The Ca:H coupling ratio of the plasma membrane calcium ATPase in neurones is little sensitive to changes in external or internal pH

To explore the effects of both external and internal pH (pH o and pH i ) on the coupling between Ca 2+ extrusion and H + uptake by the PMCA activity in snail neurones H + uptake was assessed by measuring surface pH changes ( ∆ pH s ) with pH-sensitive microelectrodes while Ba 2+ or Ca 2+ loads were extruded. Ru360 or ruthenium red injection showed that injected Ca 2+ was partly taken up by mitochondria, but Ca 2+ entering through channels was not. External pH was changed using a mixture of three buffers to minimise changes in buffering power. With depolarisation-induced Ca 2+ or Ba 2+ loads the ∆ pH s was not changed significantly over the pH range 6.5 to 8.5. With Ca 2+ injections into cells with mitochondrial uptake blocked the ∆ pH s were significantly smaller at pH 8.5 than at 7.5, but this could be explained in part by the slower rate of activity of the PMCA. Low intracellular pH also changed the ∆ pH s responses to Ca 2+ injection, but not significantly. Again this may have been due to reduced pump activity at low pH i . I conclude that in snail neurones the PMCA coupling ratio is either insensitive or much less sensitive to pH than in red blood cells or barnacle muscle. 196 words


Introduction
Cell calcium must be kept very low to allow fast and efficient intracellular Ca 2+ signalling. In nerve cells two ATPase pumps are involved in removing Ca 2+ from the cytoplasm. These are the sarco-endoplasmic reticulum calcium ATPase (SERCA) which drives uptake into the endoplasmic reticulum and the plasma-membrane calcium ATPase (PMCA) which extrudes Ca 2+ across the plasma membrane. As first shown by Niggli et al many years ago the latter couples the efflux of Ca 2+ to the uptake of H + ions [1]. The properties of the PMCA have been reviewed recently [2,3].
The PMCA is inhibited by extracellular alkalinisation and intracellular acidification [4,5,6,7,8]. In some cells there is a second plasma membrane Ca 2+ pump driven by Na + influx rather than by ATP, the Na + /Ca 2+ exchanger, but in many nerve cells the PMCA is the principle Ca 2+ extrusion mechanism at low Ca 2+ loads. [4,7,9 10].
In snail neurones the PMCA is the only mechanism for calcium extrusion from the cell body [11], even though it is sometimes described as a low-capacity system [3]. The ratio of the coupling between Ca 2+ extrusion and H + uptake at normal pH levels by the PMCA remains controversial. While the early workers [1] concluded that the Ca 2+ : H + ratio was 1:2, some subsequent work has suggested fewer H + ions are transported [12,13,14]. A recent review [2] states categorically that the ratio is 1:1, as does a recent physiology textbook [15]. Many biochemistry textbooks still report that the PMCA is a uniporter (see [14]). There have also been several reports that the PMCA is electrogenic, for example in hair cells [16] and in red blood cell preparations [17,18]. Electrogenicity implies that the coupling is not 1Ca 2+ : 2H + . In contrast, in snail neurones I recently found that the ratio under normal conditions is 1Ca 2+ : 2H + [19].
The effects of extracellular pH (pH o ) on the coupling ratio have been studied in red blood cells and barnacle muscle, but not in nerve cells. In red blood cell PMCA preparations the Ca 2+ : H + ratio is changed by pH o from about 1 : 2 at pH 6.5 to 1 : zero at pH 8.5 [20]. Similarly in barnacle muscle at external pH 6.5 the PMCA coupling ratio was 1 : 3, while at pH 8.2 the ratio was 1 : 1 [21]. The effects of changes in intracellular pH (pH i ) have not apparently been investigated in any cell.
The quantity of H + ions pumped into neurones by the PMCA is important because many channels and carriers are sensitive to small pH changes [22,23]. The possible pH sensitivity of the coupling ratio may also have important consequences for the molecular mechanism of the PMCA. 3 To investigate the effects of pH on the coupling ratio in nerve cells I have extended my recent experiments [19] on the large neurones of the common snail Helix aspersa. Using both conventional and ion-sensitive microelectrodes I have recorded membrane potential, clamp currents, intracellular and surface pH and intracellular Ca 2+ in intact cells. The PMCA was stimulated by depolarisation in Ca 2+ or Ba 2+ solutions or by direct iontophoretic injection of Ca 2+ . The subsequent PMCAgenerated surface pH changes (∆pH s ) were measured. The two other processes that might change surface pH, namely intracellular pH regulation and H + channels, were inhibited or kept inactivated. Normal pH i regulation was inhibited by removal of bicarbonate [24] which leaves only a relatively weak Na + /H + exchanger [25]. H + channels were kept closed by avoiding large depolarisations [26]. To ensure that injected Ca 2+ was not taken up by mitochondria, cells were injected with ruthenium red [27] or Ru360 [28]. This proved unnecessary for cells loaded by depolarisation.
For the experiments on the effects of pH o I measured the ∆pH s for equal depolarisations or Ca 2+ injections at different pH o values. For the experiments on pH i effects I measured the ∆pH s induced by equal Ca 2+ injections while changing pH i by HCl injection or application and removal of CO 2 /bicarbonate. With pH o between 7.5 and 8.5 I have found significant changes in the PMCA-induced ∆pH s only with injected Ca 2+ loads. Allowing for variation in the pump rate with pH, these findings suggest that the snail neurone PMCA coupling ratio is little changed by external or internal pH

General
Experiments were done on large (150 -250 µm diameter) neurones in isolated sub-oesophageal ganglia of the common snail, Helix aspersa [8,29,19]. Cells were voltage-clamped to -50 mV using two microelectrodes. The PMCA was stimulated either by depolarisations in Ca 2+ or Ba 2+ snail Ringer's solution, or by iontophoretic injection of Ca 2+ ions. The changes in surface pH (∆pH s ) resulting from the subsequent extrusion of Ba 2+ or Ca 2+ were measured using pH-sensitive microelectrodes which were pressed against the surface of the chosen cell. The effects of changing pH o were assessed by changing the superfusate pH using a mixture of three different buffers (with different dissociation constants) to minimise changes in buffering power. To reduce pH i , H + ions were injected by iontophoresis, while to increase pH i the preparation was first equilibrated with bicarbonate-buffered 4 saline and then returned to normal snail Ringer's solution. The consequent loss of accumulated intracellular bicarbonate as CO 2 caused a pH i increase of 0.4-0.5 units.
In some experiments changes in intracellular Ca 2+ were followed with a Ca 2+sensitive microelectrode (CaSM).

Preparation
An aestivating snail was killed by rapid removal of the circumoesophageal ring of ganglia, and the large cells on the dorsal side of the suboesophageal ganglia exposed as previously described [19]. All experiments were carried out at room temperature, 18-23°C, starting at least 1 h after the dissection.

Solutions
The normal snail Ringer solution contained (mM): 80 NaCl, 4 KCl, 7 CaCl 2 , 5 MgCl 2 , and 20 Hepes, titrated with NaOH to pH 7.5. Solutions of pH 6.5, 7.5 and 8.5 with a low buffering power had 2 or 5 mM of Pipes, Hepes and Taps, with additional NaCl to maintain tonicity. pH 9.5 Ringer was buffered with 20 mM CHES. Ba Ringer solutions had the same ionic composition but with BaCl 2 replacing CaCl 2 . The CO 2 Ringer solution was the same as normal except that it had 20 mM NaHCO 3 instead of Hepes, was bubbled with 2.5% CO 2 in air and contained 0.1 mM NaH 2 PO 4 .
Ruthenium red was dissolved in 0.1 M KCl at 10 mg.ml -1 [31], and Ru360 was made up at 500 nM in 0.1 M KCl with 0.1% Fast green FCF to make it visible when injected.
Aliquots of 0.1 ml of the solution were kept frozen until use. Ruthenium red injection pressure-injection electrodes often blocked, so Ru360 was preferred in later experiments. Both compounds had the same effects.

Microelectrodes
Conventional micropipettes were pulled from 1. Intracellular CaSMs were made from quartz glass and pHSMs from borosilicate glass [30,31]. For surface pH, both liquid ion sensor and Hinke-style glass pH microelectrodes [32] were used. The latter had the advantage of recording from a large area, and being less likely to penetrate the cell membrane.

Data Collection and Analysis
Potentials from the conventional microelectrodes, CaSMs and pHSMs were recorded as voltages referred to membrane potential, with the voltages from the pHSMs converted to pH before display. Potentials from the surface pHSMs were referred to the bath potential. Results were discarded if on withdrawal of an electrode its potential in Ringer solution had changed by more than 7 mV.
Potentials from the voltage-recording microelectrodes were led via preamplifiers in the Faraday cage to an 8-pole Bessel filter and recorded at 20 Hz on a PC via a CED micro 1401 interface and Spike 2 data collection program (Cambridge Electronic Design, UK). The clamp current was recorded at 100 Hz as above.
Figures were prepared from the CED data after loading into Microsoft Excel. Spikes in the V Ca or pH records generated by electronic pickup were partially erased, and the clamp current records were in some cases restricted in range. Data are presented as means ± S.E.M. of n observations. The statistical significance of observed differences was determined by a paired two-sample two-tailed t test for means. Differences between means were considered significant when P < 0.05.

Ca 2+ uptake by mitochondria or extrusion by the PMCA
The PMCA is continually active, maintaining the normal low [Ca 2+ ] i in the face of a continous influx even in a cell voltage-clamped at -50 mV. To investigate the coupling between Ca 2+ extrusion and H + uptake I have chosen to stimulate the PMCA with a brief, repeated, additional load of Ca 2+ or Ba 2+ ions. I assume that the coupling does not change with rate of activity. A Ca 2+ load can be delivered in several ways. The simplest way is by a brief depolarisation to open Ca 2+ channels, while a second more controllable method is by iontophoretic injection of Ca 2+ ions [32]. 6 In initial experiments I measured the intracellular pH change resulting from the uptake of H + while an injected Ca 2+ load was extruded. With a number of assumptions, including that all the injected Ca 2+ was extruded, I was able to assess possible changes in coupling ratio. This last assumption, however, proved untenable, as was indeed suggested by earlier work with pressure-injected Ca 2+ [33]. My later experiments confirmed that some of the injected Ca 2+ , but not Ca 2+ entering through channels, was taken up by mitochondria, as shown in Fig 1. In this experiment I recorded both pH i and V Ca (the potential from the CaSM referred to the membrane potential) and stimulated the PMCA by both depolarisations and Ca 2+ injections. The first three depolarisations each caused an increase in V Ca of 13 mV (i.e an increase in intracellular Ca 2+ ) and a small fall in pH i . Similarly the two iontophoretic injections of Ca 2+ each caused an increase in V Ca of 5 mV. Although the pH i decrease was larger for the two injections than for the depolarisations, the increase in V Ca for the injections was smaller than for the depolarisations. I then injected Ru360, which blocks Ca 2+ uptake by mitochondria [28].. This greatly increased the V Ca increase seen with an injection, to 19 mV, but did not alter the depolarisation-induced Ca 2+ increases. The pH i responses were not obviously changed. (The injection of Ru360 itself had a small effect on pH i , and a similar but slower effect on V Ca , to those seen with the Ca 2+ injections.) Similar results were obtained with ruthenium red. This suggests that a large fraction of the injected Ca 2+ was taken up by the mitochondria.
Mitochondrial Ca 2+ uptake blockers also increased the surface pH changes induced by depolarisation or injection. Such ∆pH s result from the H + being pumped into the cell as the Ca 2+ is extruded (Schwiening et al 1993, Thomas 2009). The size of the ∆pH s depends on the location of the pH electrode, the local buffering power and the activity of the PMCA. If most of the injected Ca 2+ is taken up by the mitochondria there will be little surface pH increase, as shown in Fig 2. In this experiment the cell was alternately depolarised and injected with Ca 2+ , and both pH s and V Ca measured. About 7 min after the start of the recording shown, ruthenium red was injected sufficient to make the cell clearly pink. (From the size of the bolus of ruthenium red seen at the end of each pressure-injection pulse I estimate that the final concentration was between 0.01 and 0.1 mM.) I then resumed alternating the methods of delivering a Ca 2+ load. The Ru360 had little effect on either the ∆pH s or Ca 2+ increases following depolarisations, but increased both the Ca 2+ transients and, spectacularly, the ∆pH s following Ca 2+ injections. Ru360.
I have also tested cyclopiazonic acid (CPA), which blocks Ca 2+ uptake by the endoplasmic reticulum. It was without effect, after stores were empty, on any of the responses to depolarisation or injection (see also Thomas 2002). It therefore seems  period in normal Ca 2+ Ringer. Ba 2+ is extruded by the PMCA more slowly than Ca 2+ [19], so the surface pH changes were sometimes easier to measure. Since Ba 2+ does not activate K + channels, I was also able to estimate the Ba 2+ influx from the charge needed to depolarise the membrane. In this experiment the depolarisations were of 20, 30, 40 and 50 mV for 5 s in each series, with the first and last series in normal Ca 2+ Ringer. The increasing size of the depolarisations was designed to reveal at 8 what potential the H + channels opened. Again the ∆pH s were smaller in pH 6.5 and 8.5 than in 7.5, but the transient caused by the largest depolarisation in pH 8.5 were contaminated with H + efflux through channels, as shown by the brief surface acidification.
The results from six experiments in low buffered Ca 2+ and five in low buffered Ba 2+ Ringer are compared in Fig 5. In each case I have measured the peak size of the ∆pH s responses to the same size depolarisation (usually 30 mV), averaging several measurements at each pH where possible. I have then calculated the ∆pH s for pH 6.5 and 8.5 as ratios of that seen at pH 7.5, and averaged these ratios for all experiments. The results show that both a pH increase and a decrease from normal apparently reduce the size of the surface pH changes, although not significantly.
One factor that may contribute to this is a reduced Ca 2+ or Ba 2+ entry at pH 6.5, since Ca 2+ channels are well-known to be inhibited by low pH (Moody 1984).
Unfortunately it is not possible to determine the size of the Ca 2+ influx from the clamp current, since Ca 2+ entry activates K + channels. With Ba 2+ entry, however, few K + channels are opened, so the charge carried in by the Ba 2+ is likely to be close to the total clamp current during the depolarisations [19]. I have therefore measured the charge carried in the experiments averaged in Fig 5B. At pH 6.5, 7.5 and 8.5 the average charges were 180 ± 27; 218 ± 32 and 230 ± 33 nC respectively. This suggests that at pH 6.5 the Ca 2+ influx is likely to be about 20% less than at pH 7.5.

The effect of external pH on surface pH and V Ca changes caused by Ca 2+ injection
The quantity of Ca 2+ injected by a given charge is unlikely to be influenced by external pH, so any variation in ∆pH s with pH will probably reflect a change in PMCA rate of activity or coupling between Ca 2+ extrusion and H + uptake. Fig 6A shows part of an experiment in which I recorded both pH s and V Ca responses to a series of Ca 2+ injections made at different external pH values. The cell had been earlier injected with ruthenium red. The peak ∆pH s occurred some seconds after the end of the injection, since the injection site was deep in the cell and the diffusion of ions from the injection site is likely to take some time [29]. The pH s responses at pH 8.5 were clearly smaller than at pH 7.5, and were followed by an acid overshoot. Fig 6B   collects the results from a total of 6 similar experiments. The responses at pH 6.5 and 7.5 were similar, but those at pH 8.5 were much smaller than at pH 7.5.
Part of the effect of high pH described so far probably results from inhibition of the PMCA by the reduction in available H + [4,5]. I have therefore measured the effects of a wide range of external pH values on the rate of recovery of V Ca after depolarisation-induced Ca 2+ loads. Figure 7 shows an example. I analysed the results from this and five other similar experiments by measuring the slope of the V Ca record at a value of V Ca which passed through all the recoveries (V Ca = -115 for the experiment of Fig 7). I have plotted the rates of recovery relative to those seen at pH 7.5 in Fig 8. It is clear that a high external pH inhibits the pump, with the rate at pH 8.5 being on average 69% of that at pH 7.5.

The effect of changes in intracellular pH on pHs changes following Ca 2+ injection.
The effects of pH i on the PMCA coupling ratio do not seem to have been investigated before. To test the effects of pH i on the PMCA H + uptake I either reduced pH i by injecting H + ions iontophoretically or increased pH i by a period of superfusion with CO 2 /bicarbonate buffered Ringer's solution. A representative experiment is illustrated in Figure 9. After three Ca 2+ injections at pH i between 7.7 and 7.4, two injections of H + lowered pH i to about 6.6. At this low pH i Ca 2+ injections caused greatly reduced ∆pH s which were followed by a large pH i overshoot beyond the pre-injection level. As pH i increased, so did the responses to Ca 2+ injection. V Ca was not recorded in this experiment, but in others and as reported earlier [8], low pH i caused an increase in the baseline V Ca. Figure 10 gives the results from four experiments on the effects of different pH i on the ∆pH s caused by Ca 2+ injections. All cells showed that low pH i reduced the pH s responses by a factor of about 2 over a range of 1 pH unit. I have previously shown that low pH i slows the V Ca recovery rate [8]. Allowing for this, the changes in surface pH with pHi were not significant.

Discussion
My results show that depolarisations provide a consistent way of stimulating the PMCA, which generates surface pH increases as H + ions are pumped in. Similarly, Ca 2+ injections are also effective so long as mitochondrial uptake is blocked. For depolarisation-induced surface pH changes I found no significant effect of altering external pH. The Ca 2+ injection-induced surface pH transients were reduced by 55% by high external pH, although much of this can be accounted for by the average V Ca recovery rate being 31% slower than at pH 7.5. These results suggest that the apparent effects of external or intracellular pH on the surface pH changes caused by activating the PMCA were likely to result from changes in the pump rate or measurement error rather than changes in the coupling ratio. Any effects on the coupling ratio were at most much smaller than previously reported. (Since it is not known how many different PMCAs might be active in snail neurones, my results strictly apply only to the total PMCA activity.) The validity of my measurements, and their relevance to the PMCA coupling ratio, depends on a number of assumptions. These concern the constancy of the Ca 2+ load and the extent of its extrusion by the PMCA, the accuracy of its measurement with a CaSM at a point in the cell, and the extent to which the surface pH change is proportional to the H + uptake. I also assume that the PMCA coupling ratio does not change with rate of activity, since this has not been reported for any ATPase. I have made no attempt to measure the coupling ratio itself, being concerned only with its possible alteration.

Possible changes in the Ca 2+ load and its extrusion
I have assumed that repeated depolarisation-induced loads are relatively unaffected by time or pH. From the measurements of the charge carried by Ba 2+ it is likely that low pH s reduces Ca 2+ influx; the sensitivity of Ca 2+ channels to pH has indeed been known for many years [34]. I have also assumed that iontophoretic injections of Ca 2+ by the same charge are not changed by either pH or time.
Following my experiments with ruthenium red and Ru360, I have assumed that the pH s increase generated by a depolarisation-induced Ca 2+ or Ba 2+ load was generated only by the PMCA. Similarly, where a cell has been injected with Ru360 or ruthenium red I assume that most of the injected Ca 2+ is rapidly extruded, and not sequestered.
In previous work with very large pressure injections of CaCl 2 it was concluded [33] that the pH i decreases observed were generated largely by mitochondrial Ca 2+ / H + exchange with some possible contribution from Ca 2+ extrusion. Others have reported that mitochondria are involved in buffering Ca 2+ loads in rat neurones [35].
Previous recordings of intracellular Ba 2+ and Ca 2+ recoveries after injection or loading by depolarisation [8,19] support the assumption that essentially all the unsequestered divalent cation load is pumped out by the PMCA, within 1 to 2 min of the end of the injection or depolarisation. This correlates with the time-course of the ∆pH i . Such an assumption cannot be relied on, however, at low pH i values : my previous recordings of V Ca show that at low pH i values the PMCA does not always return the V Ca to its pre-injection level.
I have shown that while Ca 2+ uptake by the endoplasmic reticulum is little changed by intracellular acidification, release is inhibited [8]. This might change the proportion of injected Ca 2+ that is rapidly pumped out by the PMCA, but I found no effect of CPA, which blocks Ca 2+ uptake by the endoplasmic reticulum. All my measurements on the effects of pH i were made as pH i fell, so that release from the endoplasmic reticulum would be unlikely while I was measuring the injection-induced ∆pH s . I also found that CPA had no effect on the pH or V Ca changes following Ca 2+ loads.

Measurement of the Ca 2+ extrusion rate
To assess the PMCA activity I measured the recovery of V Ca after an injection or a depolarisation. The V Ca was recorded at a single point inside each neurone, close to the cell membrane on the far side of the cell. Thus its recovery can be only an indirect measure of pump rate. Previous experiments with the CaSM method do suggest, however, that the decline in V Ca can be fitted closely by models including intracellular diffusion and pump rate [29].
4.3 pH s changes as a measure of H + uptake.
As well as the PMCA, two other membrane pathways may influence surface pH. The first, pH i regulation, was active only at a low level since bicarbonate was absent during PMCA activation. (I added bicarbonate only to accelerate pH i recovery from large acidifications, as in Fig 9.) The pH s overshoots seen at pH o 8.5 in Fig 6 may have been due to pH i regulation activated by the transiently large H + gradient before the H + diffused away from the interior of the cell membrane. They suggest that residual pH i regulating mechanisms may be driven by a local pH gradient caused by the PMCA. If this was activated rapidly it would reduce the size of the observed ∆pH s . I avoided using inhibitors of pH regulation since they are only slowly reversible and tend to have multiple effects. The second non-PMCA pathway is the H + channel, opened by large depolarisations [26]. At pH o 8.5 I did see evidence of H + channel opening, e.g. in Fig 4 with the 50 mV depolarisation. The H + efflux produced a brief spike in pH s which may well have reduced the total PMCA effect. Measurements showing such signs of H + efflux were excluded from the analysis. From experiments on surface pH reported earlier [36] it is clear that pH s changes produced by very large depolarisations are proportional to H + flux across the cell membrane via H + channels and to external buffering power. Previously such measurements were made at essentially the same pH. Those reported here, however, were done in three solutions of different pH using buffers with pK values of 6.8, 7.5 and 8.4. These gave a relatively constant buffering power between pH 6.5 and 8.5. Thus the same ∆pH s in this range should represent that same flux of H + ions. This makes no allowance for a possible contribution of fixed buffers on the cell surface, about which little is known.
Perhaps they make a significant contribution at low pH values.
If the reduction in PMCA activity at high pH explains in part the reduction in the ∆pH s at pH 8.5, then the increased pump rate at pH 6.5 should theoretically make the ∆pH s much larger. Failure to demonstrate this possibly suggests that the peak rate of the PMCA, which determines the size of the ∆pH s , was limited by availability of ATP or another factor.

Comparison with previous measurements
These findings differ from those reported for the effects of extracellular pH on the coupling ratio of the PMCA of red blood cells and barnacle muscle fibres. With red blood cells it was reported that the number of H + ions pumped in for each Ca 2+ extruded was about 1.9 at pH o 6.3, but that this number fell to zero as pH o was increased to 8.0 [20]. In other words the more H + that were available outside the cell, the more were pumped in. With barnacle muscle similar findings have been reported [21]. At a pH o of 6.5 the authors estimated that 3 H + were taken up for each Ca 2+ extruded, while at pH 8.2 only 1 H + ion was pumped in. No previous study has been published on the effects of pH i on the coupling ratio.
The sensitivity of the PMCA to pH may well be different in neurones. It is also worth noting that my methods were quite different from those used previously, and perhaps involved less disruption to the normal physiology of the cell. Both previous reports measured Ca 2+ efflux with 45 Ca. For the red blood cell studies the H + uptake of a large number of cells was measured by clamping extracellular pH and measuring 13 how much acid must be added as the 45 Ca was pumped out [20]. With barnacle muscle the pH i of a single muscle fibre was followed with a microelectrode, and measured buffering power used to estimate H + uptake [21]. A variety of treatments was used to block all pathways, except the PMCA, that might allow Ca 2+ or H + movement across the cell membrane. External Na + and Ca 2+ were both removed.
My experiments were done with normal Na + and Ca 2+ gradients, on single voltageclamped cells to minimise passive ion fluxes. Cells were unavoidably loaded with Clvia the microelectrodes, but it is hard to see how this would influence the PMCA.
If my assumptions are correct, I conclude that in my preparation the PMCA coupling ratio is relatively insensitive to changes in either intracellular or extracellular pH. This has important implications for the power of the PMCA as a source of intracellular H + , as well as for its molecular mechanism. passed by the voltage clamp electrode to depolarise the cell (+ve current) or inject Ca 2+ iontophoretically (-ve current). The third record shows intracellular pH recorded with a microelectrode. The bottom record shows intracellular Ca 2+ as the potential (V Ca ) recorded by an intracellular Ca 2+ -sensitive microelectrode. Electrical noise from the injection current seen on the pH and V Ca recordings has been partially delected.