Mechanism for proton conduction of the M(2) ion channel of influenza A virus.

The M(2) integral membrane protein of influenza A virus forms a proton-selective ion channel. We investigated the mechanism for proton transport of the M(2) protein in Xenopus oocytes using a two-electrode voltage clamp and in CV-1 cells using the whole cell patch clamp technique. Membrane currents were recorded while manipulating the external solution to alter either the total or free proton concentration or the solvent itself. Membrane conductance decreased by approximately 50% when D(2)O replaced H(2)O as the solvent. From this, we conclude that hydrogen ions do not pass through M(2) as hydronium ions, but instead must interact with titratable groups that line the pore of the channel. M(2) currents measured in solutions of low buffer concentration (<15 mM in oocytes and <0.15 mM in CV-1 cells) were smaller than those studied in solutions of high buffer concentration. Furthermore, the reversal voltage measured in low buffer was shifted to a more negative voltage than in high buffer. Also, at a given pH, M(2) current amplitude in 15 mM buffer decreased when pH-pK(a) was increased by changing the buffer pK(a). Collectively, these results demonstrate that M(2) currents can be limited by external buffer capacity. The data presented in this study were also used to estimate the maximum single channel current of the M(2) ion channel, which was calculated to be on the order of 1-10 fA.

The M 2 protein of influenza A virus is thought to function as an ion channel that permits protons to enter virus particles during virion uncoating in endosomes. In addition, in influenza virus-infected cells, the M 2 protein causes the equilibration of pH between the acidic lumen of the trans-Golgi network and the cytoplasm (1,2). The M 2 protein consists of a 24-residue N-terminal extracellular domain, a single internal hydrophobic domain of 19 residues that acts as a transmembrane domain and forms the pore of the channel, and a 54-residue cytoplasmic tail (3). Chemical cross-linking studies showed the M 2 protein to exist minimally as a homotetramer (4 -6). More recently, statistical analysis of the ion channel activity of mixed oligomers indicated that a homotetramer is also the minimal active oligomeric form of the protein (7).
Despite the small size of the active M 2 oligomer, several pieces of evidence indicate that ion channel activity is intrinsic to the M 2 protein. First, ion channel activity has been observed in three different expression systems: Xenopus oocytes (8 -10), mammalian cells (11,12) and yeast (13). Second, M 2 channel activity has also been recorded in artificial lipid bilayers from a reconstituted peptide corresponding to the transmembrane domain of the M 2 protein (14) and from purified M 2 protein (15). Thus, due to its structural simplicity, the M 2 ion channel is a potentially useful model for the study of ion channels in general.
Based on calculations using the Goldman-Hodgkin-Katz equation and measurements of current reversal voltage, the M 2 ion channel is thought to be at least 10 5 -fold selective for protons (8,11), although other monovalent cations may also permeate the channel (8). M 2 ion channel activity is increased when the pH of the extracellular domain is lowered (10,11,16). This increase in activity occurs within the range of pH values expected for titration of histidine (17). The only amino acid in the transmembrane domain of the M 2 protein with a titratable group in this pH range is His 37 ; and when His 37 is replaced by Ala, Gly, or Glu, the proton selectivity of the channel is greatly reduced, and the channel is conductive over a wider range of pH values (10,11). It has been proposed that His 37 forms a selectivity filter for protons (18).
Two possible mechanisms could account for the high proton selectivity of the M 2 ion channel. First, it is possible that certain residues of the pore form a narrow selectivity filter through which only hydronium ions can pass. Second, a residue in the pore region might form part of a conducting pathway, perhaps a proton wire (19), by providing a site highly favorable for interactions with protons. We distinguished between these possibilities by replacing the water solvent with D 2 O.
We have observed that inward M 2 currents sometimes decrease during a constant voltage clamp pulse to large negative voltages. Since M 2 currents do not display rapid voltage-or time-dependent activation/inactivation, it is possible that the decrease in current is the result of insufficient buffer capacity at the external mouth of the channel. We tested this possibility by examining the effect of decreased buffer capacity on M 2 current amplitude.

mRNA Synthesis and Site-specific Mutagenesis
The cDNA to the influenza A/Udorn/72 mRNA was cloned into the BamHI site of pGEM3 such that mRNA sense transcripts could be generated by using the bacteriophage T7 RNA polymerase promoter and T7 RNA polymerase. For in vitro transcription, plasmid DNAs were linearized downstream of the T7 promoter and the M 2 cDNA with XbaI. In vitro synthesis and quantification of m 7 G(5Ј)ppp(5Ј)G-capped mRNA were carried out as described previously (10). * This work was supported by United States Public Health Service Research Grants GM56423-02 (to J. D. L.), AI-20201 (to R. A. L.), and AI-31882 (to L. H. P.) from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Culture and Infection of CV-1 Cells
CV-1 cells were cultured and infected with recombinant simian virus 40 expressing the M 2 protein from influenza A/Udorn/72 (rSV40-M 2 ) as described previously (12). Briefly, CV-1 cells grown to confluency at 37°C and 5% CO 2 in culture medium (Dulbecco's modified Eagle's medium, 10% fetal calf serum, penicillin, and streptomycin) were trypsinized, pelleted, and resuspended in culture medium. Resuspended cells were incubated in the presence of high titer rSV40-M 2 (100 l of resuspended CV-1 cells and 1 ml of virus stock) for 4 h. Infected cells were then diluted 1:1 in culture medium and seeded onto 5-mm square glass coverslips arranged in 3.5-cm Petri dishes (2-ml total volume/dish). Infected cells were incubated for 48 h before recording to ensure adequate M 2 protein expression. The presence of M 2 protein at the surface of infected CV-1 cells was confirmed by indirect immunofluorescence and flow cytometry as described previously (12).

Measurement of Membrane Current
Oocytes-Whole cell currents were measured using a two-electrode voltage clamp. Electrodes were filled with 3 M KCl, and the oocytes were bathed in either Barth's solution (88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO 3 , 0.3 mM NaNO 3 , 0.71 mM CaCl 2 , 0.82 mM MgSO 4 , and 15 mM HEPES (pH 7.5), osmolality of ϳ210 mosmol/kg) or a modified solution during the recording. Continuous current-voltage relationships were measured with ramps of membrane voltage since the M 2 channel shows no rapid voltage-or time-dependent gating. These ramps typically spanned a range of 120 mV in 2 s. Oocyte holding potential was Ϫ20 mV unless stated otherwise.
CV-1 Cells-M 2 currents were recorded from CV-1 cells using the whole cell patch clamp technique as described previously (12). Briefly, patch pipettes with tip diameters of ϳ2-3 m were pulled from borosilicate capillary glass, fire-polished, and then partially filled with pipette solution (145 mM KCl, 5 mM EGTA, 1 mM MgCl 2 , 5 mM NaCl, and 15 mM HEPES (pH 7.4) adjusted with KOH, osmolality of 300 -310 mosmol/kg). Pipettes filled with this solution typically had resistances of ϳ3-4 megaohms. CV-1 cells attached to glass coverslips were transferred to a recording chamber filled with "high buffer" bath solution (140 mM NaCl, 5.3 mM KCl, 0.55 mM MgSO 4 , 5.5 mM glucose, and 15 mM HEPES (pH 7.4) or 15 mM MES 1 (pH 6.2), osmolality of ϳ300 mosmol/ kg). Also used in experiments was a "low buffer" pH 6.2 solution that contained 25 M bromcresol purple instead of 15 mM MES and additional NaCl to adjust the solution osmolality to 300 mosmol/kg. Seals (in excess of 10 gigaohms) were made by gently pressing the patch pipette against a CV-1 cell and then rapidly applying ϳ12 mm Hg suction. The whole cell configuration was achieved by using a brief pulse of high voltage combined with gentle pipette suction. In the whole cell configuration, cells had access resistances of Ͻ10 megaohms. Cells were generally bathed in pH 7.4 solution and held at Ϫ20 mV. Whole cell currents were recorded after the bathing solution was changed from pH 7.4 to pH 6.2 using a Fast Step perfusion system (Model SF77B, Warner Instruments Corp., Hamden, CT). Using this system, solution changes could be made in Ͻ100 ms. M 2 -specific currents were identified by sensitivity to block by 100 M amantadine.

pD Measurement
The nominal reading taken from a glass pH electrode (pH nom ) deviates from the true pD of D 2 O solution by 0.4 units such that pD ϭ pH nom ϩ 0.4 (21). Our pH meter read 0.41 Ϯ 0.02 units (mean Ϯ S.D., n ϭ 4) higher when 0.01 M HCl was added to H 2 O than when added to D 2 O. We therefore corrected the pD in D 2 O solutions by adding 0.41 to the nominal reading of our pH meter.

RESULTS
pH Activation of M 2 Currents-As we have shown previously (10,17), whole cell currents recorded in oocytes expressing the M 2 protein increase when the pH of the bathing medium is lowered (Fig. 1). This current is effectively blocked in the presence of 100 M amantadine, an anti-influenza virus compound and inhibitor of M 2 ion channel activity (10).
Deuterium Currents of the M 2 Ion Channel-To test the possibility that protons interact directly with the M 2 ion channel protein when traversing its pore region, we took advantage of some differences in the physical properties of H 2 and D 2 . First, D 2 O is 1.25 times more viscous than H 2 O. If protons pass through the M 2 channel as hydronium ions, then only a modest decrease in current on changing from H 2 O to D 2 O solvent would be expected according to the ratio of their viscosities. On the other hand, if protons traverse the M 2 channel via a proton wire, involving the exchange of protons between H 2 O molecules occupying the channel pore, or if protons interact directly with the channel, then a larger conductance decrease upon changing from H 2 O to D 2 O should occur since there exists a large difference in both the mobilities and the zero point energies of protons and deuterons, respectively (19,22). To test between these possibilities, we compared the conductance and reversal voltage of M 2 -expressing oocytes bathed in water and D 2 O solvents. The water permeability of the oocyte membrane is very high, and it has been demonstrated that the external solvent will determine the internal solvent of the cell in these experiments (19). An alteration of the internal solvent would be expected to modify the internal hydrogen ion concentration of the cell since the dissociation constant for many buffers is lower for protons than for deuterons. Thus, simply changing the external solvent from H 2 O to D 2 O at constant pL out (pH out or pD out ) would be expected to increase the pL in of the ooplasm. We compared the conductance of M 2 -expressing oocytes at three or more values of pL out in the same oocyte between pL out ϭ 5.0 and pL out ϭ 7.5. This was done by first determining the current-voltage relationship at two or more values of pH out (Fig. 2). Next, the current-voltage relationship was determined in D 2 O solutions with the same values of pL out (Fig. 2). At least 20 s was allowed for equilibration of the solvent. This time was sufficient, as no differences were observed in the current-volt- 1 The abbreviation used is: MES, 4-morpholineethanesulfonic acid. age relationships obtained 20 s and 2 min after changing the solvent. To check that irreversible changes did not occur in D 2 O, the relationship was again determined in water at pH 6.2. Finally, the relationship was determined in the presence of amantadine (pH 6.2). For every value of pL out tested, the current-voltage relationship of the amantadine-sensitive current had the same overall shape in both solvents, with a higher slope conductance for more negative membrane voltages (Fig.  2). Also, for every value of pL out tested, the relationship shifted to more positive voltages (Fig. 3A), and its slope decreased (by 50 -60%) when D 2 O was substituted for H 2 O (Fig. 3B).
Effects of Buffer Concentration on M 2 Currents-We have often observed that the inward current of M 2 -expressing oocytes decreases slightly during a constant voltage clamp pulse to large negative voltages. This effect becomes more prevalent as the flow rate of the solution bathing the oocyte is decreased. One possible explanation for this observation is that the buffer concentration was inadequate to supply the channel with protons. Unlike ion channels that conduct Na ϩ , Cl Ϫ , Ca 2ϩ , etc., proton-selective ion channels are supplied with the conducting ion at a free concentration in the micromolar range. Free protons are supplied by dissociation from buffer molecules, enabling the total concentration of protons to greatly exceed the free concentration. To test this, we measured the dependence of M 2 current amplitude on external buffer concentration.
In solutions of low buffer concentration (Ͻ15 mM), amantadine-sensitive M 2 currents measured in Xenopus oocytes decreased with time and had smaller final current amplitudes (Fig. 4A). M 2 current decrease in low buffer was measured at a given voltage as the difference between the current amplitude in 15 mM buffer and the final current amplitude measured at lower buffer concentrations. At higher buffer concentrations (Ͼ15 mM), amantadine-sensitive M 2 currents had approximately the same amplitude as those measured in 15 mM buffer (Fig. 4B). For a given buffer concentration, the magnitude of the current decrease in low buffer was proportional to the inward current amplitude in high buffer (Fig. 5). The relationship between inward current decrease in low buffer and inward current amplitude in high buffer was steeper for lower buffer concentrations (Fig. 5). No change in current amplitude was observed in low buffer when cells were held at a voltage that corresponded to zero current, which for most cells was near the calculated proton equilibrium potential. Low external buffer concentrations also had no effect on the amplitude of outward M 2 currents (data not shown).
Are current by external buffer concentration was specific to oocytes, we compared the effects of low buffer concentration on M 2 currents recorded in mammalian cells infected with rSV40-M 2 (12). M 2 currents recorded in rSV40-M 2 -infected CV-1 cells using the whole cell patch clamp technique were severely reduced in amplitude when the external buffer concentration was decreased from 15 mM to 25 M using bromcresol purple as a pH buffer (Fig. 6). We chose to use bromcresol purple as a buffer since it undergoes a distinct color change below pH 6.0, allowing the approximate pH of the solution to be easily monitored during the experiment. On average, M 2 currents recorded in the low buffer solution at Ϫ60 mV were decreased in amplitude to 31.0 Ϯ 8.1% (n ϭ 7) of those recorded in 15 mM buffer. Only a slight decrease in current was observed when the buffer concentration was reduced from 15 to 0.15 mM. On average, M 2 currents recorded in 0.15 mM buffer at Ϫ60 mV were decreased in amplitude to 88.98 Ϯ 2.17% of those recorded in 15 mM buffer (n ϭ 4). M 2 currents measured in CV-1 cells in low buffer were decreased in amplitude at all times during a constant voltage clamp pulse. This result is in contrast to oocytes in which the current in low buffer was seen to decrease over a matter of seconds during a voltage clamp pulse. The results demonstrate that the limitation of M 2 current by buffer concentration is not restricted to oocytes and therefore suggest a common mechanism.
Mechanism of M 2 Current Decrease-The dependence of M 2 current amplitude on external buffer concentration suggests that M 2 proton currents can exceed the capacity of the buffer to deliver protons at the channel mouth, thus resulting in a transient extracellular alkalinization. The occurrence of a transient alkalinization in low buffer has already been suggested in Fig.  4A (arrow), where there was an "overshoot" of current following a hyperpolarizing pulse. If alkalinization occurs, then the reversal voltage for the M 2 current (V rev ) measured in low buffer should be shifted to a more negative voltage since the pH gradient across the channel would be transiently decreased under this condition. To test this hypothesis, the following experiment was performed. V rev was measured using voltage ramps evoked first while the oocyte was bathed in high buffer, then while it was bathed in low buffer, and finally again while it was bathed in high buffer (Fig. 7). Rapid voltage ramps were necessary to measure the effect of depletion on V rev since there should be no depletion occurring at true V rev . We found that at a holding potential of Ϫ20 mV, which produces a large inward In this case, currents were evoked by a 2-s test pulse to Ϫ120 mV from a holding voltage of Ϫ20 mV. For a given final current, the decrease in M 2 current became larger as the buffer concentration was decreased. current at pH 5.8, V rev shifted in the low buffer solution toward more negative voltages by Ϫ12.5 Ϯ 1.0 mV (n ϭ 9). The transient shift of V rev in the low buffer solution was larger for more negative holding potentials, which was consistent with greater depletion. The transient shift of V rev in low buffer could be prevented by holding cells at a voltage that corresponded to zero inward current (data not shown). The incomplete recovery of V rev in high buffer following exposure to the low buffer solution is thought to be the result of gradual oocyte acidification while bathed in low pH solutions (8). Consistent with this explanation was a small change in V rev (2-5 mV) toward more negative voltages observed in oocytes exposed only to a highly buffered solution (data not shown). The mostly reversible shift of V rev in low buffer to more negative values is consistent with the notion that the M 2 current decrease in low buffer is the result of proton depletion in the bulk solution near the extracellular mouth of the M 2 channel.
Does the Vitelline Membrane Limit the Diffusion of Buffer to the M 2 Channel?-Xenopus oocytes are surrounded by the vitelline membrane, which, among other functions, serves to maintain oocyte structural integrity. For this reason, the vitelline membrane is generally left intact for ion channel recording. It was possible that the greater sensitivity of M 2 currents in oocytes to low external buffer concentrations and the slower time course of current decrease compared with those recorded in mammalian cells may have been the result of the vitelline membrane acting as a diffusion barrier for protonated buffer molecules. We therefore tested the effect of decreased buffer concentration in oocytes from which the vitelline membranes had been removed and compared the results to those obtained in vitelline-intact oocytes. M 2 currents recorded in oocytes with no vitelline membrane decreased in a similar manner to those recorded in vitelline-intact oocytes when bathed in buffer at concentrations below 15 mM (n ϭ 10; data not shown). This result demonstrates that the greater sensitivity of M 2 currents to low external buffer concentrations seen in oocytes is not the result of the vitelline membrane impeding buffer diffusion to the mouth of the M 2 channel.

Is Proton Depletion Localized to the M 2 Channel
Macromolecular Complex?-We wished to determine if proton depletion in low external buffer occurred in the macroscopic solution shared by all channels as the result of combined proton flux through multiple M 2 channels or whether the deduced depletion was localized to the bulk solution surrounding the macromolecular complex of each M 2 channel. Assuming the same mean channel conductance for all M 2 channels, oocytes expressing a greater total number of channels (i.e. expression level) should, for a given voltage and pH, have proportionally larger currents. If proton depletion in low buffer is channellocalized, then the ratio of current amplitude in low buffer to current amplitude in high buffer should be constant for a given voltage and pH, regardless of the expression level. On the other hand, if current decrease is due to the cooperative effect of many channels to deplete protons from the macroscopic solution, then the ratio of current amplitude in low buffer to current amplitude in high buffer should decrease with expression level. To distinguish between these two possibilities, M 2 current amplitude was measured at high and low buffer concentrations at a constant voltage and pH in individual oocytes, each with many levels of expression. The ratio of current amplitude in low buffer to current amplitude in high buffer was then plotted for each oocyte against the amplitude of the current measured in the high buffer solution (Fig. 8). We found that the ratio of current amplitude in low buffer to current amplitude in high buffer for oocytes with different levels of expression was approximately constant, suggesting that proton depletion is a channel-localized phenomenon.
Is M 2 Current Amplitude Dependent on Buffer Capacity?-Buffer capacity at a given pH is dependent on both the concentration of protonated buffer species and the buffer pK a . We therefore tested whether M 2 currents measured at a buffer concentration found previously to adequately supply the channel (e.g. 15 mM) could also be limited by changing the buffer pK a at a constant pH (Fig. 9). This was done by measuring M 2 currents at pH 6.2 and pH 4.4 in the presence of 15 mM MES (pK a ϭ 6.1), HEPES (pK a ϭ 7.4), or tartaric acid (pK a ϭ 4.2). At pH 6.2, currents measured in 15 mM MES did not decrease measurably at a given voltage. However, a smaller current amplitude was observed when the external solution was buffered with HEPES or tartaric acid. The smallest M 2 current amplitude at pH 6.2 was observed when tartaric acid was used as a buffer. At this pH, tartaric acid would be the least protonated of the three buffer species. In contrast, tartaric acid provided the largest current of the three buffers when the experiment was repeated at pH 4.4, even though at this pH, both HEPES and MES are highly protonated. We chose to measure currents at pH 4.4 instead of pH 4.2 (the pK a of tartaric acid) in order to minimize acidification and subsequent oocyte damage, which occurs at low pH values. The results thus demonstrate that M 2 currents are limited by buffer capacity and not just the concentration of protonated buffer species.

DISCUSSION
Our results are consistent with a model for proton conduction across the M 2 ion channel in which protons donated from buffer molecules interact with a titratable group within the channel pore. At least one such group lies within the electric field of the membrane and is probably His 37 .
Comparison of the conductance measured in water and D 2 O suggests that hydrogen ions do not pass through the channel in the form of hydronium ions. If this were the case, only a modest decrease in conductance would be expected due the greater viscosity of D 2 O, ϳ20%. Instead, we found that the slope conductance, measured over a range of ϳ1.5 pH units, decreased by 40 -50% (Fig. 3B). The greater decrease in conductance could be explained either by deuterons having a greater affinity than protons for a titratable group lining the M 2 channel pore (22) or by the large difference in the mobilities of protons in H 2 O and deuterons in D 2 O (19).
Several lines of evidence indicate that one of the ionizable groups of the channel that binds the conducting protons lies within the electric field of the membrane. First, the slope conductance of the current-voltage relationship increases for negative voltages (Fig. 1). The voltage dependence that we measured is similar to that found previously in mouse erythroleukemia cells (11). Second, site-directed mutagenesis experiments have shown that His 37 is essential for the activation of the current of the M 2 channel by low pH (17). Experiments with various transition metals have shown that the channel can be inhibited with Cu 2ϩ and that the high affinity site for inhibition results from coordination of Cu 2ϩ with His 37 (23). This inhibition is strongly voltage-dependent, indicating that His 37 lies within the electric field of the membrane. It thus seems likely that His 37 binds protons as they traverse the pore of the channel (18).
The conclusion that hydrogen ions interact with the pore region of the M 2 ion channel has implications for the single channel conductance expected from this proton-selective channel. The M 2 single channel current can be estimated in the following three ways.
1) The maximum theoretical flux of hydrogen ions would be limited by the reverse reaction rate constant of the titratable group to which the hydrogen ions bind. If this group is His 37 (pK a ϳ 6), then the calculated value of the reverse reaction rate constant (His⅐H ϩ ϩ H 2 O 3 His ϩ H 3 O ϩ ) is on the order of 1.7 ϫ 10 4 /s (22). The maximum single channel current to be expected based upon this value and assuming four His 37 residues per M 2 channel macromolecular complex is ϳ10 fA.
2) We found that M 2 currents measured at a given voltage in CV-1 cells diminished if the buffer concentration of the oocyte bathing solution was below ϳ0.15 mM. Thus, the number of protons supplied to a single M 2 channel macromolecular complex in 0.15 mM buffer must be roughly equal to the number of protons traversing the channel during maximum flux. Assuming that the mobility of a proton in H 2 O is very high, ϳ10 Ϫ4 (cm/s)/(V/cm) (24), then the limiting step in the transfer of protons to His 37 would be the dissociation of protons from the buffer molecules. Assuming (a) that the proton dissociation rate is 10 5 /s for a buffer of pK a ϭ 6 and (b) that protons diffuse into the channel from a hemispherical sink with a radius of 5 Å, then the maximum current possible from 0.15 mM buffer would be equal to the number of protonated buffer molecules in the sink multiplied by the proton dissociation rate, a value of ϳ1 fA.
3) The final way in which the M 2 single channel current can be estimated is by dividing the total current recorded in M 2expressing oocytes by the number of M 2 channels per oocyte. The total current measured at pH 6.2 and a membrane voltage of Ϫ130 mV is ϳ0.7 A. The number of M 2 channels per oocyte can be calculated by assuming that there is ϳ3 ng of M 2 protein expressed per oocyte and by using a molecular weight for the M 2 tetrameric channel of 60,000 (9). Given that ϳ50% of the M 2 protein is present at the cell surface (7), the calculated current per M 2 channel is ϳ0.5 fA. All three calculations suggest that the M 2 single channel current is very small. Two reports of a larger single channel activity of the M 2 ion channel have appeared in the literature. The first study used a synthetic transmembrane peptide and reported a conductance of 10 picosiemens (14). However, this value was obtained at pH 2.3 in glycine buffer, far from the pH range accessible to electrophysiological investigations. The second study utilized affinity-purified M 2 protein and did not report consistent single channel conductances (15). Such inconsistency of single channel conductance perhaps resulted from aggregation of the protein in the membrane (25). In a patch clamp study of stably transfected mammalian cells, Chizhmakov et al. (11) reported that M 2 single channel currents were not detectable. Our results lend support to the expectation that the M 2 single channel current would be below the detectable range of conventional single channel recording techniques.
We found some differences in M 2 currents recorded in Xenopus oocytes and CV-1 cells. First, M 2 currents in CV-1 cells are not as sensitive to external buffer concentration as those re- FIG. 9. Dependence of M 2 current amplitude measured in M 2 -expressing oocytes on the difference between the solution pH and the buffer pK a . Examples of M 2 currents measured in a single oocyte in three different buffers at pH 6.2 and pH 4.4 are shown in A. Currents shown were evoked by a 2-s voltage pulse to Ϫ60 mV from a holding potential of Ϫ20 mV. The buffers used were 15 mM HEPES (pK a ϭ 7.4), 15 mM MES (pK a ϭ 6.1), and 15 mM tartaric acid (TA; pK a ϭ 4.2). Dotted lines show zero current level. Note the overshoot of current (arrows) following the voltage pulse that occurred in tartaric acid at pH 6.2 and in HEPES at pH 4.4. A summary of the results is shown in B. The average ratios of current amplitude in 15 mM HEPES (white bars), 15 mM MES (black bars), and 15 mM tartaric acid (gray bars) to the maximum current amplitude obtained with any of the three buffers are shown at both pH 6.2 (maximum current in MES) and pH 4.4 (maximum current in tartaric acid). Data were obtained from eight oocytes, each being exposed to all three buffers at both pH values. Error bars show S.E. The greatest M 2 current amplitude was observed when the buffer pK a matched the solution pH. corded in Xenopus oocytes (Figs. 4 -6). Second, M 2 currents recorded in CV-1 cells in low buffer decrease and reach a lower steady-state current almost instantaneously compared with those recorded in Xenopus oocytes, which decrease over a matter of seconds. This was not the result of the oocyte vitelline membrane acting as a diffusion barrier, as removing the vitelline membrane made no difference to the time course of current decrease. One possibility is that there is a greater M 2 current density in oocytes, which leads to more protons being depleted from the surrounding solution. This can be tested by calculating the M 2 current density of M 2 -expressing oocytes and comparing this with the equivalent value obtained from CV-1 cells. M 2 current density can be calculated by dividing the whole cell current at a given voltage and pH by the cell-surface area. The latter can be calculated using the standard cell membrane capacitance of 1 microfarad/cm 2 . The calculated values for M 2 current density in oocytes and CV-1 cells are similar, being 0.04 and 0.06 pA/m 2 , respectively.
A third possibility is that the presence of surface villi in oocytes (but not in CV-1 cells) increases the unstirred layer at the oocyte surface, leading to greater ion depletion through insufficient mixing with the bulk solution (26). If this is the case, then the protonated buffer concentration in the unstirred layer near the mouth of the M 2 channel may be different from the protonated buffer concentration in the macroscopic solution shared by all channels. Thus, the actual concentration of buffer below which oocyte M 2 currents decrease may be closer to the value obtained in CV-1 cells, which have a comparatively smooth surface. The presence of villi causing a diffusion barrier might also explain the slower time course of current decrease observed in oocytes. Although the effect of surface villi on M 2 currents could not be tested directly, the presence of an unstirred layer is suggested by the observation that oocyte currents could be decreased by reducing the bathing solution flow rate even in 15 mM buffer. M 2 current amplitude at a given pH was found to be strongly dependent on the buffer pK a . M 2 currents decreased in amplitude when the buffer pK a differed from the solution pH ( Fig. 9) even though a relatively high buffer concentration was used. This result can be explained in the same way as the dependence of M 2 current amplitude on external buffer concentration by assuming that a buffer is at maximum capacity to provide protons to the solution only when the pK a of the buffer equals the pH of the solution. Our results cannot, however, distinguish between the following mechanisms: 1) protons are transferred from the buffer to water molecules to form hydronium ions, which then donate protons to titratable groups in the channel pore; and 2) protons are transferred directly from the buffer to a group in the channel with a pK a similar to that of water.
Our results suggest that protons passing through the M 2 ion channel interact directly with the channel protein instead of passing through the channel as hydronium ions. We demonstrate that M 2 currents can be limited by buffer capacity and are in contrast to those obtained with other proton-specific channels through which proton flux was not limited by modest decreases in buffer capacity (27,28). In previous work (10,17), we have demonstrated that His 37 is an essential residue for pH sensitivity of the M 2 channel, and thus, this residue would be expected to be a site for proton binding to the channel molecule. Modification of the imidazole side chain of His 37 may reveal the atoms that participate in this interaction.