic pH.
It is well established that the S5-S6 linker contains pore-forming region and also contributes to the formation of external vestibule of the channel pore. As shown in Fig. 1B, there are only nine amino acid differences in the S5-S6 linker between Kv1.2 and Kv1.4, but a cluster of six amino acids, including five different residues is found in the first half of the S5-S6 linker. Therefore, we constructed a chimera in which the first half of the S5-S6 linker of Kv1.4 was replaced with that of Kv1.2 to examine whether the region is responsible for high sensitivity of Kv1.4 to acidic pH. Representative traces and relative currents plotted against external pH are shown in Fig. 4A. The sensitivity to acidic pH of the chimera channel was far less than Kv1.4. Lowering external pH from 7.5 to 5.5 decreased currents through the chimera channel by 43.6 ± 8.5% at +40 mV and by 69.7 ± 7.1% at 0 mV (n = 7). Similar to Kv1.2, inhibition of the chimera channel by acidic pH was apparently voltage dependent (Fig.4A). Current-voltage relationships under different pH are shown in Fig.4B and those corrected by offsetting the voltage shifts of activation (Fig. 4C) are in Fig. 4D. The corrected current-voltage relationships indicate that the current inhibition observed at pH 5.5 is largely due to the voltage shift but that at pH 4.5 has a considerable component independent of the voltage shifts.
Histidine 508 Is Responsible for High Sensitivity to Acidic pH.
Since replacement of the first half of the S5-S6 linker caused a loss of high sensitivity of Kv1.4 to acidic pH, we then introduced point mutations at the five unmatched residues located in the region. Substitution of the each amino acid of Kv1.4 with the corresponding residue of Kv1.2 generated P505R, T506D, T507S, H508Q, and Q510P. Among them, P505R and Q510P expressed in Xenopus oocytes did not produce currents large enough to evaluate the effects of acidic pH, but the other three mutants produced currents suitable for the evaluation. Effects of acidic pH on the currents through T506D and T507S were very similar to Kv1.4 (Fig. 5). Inhibition of both channels was not voltage dependent and their currents were reduced by about 35% at pH 6.5 and about 90% at pH 5.5. In contrast to T506D and T507S, current of H508Q was far more resistant to acidic pH. Although H508Q was not subjected to external pH 4.5, its response to acidic pH down to 5.5 was very similar to the chimera channel. Lowering external pH from 7.5 to 5.5 decreased currents of H508Q by 44.0 ± 2.6% at +40 mV and by 65.8 ± 4.0% at 0 mV (n = 8) (Fig. 6A). Inhibition of currents of H508Q is apparently voltage dependent (Fig. 6A). Current-voltage relationships and those corrected by offsetting the voltage shifts of activation (Fig. 6C) are also similar to those of the chimera channel (Fig. 6, B and D).
Tandem Channel Shows Intermediate Sensitivity to Acidic pH.
Since Kv1.2 and Kv1.4 currents showed a marked difference in the response to acidic pH, we examined the effects of acidic pH on the currents through Kv1.4-Kv1.2 tandem channel, which is thought to be composed of Kv1.4 and Kv1.2 with 1:1 stoichiometry. When the tandem channel was expressed in Xenopus oocytes, depolarizing pulses produced transient outward currents as previously reported (Nunoki et al., 1994). Actual traces of the tandem channel at +40 mV are shown in Fig. 2C and relative currents against external pH are shown in Fig. 2F. Lowering external pH from 7.5 to 5.5 reduced the current by 44.0 ± 3.1% at +40 mV and by 67.4 ± 4.4% at 0 mV (n = 6). Overall characteristics of the response to acidic pH of the tandem channel were intermediate between Kv1.2 and Kv1.4. Currents through the tandem channel were inhibited in an apparently voltage-dependent manner at pH 5.5 similar to Kv1.2, but the voltage dependence of the suppression at pH 4.5 was not prominent (Fig. 2F). Current-voltage relationships and those corrected for the voltage shifts of activation (Fig. 3C) are shown in Fig. 3, F and I. Although each curve at acidic pH moved to the left when the voltage was offset, there clearly exists considerable current reduction independent of the voltage shifts.
The present study demonstrates marked differences in the responses to acidic pH between Kv1.2 and Kv1.4. The differences include not only the sensitivity to pH but also the apparent voltage-dependence of the pH effect. Inhibition of Kv1.4 currents by acidic pH was remarkable and voltage-independent, whereas the inhibition of Kv1.2 currents was small and apparently voltage-dependent. However, when current-voltage relationships for Kv1.2 were corrected by simply offsetting the voltage shift of activation curves, the relationships under various external pH (from 7.5 to 5.5) overlapped each other. This result means that the apparent voltage-dependent block of Kv1.2 current is mainly due to the shift of activation voltage, which probably results from surface charge effects of H+ ions (Green and Andersen, 1991;Hille, 1992). In contrast, it seems that the inhibition of Kv1.4 currents by acidic pH is mostly independent of surface charge effects. Thus, inhibition of the two channels by acidic pH differs qualitatively.
Replacement of the S5-H5 linker of Kv1.4 with that of Kv1.2 markedly reduced the sensitivity of Kv1.4 to acidic pH and also conferred apparent voltage-dependence. Further mutagenesis experiments revealed that replacement of a histidine residue (H508) in the S5-H5 linker is responsible for the changes in the chimera channel. The calculated pK a value for inhibition of Kv1.4 currents by H+ ions is 6.3, a similar value to the pK a of histidine (6.5), which supports the importance of the histidine residue in the inhibition of Kv1.4 by acidic pH. Since the apparent voltage dependence is due to the voltage shift for activation, which is probably ascribed to surface charge effects of H+ ions, the results with the chimera and H508Q indicate that the histidine residue rather masks the surface charge effects of H+ ions. Although it seems strange that removal of a titratable residue histidine strengthens surface charge effects of H+ ions instead of diminishing it, another article shows no role of a histidine residue in surface charge effects of H+ ions (Steidl and Yool, 1999). They demonstrated that a histidine residue (H452) in the S5-H5 linker is an external pH sensor in Kv1.5 but steady-state activation curves for wild type and Kv1.5H452Q were equally shifted by acidic pH. The H452 of Kv1.5 is an equivalent residue to H508 of Kv1.4, which we identified as a determinant of pH sensitivity. Taken together, from our result on Kv1.4 and their result on Kv1.5, it could be concluded that a histidine residue at the site does not participate in the positive shift of activation curve by H+ions. A major point of the study on Kv1.5 concerns C-type inactivation. The study on Kv1.5 demonstrated that the histidine residue plays a role in C-type inactivation of Kv1.5 currents and acidic pH augments the C-type inactivation to result in the current reduction of Kv1.5. In our results on Kv1.4 whose prominent character is a rapid N-type inactivation, acidic pH seems not to have significant effects on the time course of N-type inactivation. In addition, changes in recovery from inactivation are probably not involved in the observed effects of acidic pH on Kv1.4 since the currents were elicited at 30-s intervals, which seems to be long enough to recover from inactivation.
We also investigated the effect of acidic pH on Kv1.4-Kv1.2 tandem channel, which is thought to be formed as heterotetramer of Kv1.2 and Kv1.4. Interestingly, its sensitivity to pH and apparent voltage-dependence were intermediate between the two parent channels. These results in turn support the heteromultimeric nature of the tandem channel. Tetrameric nature of K+ channels tempted us to speculate that Kv1.4 homotetramer has four H+ ion binding sites to produce its high sensitivity to pH, whereas Kv1.4-Kv1.2 heterotetramer has two binding sites, which might account for the reduced sensitivity. However, the calculated Hill coefficients for their inhibition curves were 1.2 (Kv1.4) and 0.9 (Kv1.4-Kv1.2), which indicate that a single H+ ion binding site is probably sufficient for inhibition of both channels. Although it is not clear by what molecular mechanisms the tandem channel exhibits the intermediate characteristics in the response to pH, at least it must acquire apparent voltage dependence from a parent channel Kv1.2. Thinking of the capability of Kv channels being formed as heterotetramer in heterologous expression system and colocalization of different Kv channel subunits in vivo (Christie et al., 1990; Isacoff et al., 1990; McCormack et al., 1990;Ruppersberg et al., 1990; Po et al., 1993; Sheng et al., 1993; Wang et al., 1993), it is noteworthy that heteromultimeric channel has intermediate properties between the component channels.
There are several reports concerning the identification of amino acid residues responsible for pH sensitivity of the ion channels (Busch et al., 1991; Coulter et al., 1995; Fakler et al., 1996; Steidl and Yool, 1999). The study on HIR (Kir2.3), an inward rectifier K+ channel, showed that a histidine at 117 is a molecular determinant of pH sensitivity of the channel (Coulter et al., 1995). It was proposed that the presence of a positive or neutral residue at position 117 of Kir2.3 favors a channel conformation that allows a different titratable group to influence pore properties. Interestingly, the histidine 117 of Kir2.3 resides at the exactly same position as H508 of Kv1.4 and H452 of Kv1.5; they are all located at 19 residues before a selectivity filter GYG (Heginbotham et al., 1992). Since inward rectifier K+ channels and voltage-gated K+ channels are distantly related, these findings might imply a fundamental importance of a histidine residue at the position in pH sensitivity of K+channels.
- Received July 13, 2000.
- Accepted October 10, 2000.
Send reprint requests to: Dr. Kuniaki Ishii, Department of Pharmacology, Yamagata University School of Medicine, 2-2-2 Iida-nishi, Yamagata 990-9585, Japan. E-mail:
This study was supported by Grants-in Aid for Scientific Research (09557210 and 10470021) from the Ministry of Education, Science, Sports and Culture, Japan.
- action potential duration
- polymerase chain reaction
- N-(2-acetamido)-2-aminoethanesulfonic acid
- 2-(N-morpholino)ethanesulfonic acid
- The American Society for Pharmacology and Experimental Therapeutics