J Biol Chem, Vol. 275, Issue 11, 7811-7817, March 17, 2000

Involvement of Histidine Residues in Proton Sensing of ROMK1 Channel^*

Sengthong Chanchevalap, Zhenjiang Yang, Ningren Cui, Zhiqiang Qu, Guoyun Zhu, Congxiao Liu, Lande R. Giwa, Latifat Abdulkadir, and Chun Jiang

From the Department of Biology, Georgia State University, Atlanta, Georgia 30302-4010

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ROMK channels are inhibited by intracellular acidification. This pH sensitivity is related to several amino acid residues in the channel proteins such as Lys-61, Thr-51, and His-206 (in ROMK2). Unlike all other amino acids, histidine is titratable at pH 6-7 carrying a positive charge below pH 6. To test the hypothesis that certain histidine residues are engaged in CO₂ and pH sensing of ROMK1, we performed experiments by systematic mutations of all histidine residues in the channel using the site-directed mutagenesis. There are two histidine residues in the N terminus. Mutations of His-23, His-31, or both together did not affect channel sensitivity to CO₂. Six histidine residues are located in the C terminus. His-225, His-274, His-342, and His-354 were critical in CO₂ and pH sensing. Mutation of either of them reduced CO₂ and pH sensitivities by 20-50% and ~0.2 pH units, respectively. Simultaneous mutations of all of them eliminated the CO₂ sensitivity and caused this mutant channel to respond to only extremely acidic pH. Similar mutations of His-280 had no effect. The role of His-270 in CO₂ and pH sensing is unclear, because substitutions of this residue with either a neutral, negative, or positive amino acid did not produce any functional channel. These results therefore indicate that histidine residues contribute to the sensitivity of the ROMK1 channel to hypercapnia and intracellular acidosis.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ROMK channels (Kir1.1 and Kir1.2), members in the inward rectifier K⁺ channel family, were first cloned in the kidney and have later been found in several organs including the central nervous system (1, 2). These K⁺ channels are believed to control K⁺ secretion in the renal tubular cells and membrane potential in excitable cells (3, 4). ROMK channels have a relatively weak inward rectification allowing significant K⁺ currents in the outward direction and are subject to extensive modulations by second messengers, protein kinases, phospholipids, ATP, and other nucleotides (5). Another important modulator of the ROMK channels is proton. A decrease in intracellular pH strongly inhibits these channels (6-11).

Proton sensing in ROMK channels requires multiple sites or residues in the channel protein to interact with protons. A lysine residue (Lys-80 in ROMK1 or Lys-61 in ROMK2) in the N terminus of channel proteins plays a critical part in the channel sensitivity to intracellular pH. Mutation of this residue to methionine greatly reduces the pH sensitivity (8, 10). However, lysine is not titratable at a physiological pH range. Thus, how this lysine residue works in pH sensing is not fully understood. A nontitratable threonine residue at position 51 of ROMK2 is also involved in pH sensing (8). Mutation of this threonine to a negatively charged residue enhances pH sensitivity, whereas switching it to a positive amino acid decreases the pH sensitivity (8). Another residue related to pH sensing is histidine 206. Mutation of this residue to glycine enhances channel sensitivity to pH (10). Supporting multiple interaction sites in proton sensing are also recent studies showing that movements of protein domains in N and C termini occur during the recovery of ROMK1 channels from the inhibition by acidic pH (11). Two cysteine residues (Cys-49 and Cys-308) are critical players in the recovery, although they do not affect the channel inhibition per se (11). Although these observations clearly demonstrate that multiple residues are involved in proton sensing in ROMK proteins, which one of them is the proton-binding site remains to be known.

Unlike all other amino acids, histidine can be protonated at pH 6-7 (side chain pK 6.04) resulting in a positive charge at this residue with acidic pH. This may in turn cause alterations in protein conformation and channel activity. Indeed, there is experimental evidence indicating that histidine is a central player in the pH-dependent modulation of a number of ion channels, including KST1, connexin 43, GABA receptor, Kv2.1, and porin (12-16). In ROMK1, there is a total of eight histidine residues in the channel protein with two of them in the N terminus and the rest in the C terminus. To test the hypothesis that certain histidine residues are involved in CO₂ and pH sensing of the ROMK1 channel, we performed experiments by systemic mutations of these histidine residues in the ROMK1 protein and studied channel response to CO₂ and pH. Our results indicate that His-225, His-274, His-342, and His-354 play a role in CO₂ and pH sensing of the ROMK1 channel.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Frogs, Xenopus laevis, were anesthetized by bathing in 0.3% 3-aminobenzoic acid ethyl ester. A few lobes of ovaries were removed after a small abdominal incision (~5 mm). Xenopus oocytes were treated with 2 mg/ml of collagenase (Type I, Sigma) in OR2 solution (82 mM NaCl, 2 mM KCl, 1 mM MgCl₂, and 5 mM HEPES, pH 7.4) for 90 min at room temperature. After washing, the oocytes were then incubated at 18 °C in ND-96 solution containing 96 mM NaCl, 2 mM KCl, 1 mM MgCl₂, 1.8 mM CaCl₂, 5 mM HEPES, and 2.5 mM sodium pyruvate with 100 mg/l geneticin added (pH 7.4).

A vector for eukaryotic expression (pcDNA3.1, Invitrogen, Carlsbad, CA) was used to express ROMK1 and Kir2.1 channels (1, 17). Site-specific mutations were carried out using a site-directed mutagenesis kit (Quickchange, Stratagene, La Jolla, CA) based on the Pfu polymerase chain reaction (18). The correct sequences in both subcloning and site-directed mutagenesis were confirmed with DNA sequencing. Systemic mutations were performed on each histidine residues in the ROMK1, in which these residues were substituted in order with a neutral (asparagine, glutamine, or alanine), acidic (aspartic acid or glutamic acid), and alkaline amino acids (lysine). After expressions of these mutant channels were confirmed in voltage clamp, their sensitivity to CO₂ was tested.

Whole-cell currents were studied on the oocytes 2-4 days after a cDNA injection (20-50 ng in 50 nl of H₂O). Two-electrode voltage clamp was performed using an amplifier (Geneclamp 500, Axon Instruments Inc., Foster City, CA) at room temperature (~25 °C). The extracellular solution contained: 96 mM KCl, 3 mM MgCl₂, and 5 mM HEPES (pH 7.4). Cells were impaled using electrodes filled with 3 M KCl. Currents were recorded using electrodes with resistance of 0.4-0.6 M, low pass-filtered (Bessel, 4-pole filter, 3 dB at 5 kHz), digitized at 5 kHz (12-bit resolution), and acquired using pClamp 6.0.3 (Axon Instruments) (18, 19). Oocytes were accepted for further experiments only if they expressed large inward rectifying currents (>3 µA) with clear superimposed current traces in response to depolarizing command potentials.

In a semi-closed recording chamber (BSC-HT, Medical System, Greenvale, NY), the oocytes were placed on a supporting nylon mesh, so that the perfusion solution bathed both the top and bottom surface of the oocytes (18, 19). The perfusate and the superfusion gas entered the chamber from the inlet at one end and flowed out at the other end. There was a 3 × 15-mm gap on the top cover of this chamber, which served as the gas outlet and an access to the oocytes for recording microelectrodes. The perfusate contained 96 mM KCl, 3 mM MgCl₂, 5 mM HEPES (pH 7.4). At baseline, the chamber was ventilated with atmospheric air. Exposure of the oocytes to CO₂ was carried out by switching the superfusion air to a gas mixture containing CO₂ balanced with 21% O₂ and N₂. The high solubility of CO₂ resulted in an evident change in intra- or extracellular acidification as fast as 10 s in these oocytes. Current responses to CO₂ were studied before, during (4-5 min), and after CO₂ exposure. Extracellular acidification was done using a buffer containing 10 mM PIPES¹ or HEPES. The pH of these solutions was titrated to the desired levels immediately before experiments. In intracellular acidification experiments, the bath solution contained 76 mM KCl, 3 mM MgCl₂, and 20 mM KHCO₃ (pH titrated to 9.5). This solution was then bubbled with 15% CO₂ for at least 30 min before recordings. The CO₂ bubbling decreased the pH level in this perfusion solution to about 7.35.

Patch clamp experiments were performed at room temperature (~25 °C) as described previously (18-20). In brief, fire-polished patch pipettes (0.5-2 M) were made from 1.2-mm borosilicate capillary glass (Sutter P-94/PC puller). Macroscopic currents were recorded from giant inside-out patches using pipettes (0.5-1.0 M). Current records were low pass-filtered (2000 Hz, Bessel, 4-pole filter, 3 dB), digitized (10 kHz, 12 bit resolution), and stored on computer disc for later analysis (PCLAMP 6, Axon Instruments). Junction potentials between bath and pipette solutions were appropriately nulled before seal formation. Vitelline membranes of the oocytes were mechanically removed after exposing to hypertonic solution (400 mosM) for 5 min. The stripped oocytes were placed in a Petri dish containing regular bath solution. The solution applied to the bath and recording pipettes contained: 40 mM KCl, 75 mM potassium gluconate, 5 mM potassium fluoride, 0.1 mM sodium vanadate, 10 mM potassium pyrophosphate, 1 mM EGTA, 0.2 mM ADP, 10 mM PIPES, 10 mM glucose, and 0.1 mM spermine (FVPP solution, pH 7.4). This solution was chosen after several others had been tested regarding channel rundown in excised patches. In a control experiment, we found that macroscopic currents recorded from giant inside-out patches were very well maintained showing less than 10% reduction over a 20-min period of recordings in such a bath solution. Data were further filtered (1000 Hz) with a Gaussian filter. Current amplitude was measured at 100 mV after averaging eight consecutive traces. Percentage changes in current amplitude with CO₂ and intracellular acidification were presented in these experiments.

A parallel perfusion system was used to administer agents to patches at a rate of ~1 ml/min with no dead space (18-20). Low pH was produced by exposing patches to the same internal FVPP solution that had been titrated to various pH levels using gluconic acid, N-methyl-D-glucamine, or KOH. Osmotic pressure was adjusted to ~300 mosM after pH titration. No significant difference in single channel conductance was seen between N-methyl-D-glucamine and KOH titrations (18, 19). PIPES buffer was used because of its appropriate buffering range and membrane impermeability. Data are presented as mean ± S.E., and differences in mean were tested with the Student's t test and accepted as significant if p  0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Inhibition of ROMK1 Currents by High CO₂ and Low pH-- Whole-cell currents were studied in the two-electrode voltage clamp mode. Depolarizing and hyperpolarizing command pulses were given to the cell (in a range from 160 to 140 mV with 20 mV increments) at a holding potential of 0 mV. Under such a condition, evident inward rectifying currents were observed 2-4 days after an injection of ROMK1 cDNA. At a hyperpolarizing potential (160 mV), the amplitude of these currents averaged 23.4 ± 3.4 µA (mean ± S.E., n = 16). These currents were sensitive to micromolar concentrations of Ba²⁺ and Cs⁺ with the IC₅₀ (50% of the maximum inhibition) for Ba²⁺ 110 µM.

Exposure of these oocytes to 15% CO₂ for 4-5 min produced a strong inhibition of ROMK1 currents, when oocytes were positioned ~100-200 µm beneath the surface of the perfusion solution in a semiclosed recording chamber. Evident inhibition of these currents started within 30 s of the CO₂ exposure. The maximum effect was reached in about 2-3 min during the exposure and maintained throughout the rest of period (Fig. 1). The CO₂ response was reversible. A complete recovery was seen in most of oocytes studied with 1-3 exposures. The CO₂-induced inhibition of the ROMK1 currents showed a clear concentration dependence with 5, 10, and 15% CO₂. At 15%, CO₂ caused an inhibition of the ROMK1 currents by 70.5 ± 2.8% (n = 9). In sharp contrast, another Kir channel, Kir2.1, was inhibited by 2.8 ± 1.1% (n = 5), a level that was similar to the response of endogenous currents in water-injected oocytes (1.6 ± 8.2%, n = 5, p > 0.05), indicating that CO₂ selectively inhibits specific K⁺ currents.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Reversible inhibition of ROMK1 currents by high CO2 and acidic pH. Whole-cell currents were recorded from a Xenopus oocyte using an extracellular solution containing 96 mM K+ and the two-electrode voltage clamp 2 days after ROMK1 cDNA injection. Membrane potential was held at 0 mV, and a series of command pulse potentials from 160 to 140 mV with a 20-mV increment were applied to the cell. When the oocyte was exposed to 15% CO2, inward rectifying currents were markedly and reversibly inhibited. a-e, current records before, during, and after CO2 exposure. f, time profile of the current amplitude with CO2 shows that the currents start to decrease almost immediately after the perfusion gas was switched to CO2, reach the maximum at ~3 min, and maintain at this level until the overflowing gas was switched back to room air. Note that arrows from left to right point where a, b, c, d, and e are sequentially taken. g-i, ROMK1 currents were reversibly inhibited by selectively reducing pHi without changing pHo. Oocytes were bathed with a solution that had been bubbled with 15% CO2 for 35 min. The pH level in this perfusion solution was reduced to 7.34 (from pH 9.5 before bubbling). Although there is not marked change in pHo, pHi was reduced to 6.59 from pH 7.20. This drop in intracellular pH caused a marked inhibition in ROMK1 currents (h).

Intracellular acidification to pH_i 6.6, a level (pH 6.6 ± 0.1, n = 5) that was measured using ion-selective microelectrodes during 15% CO₂ exposure (21), produced a drop in ROMK1 currents by 56.7 ± 2.8% (n = 4) without significant change in pH_o (Fig. 1, g-i). Acidification of the perfusate to pH 6.2, the same level of pH_o as seen during 15% CO₂ exposure, had no effect on the ROMK1 currents (3.3 ± 3.1%, n = 5, p > 0.05). Thus, these results indicate that a drop in intra- but not extracellular pH is the primary cause for the inhibition of ROMK1 current by CO₂.

The inhibition of ROMK1 by CO₂ or intracellular acidification can be a direct effect of protons on the channel protein or an indirect effect through changes in concentrations of intracellular second messengers, protein kinases, phosphatases, and other cytosol-soluble factors. To delineate the mechanisms underlying the channel modulation by protons, we performed experiments using cell-free excised patches. In the inside-out patch configuration, ROMK1 currents were strongly inhibited when the cytosolic surface of the plasma membrane was exposed to low pH solutions. This inhibition was fast, reversible, and concentration-dependent (Fig. 2). A similar reduction in pH_i did not affect the Kir2.1 currents. Because cytosol-soluble factors were vastly diluted or washed out under such an experimental condition, the inhibition of ROMK1 channels is unlikely to be mediated by second messengers and other cytosol-soluble factors. Moreover, because our intracellular solutions contained chemicals that were unfavorable to protein dephosphorylation (see "Materials and Methods"), the modulation of ROMK1 channel activity by pH may not be related to the fast turnover of protein phosphorylation and dephosphorylation.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Inhibition of ROMK1 currents by decreases in intracellular pH (pHi) in cell-free excised patch. A, ROMK1 currents were recorded from an inside-out patch with symmetric K+ concentration on both sides of the plasma membrane. Ramp command potentials from 100 to 100 mV were applied to the patch with a holding potential of 0 mV. Exposure of the internal membrane to solutions with various pH levels produced a graded inhibition of inward rectifying currents. Note that eight superimposed traces are shown in each panel. B, the amplitude of these currents can be expressed as a function of pHi using the Hill equation: y = 1/{1 + (pK/x)h}, where y is the normalized current amplitude, pK the midpoint pH value for channel inhibition, x is pHi, and h is the Hill coefficient. The pK and h here are pH 6.73 and 3.6, respectively. Change in pHi in the same range had a very little effect on Kir2.1 currents. Data are presented as means ± S.E. (n = 4).

Histidine Residues in the N Terminus-- Our results have suggested that ROMK1 is inhibited during hypercapnia, and this is likely to be mediated by a decrease in intracellular pH. The proton-sensing sites thus should be located on the cytosolic side of the plasma membranes. Because both N and C termini of ROMK1 are intracellular, we examined histidine residues in the N- and C-terminal regions by systematical mutations of these residues using the site-directed mutagenesis. In these experiments, one histidine residue was substituted with a neutral, acidic or alkaline amino acid sequentially (nonpolar residues were not extensively attempted, because they may change the hydrophobicity at local areas). Then, channel sensitivity to CO₂ was examined. If any of these mutants was not expressed after examining 30-40 oocytes injected, another injection of the same mutant from a different colony was followed. If there was still no expression, we believed that the mutation was too severe to produce functional channels. There are two histidine residues in the N terminus. These histidines were replaced with asparagine, aspartate, or lysine that are either nontitratable or titratable at extremely low or high pH. When the sensitivity of these mutant channels to CO₂ was examined, we found that none of these mutations had an effect on the channel sensitivity to CO₂ (Fig. 3). Neither did a simultaneous replacement of both of these histidines (H23D/H31D), suggesting that His-23 and His-31 are not the pH sensor in ROMK1.

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Potentially titratable histidines of the N terminus in CO2 sensitivity. Two histidine residues are found in the N terminus of ROMK1. They were replaced with aspartic acid, asparagine, or lysine. CO2 sensitivity was then examined using a two-electrode voltage clamp. Simultaneous mutations of both of them or replacements of individual ones had no effect on the sensitivity of these mutant channels to CO2, indicating that these histidine residues are not involved in pH sensing.

Histidine Residues in the C Terminus-- There is a total of six histidine residues in the C terminus of the ROMK1 channel. All of these histidine residues were modified systematically, and sensitivities of these mutant channels to CO₂ and pH_i were then studied. His-274 is a nonconserved residue. Because an acidic residue is found at the same position in Kir2.1 and Kir2.3, we created a mutation by substituting His-274 with an aspartic acid (H274D), which expressed active channel with current amplitude of ~13 µA (Table I). The H274D mutation significantly reduced channel sensitivity to CO₂ by ~15% (Fig. 4). We also mutated the His-274 to an alkaline lysine and neutral polar asparagine. The H274K mutation reduced the CO₂ sensitivity by ~30%, but the H274N did not yield a functional channel. His-342 and His-354 are two nonconserved residues. Mutation of the His-342 to a neutral polar residue (H342Q) markedly reduced the CO₂ sensitivity of the channel (Fig. 4). Similar effects were seen with mutations to either basic or acidic residues (H342K and H342D). Mutations of His-354 to a neutral, positive, or negative residue (H354N, H354K, or H354D) also decreased the channel sensitivity to CO₂ with the H354N being more obvious (Fig. 4). Although the His-225 is conserved in the Kir family, its mutations produced function channels. H225K and H225D mutations reduced the channel sensitivity to CO₂ by 30-40% (Figs. 4 and 5). The H225N mutation was special among all histidine mutants of the ROMK1. This mutation significantly enhanced channel sensitivity to CO₂ and generally eliminated the inward rectification (Figs. 4 and 5). Interestingly, the inward rectification that was not seen at the baseline (Fig. 5a) appeared during hypercapnia (Fig. 5, b and c), and the emergence of the rectification was much faster than the inhibition of the inward rectifying currents (Fig. 5b). Washout made the H225N currents return a to nonrectifying pattern (Fig. 5, d and e). This phenomenon was not observed in H225D and H225K (Fig. 5, f-j), suggesting that a charged status at this residue contributes to the rectification mechanisms of the ROMK1. When several of these histidines were mutated together, greater effects were observed. The CO₂ sensitivity was reduced by ~70% in the H342Q/H354N mutant and by ~85% in the H225K/H342Q/H354N mutant. When all these four histidines were mutated (H225K/H274K/H342Q/H354N), the channel became CO₂-insensitive (Figs. 4 and 6). Its response to 15% CO₂ was similar to that of Kir2.1 (p > 0.05, n = 6).

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Potentially titratable histidines of the C terminus in channel sensitivity to CO2. There are six histidine residues in the C terminus of ROMK1. They were replaced with asparagine, glutamine, aspartic acid, or lysine. CO2 sensitivity was significantly reduced in His-225, His-274, His-342, and His-354 mutations. Simultaneous mutations of two or more of them had a greater effect (H342Q/H354N and H225K/H342Q/H354N). His-280 is not involved in CO2 sensing, because mutations of this residue to either neutral, positive, or negative residue did not affect the CO2 sensitivity. These effects are specific, as combined mutations of some of these CO2-sensitive histidines with His-280 did not increase the CO2 sensitivity (H280K/H342Q). The decrease in CO2 sensitivity is statistically significant in mutants with an asterisk in front of them. Data are presented as mean ± S.E.

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Effects CO2 on His-225 mutations. a-e, whole-cell currents were recorded from an oocyte with H225N injection. Whereas the currents show almost a linear current-voltage relationship at baseline (a), CO2 exposure resumed the inward rectification (b, 30 s; and c, 4 min into 15% CO2 exposure). Washout brought the currents back to their baseline status (d and e). f-j, current inhibition by CO2 in another oocyte with H225K injection. Unlike H225N, the H225K currents showed a clear inward rectification at the baseline (a), during 15% CO2 exposure (g and h), and after washout (i and j). Note that CO2 exposure produced a much larger inhibition in H225N (c) than H225K (h).

His-280 is a nonconserved histidine in the C terminus. Mutations of the His-280 to aspartate, lysine, or asparagine (H280D, H280K, or H280N) had no effect on channel sensitivity to CO₂. This was further confirmed by the double mutation of H280N/H342Q, which revealed a similar CO₂ sensitivity as the H342Q mutation (Fig. 4). His-270 is a conserved residue. Mutations of the His-270 to neutral, positive, or negative amino acid (H270N, H270K, and H270D) did not produce any functional channel. Thus, His-280 is not involved in CO sensing, whereas the function of His-270 in CO₂ and pH sensing remains unclear.

In inside-out patches, H225K, H274K, H342Q, H342D, H354K, and H354N showed similar inward rectifying K⁺ currents as their wild-type counterpart. Their pH sensitivity, however, was much smaller than that of the wild-type ROMK1. At pH 6.6, about 75% wild-type ROMK1 currents were suppressed, whereas only 20-40% inward rectifying currents were inhibited as one of these histidines was mutated (Fig. 7A). Their pK values were between pH 6.40 and 6.54 with an h of 3.2-3.4, in contrast to pK 6.73 and h 3.6 in the wild-type ROMK1 (Table II). A combined mutation of both these histidine residues (H342Q/H354N) markedly reduced the pH sensitivity (pK 6.37, h 3.0). This was even more evident in triple and quadruple mutations of these histidine residues (pK 6.25 and h 2.7 in the H225K/H342Q/H354N; pK 6.20 and h 2.7 in the H225K/H274K/H342Q/H354N). Although most of the wild-type ROMK1 currents were suppressed at pH 6.6 (Fig. 2A), this level of intracellular acidification had barely any effect on H225K/H342Q/H354N and H225K/H274K/H342Q/H354N currents (Figs. 6 and 7A). When baseline currents were compared between H342Q versus H342K, H354N versus H354K, and H225N versus H225K, we found that inward rectifying currents were significantly reduced in the mutants with a positive lysine to replace the histidine in comparison to mutations with a neutral amino acid (Table I).

View larger version (26K): [in this window] [in a new window]   Fig. 6.   Decrease in CO2 and pH sensitivities with histidine mutations. A, whole-cell currents were studied in the mutant ROMK1 (H225K, H274K, H342Q, and H354N). Exposure to 15% CO2 did not produce any significant inhibition of these currents. B, macroscopic currents were recorded from an inside-out patch under the same condition as in Fig. 2. Current amplitude remained the same in a pH range from 7.4 to 6.6. Further reduction in pHi to 6.2 caused an inhibition of these currents to about 50% of their baseline levels. Currents were completely suppressed at pHi 5.8. Note that eight superimposed traces are shown in each panel.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

In these studies, we have shown that ROMK1 channel is inhibited during hypercapnia. This is likely to be mediated by interactions of protons with the ROMK1 protein. We have demonstrated that four histidine residues in the C terminus of ROMK1 are involved in CO₂/pH sensing.

Inhibition of ROMK1 during Hypercapnia-- Previous studies have shown that certain unidentified Kir channels in the kidney are inhibited by hypercapnia (22-24). ROMK channels were cloned from the kidney and are known to be inhibited by a drop in intracellular pH (6-11). In our current studies, we have shown that the ROMK1 is indeed inhibited during hypercapnia. This inhibition is mediated by a decrease in intracellular pH, as selective decrease intra- but not extracellular pH inhibits this channel in whole-cell recordings and excised patches. It is known that protons can affect K⁺ currents by directly acting on channel proteins or by involving another intermediate molecules such as polyamines (8, 9, 10, 18, 25). The possibility of involving polyamines appears small, because both inward and outward rectifying currents are affected by CO₂ and pH in ROMK1 and its mutants. Also, we do not believe that the proton-induced inhibition is mediated by second messengers and protein phosphorylations, because the inhibition is seen in cell-free membrane patches in which cytosol-soluble factors are largely washed out. Furthermore, there are inhibitors for phosphatases and phosphodiesterases in our intracellular solutions, so that the turnover of protein phosphorylation and dephosphorylation may not occur at least in our experimental periods of pH exposures (1-2 min). Thus, it is likely that the ROMK1 channel is inhibited by CO₂ through interactions of intracellular protons with the channel protein.

Proton Bindings to Titratable Residues in ROMK1-- Protons can affect protein conformation and activity via their interactions with amino acid residues in the peptide chain. The potential amino acid residues that may interact with protons in the ROMK1 channel have been examined (8-10). In our current studies, we have revealed four histidine residues in the C terminus of ROMK1, which are likely to be the proton-binding sites. First, mutations of any of these residues significantly reduce channel sensitivity to CO₂ and pH. Second, greater reductions in pH sensitivity are seen when more histidine residues are mutated. The channel with three or four histidines mutated generally becomes CO₂-insensitive. Third, these effects of histidine mutations on channel sensitivities to CO₂ and pH were observed when Lys-80 and Thr-70 were intact, two residues known to be responsible for pH sensing in the N terminus of the ROMK1. In addition, baseline currents after a replacement of these histidine residues with positive amino acids are significantly smaller than those with neutral amino acids. Because the expression density of these mutant channels was not determined, it is unclear whether this is a result of surface expression of the channels or protonation of these histidine residues. When all these four histidines are mutated, we have found that the H225N/H274K/H342Q/H354N channel completely loses its CO₂ sensitivity. This is likely to result from the left shift of the pH titration curve in which this mutant channel does not respond to acidic pH unless it drops below pH 6.6 (Fig. 7A). We have also realized that extremely acidic pH can still inhibit the mutant channel, implying that there may exist unidentified proton sensor(s). We believe that the His-270 may be one of them. This histidine is conserved among Kir channels. We have tried to mutate it to positive, negative, or neutral amino acids but did not see any functional expression. Because another conserved histidine residue His-225 is involved in pH sensing, it is possible that the His-270 also has the proton-binding capability.

View larger version (26K): [in this window] [in a new window]   Fig. 7.   Concentration-dependent inhibitions of K+ currents in inside-out patches. The wild-type ROMK1 was strongly inhibited by low pHi with pK 6.73 and h 3.6, while the Kir2.1 barely responded to intracellular acidification in a pH range from 7.4 to 5.8. The pH sensitivity in ROMK1 was reduced when His-225, His-342, and His-354 were mutated in the channel protein. Single mutation of each of these residues shifted the pK values by about 0.2 pH units to the left. Double mutations (H342Q/H354N) decreased the pH sensitivity by 0.4 pH units. Mutations of all of them together (H225K/H342Q/H354N/) left-shifted the titration curve by 0.5 pH unit (pK 6.25, h 2.7). These effects are specific, because mutation of another histidine residue (H280N) had no effect. Also, mutation of this residue together with His-342 (H280K/H342Q) did not produce any additional effect on the titration curves for the H342Q. Data are presented as mean ± S.E. (n  4 for each test).

The His-225 is located in the corresponding position of His-206 in ROMK2. Mutation of the latter histidine to glycine (H2O6G) has been shown to enhance pH sensitivity (10). In our current studies, we have mutated it to another neutral and polar amino acid asparagine and found that the H225N mutant became more sensitive to CO₂ and pH_i. In addition, we have realized that the H225N mutant loses most of its inward rectification. This is not a result of leaky membranes, because whole-cell currents are strongly inhibited by CO₂ with clear inward rectification seen during the inhibition. Hence, the His-225 may be involved in the regulation of inward rectification in the ROMK1, as demonstrated recently on the His-216 in the Kir6.2 channel, which however apparently works in an opposite way (25). Interestingly, mutations of the His-225 to positive or negative amino acid suppress channel sensitivities to CO₂ and pH. What causes the different effects of these mutations is unclear. It is possible that this residue is normally engaged in the inward rectification. This engagement may limit the availability of the protein segment around this residue for the channel inhibition by intracellular protons. Mutation of this residue to a noncharged amino acid may lead to a release of this engagement allowing it to contribute to the channel inhibition during hypercapnia and intracellular acidification.

Fakler et al. (8) have shown that Lys-80 in ROMK1 is crucial in the pH-mediated channel inhibition. Mutation of this lysine residue to methionine virtually eliminates the channel sensitivity to pH_i (8, 10). However, subsequent studies have raised a question as to whether the Lys-80 is the only proton-binding site responsible for ROMK1 channel inhibition. First, several other residues are also involved in pH sensing in ROMK channels, such as Thr-51 and His-206 in ROMK2 (9-10) and His-225, His-274, His-342, and His-354 in ROMK1 as shown in the present study. Second, pH sensing in ROMK1 channel requires movements of both the N and C termini, in which structures other than the Lys-80 are required for the pH sensing (11). Multiple interaction sites of protons with the channel protein seem to be needed to produce such a large scale of movements, as has been demonstrated in the pH-dependent gating of the KcsA K⁺ channel (26). Third, the high Hill coefficient (h > 3) for pH-dependent inhibition of ROMK channels suggests the existence of multiple proton-binding sites (8-10). Finally, lysine has a pK value of 10.8 in its side chain and may not be titratable at physiological pH levels such as during an exposure to 5-15% CO₂. Although it has been postulated that Thr-51 and other unidentified residues might be responsible for lowering the pK point of this lysine residue (Lys-61) in ROMK2 (8, 9), there is still a lack of information indicating that this lysine is indeed protonated at near physiological pH levels.

If the Lys-80 is not the only proton-binding site, then how can we explain the phenomenon that the K80M mutant ROMK1 becomes pH-insensitive? Although a number of amino acid residues in the ROMK channels are involved in the pH sensing, clearly none of them has an effect as potent as the Lys-80. For instance, we have found that channel sensitivity to 15% CO₂ disappears in the K80M mutant with or without histidine mutations (Fig. 4). Because mutations of all other titratable and nontitratable residues diminish but not abolish the pH sensitivity in the mutant channels, it is possible that the Lys-80 does not work in parallel with other residues that are related to pH sensing. Instead, this lysine residue seems to act as a common final path or a gate in the modulation of the channel activity by other residues. The Lys-80 is located in an area in which several important channel modulations take place. For example, the presence of a threonine residue in this area makes the Kir2.3 channel sensitive to protein kinase C modulation. Introducing this threonine to a Kir2.1 channel that does not own this threonine in its wild type enables a PKC sensitivity in the mutant channel (27). This threonine is also important in the pH sensitivity of the Kir2.3 channel, despite the fact that it is not titratable at all (18). These observations therefore suggest that because of its location, this lysine residue becomes so critical in Kir channel gating by protons.

Functional Implication-- The finding of a high density of histidine residues modulated by protons in only the C but not N terminus is not only a surprise but also opens a question as to whether these two parts of intracellular segments play a different role in the pH-dependent channel gating. The pH sensing in ROMK channels requires four histidines that scatter over the C terminus. The pH sensing also requires lysine and a threonine residues that may not be titratable at physiological pH levels and are located at near the M1 region in the N terminus. A straightforward explanation of these experimental results would be that the lysine and threonine residues in the N terminus play a role in the pH-dependent interactions of N and C termini, whereas histidine residues in the C terminus are the proton-binding sites. Protonations of these histidine residues may lead to a change of the C terminus and convey it to a conformation with an enhanced binding affinity to the N terminus. The N-C interaction at the position near to the first membrane-spanning sequence around Lys-80 and Thr-70 may subsequently cause an inhibition of the channel activity. Because some of these histidine residues are present in Kir2.1 channel, this hypothesis may be helpful in explaining the phenomenon that Kir2.1 C terminus in the Kir1.1-Kir2.1 and Kir2.3-Kir2.1 chimerical channels can carry out some pH sensitivity like its counterpart in the ROMK1 and Kir2.3 channels (8, 18).

In conclusion, the ROMK1 channel is inhibited during hypercapnia and intracellular acidification. This effect is likely to be mediated by interactions of protons with the ROMK1 protein, and four histidine residues in the C terminus of ROMK1 are involved.

	ACKNOWLEDGEMENTS

We thank Drs. Steven Hebert and Lily L. Jan for their gifts ROMK1 and Kir2.1 cDNAs.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant HL58410-01 and the Grant-in-Aid Award 9950528N from the American Heart Association.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. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Dept. of Biology, Georgia State University, 24 Peachtree Central Ave., Atlanta, GA 30303-4010. Tel.: 404-651-0913; Fax: 404-651-2509; E-mail: biocjj@panther.gsu.edu.

	ABBREVIATIONS

The abbreviation used is: PIPES, 1,4-piperazinediethanesulfonic acid.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES