Materials and methods

Tissue preparation

Human sweat glands were obtained from slivers of apparently normal skin obtained from adults undergoing surgery who gave informed consent. Ethical committee approval was obtained for this procedure. To prepare single glands from this skin, the shearing technique was used (^{Lee et al, 1984}); small slivers of skin were sheared with steel surgical scissors continuously for about 10 min in a small quantity (about 2 ml) of Dulbecco's phosphate-buffered saline to yield a slurry. This was then diluted further and examined under a dissecting microscope, and individual eccrine glands were identified. From these glands single reabsorptive ducts were dissected using fine steel forceps.

Duct set-up

Experiments were performed in a thermally controlled Perspex chamber (White Instrument, MD), on the stage of an inverted microscope (Olympus OM-2), and ducts were mounted on a conventional microperfusion apparatus (White Instrument). Access of the bathing medium was denied to the apical membrane by crimping the ends of the ducts in glass micropipettes using the technique first described by^{Dellasega and Grantham (1973)} for renal proximal tubules. Ducts were superfused with saline at a rate of 8 ml per min with a bath volume of 200 l. The control saline contained (in mM) NaCl 114; K₂HPO₄ 2.5; MgCl₂ 1; NaHCO₃ 25; glucose 5; Na lactate 4; CaCl₂ 1. This solution was bubbled with 5% CO₂ in oxygen and adjusted to pH 7.4. Changing solutions was rapidly performed with a pneumatically driven minislider valve (Omnifit, Cambridge, U.K.) All experiments were performed at 37°C.

PH_i measurements

Generally, pH_i was measured as previously described for nonperfused renal proximal tubules (^{Macri et al, 1993}). Briefly, ducts were loaded with the AM form of the pH sensitive dye 2'-7'-bis(carboxyethyl)-5(6)-carboxyfluorescein (BCECF) (Molecular Probes, Oregon) for up to 10 min at room temperature. A PTI dual monochromator spectrofluorometer (PTI, Surbiton, U.K.) was used to excite the dye at 450 and 500 nm. The resulting fluorescence was passed through a 535 nm bandpass filter, detected with a photomultiplier tube, and digitized for analysis. The ratio of emitted light from the two excitation wavelengths, after correction for background fluorescence, was calculated using PTI software. Calibration of the dye was performed by equilibration of the duct with solutions adjusted to pH 6.75, 7.0, and 7.25 containing (in mM) NaCl 30; MgCl₂ 1; glucose 5; Na lactate 4; CaCl₂ 2; HEPES 10; KCl 120. The calibration solutions also contained 10 M of the ionophore nigericin to clamp pH_i to extracellular pH (^{Thomas et al, 1979}).

Reagents

BCECF-AM and ethylisopropyl amiloride (EIPA) were obtained from Molecular Probes (Oregon). All other reagents were obtained from Sigma (Dorset, U.K.).

Statistics

All data are expressed as mean SEM. Statistical analysis was performed using Student's t test or ANOVA followed by Dunnett's multiple comparison test as appropriate. Statistical significance was assumed for p <0.05.

Results

Sodium-coupled acid-base transport

Experiments were performed whereby extracellular sodium was reduced from 143 mM to 4 mM by isosmotic replacement of sodium with N-methyl-D-glucamine for a period of 2 min to investigate the presence of sodium-coupled acid-base transport. These experiments were repeated for ducts that had been exposed to 0.5 mM of the anion transport inhibitor 4-acetamido-4'-isothiocyanatostilbene-2,2'-disulfonic acid (SITS) for 5 min and for ducts bathed in a bicarbonate-free medium (bicarbonate replaced with chloride and nominally CO₂ free). In bicarbonate-containing conditions, baseline pH_i was 7.23 0.05 and sodium reduction led to a significant (p <0.02, n = 5) acidification of 0.24 0.06. This observation clearly demonstrates the presence of some form of sodium-coupled acid-base transport on the basolateral membrane of this preparation. As shown in Figure 1 and Table I, neither preincubation with 0.5 mM SITS nor removal of external bicarbonate significantly altered the degree of acidification, suggesting that at least a component, if not all, of this response is bicarbonate independent.

Figure 1.

Effect of reducing bath sodium on pH_i in human sweat ducts. Sodium was lowered during periods indicated by horizontal bars under control conditions, absence of bicarbonate, and presence of 0.5 mM SITS. Each trace represents results from five ducts with mean SEM.

Full figure and legend (20K)

Table I - Summary of data for sodium removal experiments.

Full table

Evidence for basolateral sodium:hydrogen exchange

A sodium-coupled acid-base transporter is present on the basolateral membrane of a wide variety of epithelial cells as a sodium:hydrogen exchanger sensitive to amiloride and its analogs (^{Clark and Limbird, 1991;Tse et al, 1993}). To test for this possibility, sodium removal experiments were performed in bicarbonate-free media but, prior to the readdition of sodium to the bathing medium, ducts were exposed to 10 mol of the amiloride analog EIPA (n = 6; baseline pH_i 7.42 0.06). As shown in Figure 2 addition of EIPA caused a further acidification in addition to that evoked by sodium removal. When sodium was added back to the bathing medium in the presence of EIPA the recovery of pH_i normally observed was almost abolished Figure 2. These observations support the presence of a sodium:hydrogen exchanger on the basolateral membrane of the human eccrine reabsorptive duct.

Figure 2.

Effect of EIPA on recovery of pH_i from sodium readdition in human sweat ducts. Sodium removal and EIPA addition indicated by horizontal bars. Note that EIPA was not present as sodium was removed. Data shown are the mean SEM from six experiments.

Full figure and legend (11K)

Chloride-coupled acid-base transport

When chloride was removed from the external medium and replaced with the impermeant anion cyclamate (gluconate in the case of MgCl₂) there was a statistically significant (p <0.02, n = 5) but very small fall in pH_i of 0.04 0.01 from a baseline pH_i of 7.38 0.02. In some cases this was mildly transient with pH_i being 7.36 0.03 after 2 min of chloride removal. pH_i returned to 7.42 0.02 when chloride was replaced in the bathing medium. This acidification was not SITS sensitive as, in four of the five experiments, when a second chloride-free pulse was applied after a 5 min preincubation with 0.5 mM SITS in the bathing medium, the second pulse caused an acidification of 0.06 0.01.

Electrogenicity of basolateral acid-base transport

The existence of channels conductive for bicarbonate has been proposed for a variety of epithelia (^{Marty et al, 1984;Saito and Wright, 1984;Brown et al, 1989}). In addition, it is conceivable that acid-base transport can be mediated by proton or hydroxyl channels (^{Lyall and Biber, 1994}). To examine whether these possibilities are likely to exist for the basolateral membrane of the sweat duct, maneuvers designed to depolarize the basolateral membrane were performed in the presence of bicarbonate. Although raising external potassium from 5 to 30 mM caused a statistically significant, but very small, alkalosis of 0.04 0.01, which was fully reversible (n = 4, p <0.05, baseline pH_i 7.48 0.07), a 2 min exposure of the basolateral membrane to 1 mM of the potassium channel inhibitor barium had no significant effect on pH_i (n = 4, p >0.1, baseline pH_i 7.45 0.04). In this analysis of the effect of membrane depolarization, however, the chloride removal experiments should also be considered as, although no alkalosis was observed in these experiments (see above), chloride removal has been shown to cause a far greater depolarization than either barium or potassium elevation (^{Reddy and Quinton, 1991}). Thus, it seems that electrogenic acid-base transport does not play a major role under these experimental conditions.

Discussion

Having the ability to acidify sweat to a pH of almost 4.5, the reabsorptive duct of the human sweat gland can generate a transepithelial bicarbonate gradient of almost 1000:1 (assuming a plasma bicarbonate concentration of 25 mM and a calculated sweat bicarbonate concentration of about 0.03 mM). This requires a powerful transport mechanism to acidify the primary sweat in conjunction with high bicarbonate reabsorption rates. With large variations in transepithelial proton fluxes there could also be significant challenges to the maintenance of an intracellular pH compatible with normal cell function. Histochemical evidence for a vacuolar type proton ATPase in the apical membrane of the reabsorptive duct has recently been reported (^{Bovell et al, 2000}), which may well play a key role in the acidification of sweat. To date, however, there has been little work examining functional aspects of acid-base transport in this tissue either for acidification of sweat or for regulation of intracellular pH of cells of the sweat gland.

In this study, we observed the presence of sodium-coupled acid-base transport on the basolateral membrane of the reabsorptive duct. At least a component of this appears to be EIPA sensitive, indicative of a sodium:hydrogen exchanger (^{Clark and Limbird, 1991}). If involved in vectorial acid-base transport across the epithelium, basolateral sodium:hydrogen exchange would act to secrete rather than absorb bicarbonate. Hence the function of this transporter, as in other epithelia, is almost certainly to regulate intracellular pH rather than to contribute directly to bicarbonate reabsorption.

Removal of bath chloride led to a small, SITS-insensitive acidification. The mechanism behind this is unclear but may involve some form of cell-shrinkage-induced inhibition of membrane transport as has been reported for some sodium:hydrogen exchange isoforms (^{Nath et al, 1996}).

Maneuvers designed to depolarize the basolateral membrane were generally ineffective at raising pH_i, suggesting the absence of any major electrogenic acid-base transporter such as a bicarbonate channel, proton/hydroxyl channel, or sodium bicarbonate cotransporter. This is consistent with the report of^{Quinton and Reddy (1993)} that neither apical nor basolateral membranes of the reabsorptive duct are electrically conductive to bicarbonate. In this study neither addition of 1 mM barium nor reduction of bath chloride resulted in an elevation of pH_i, even though both would be expected to cause an intracellular alkalosis if any functional electrogenic transporters existed. On the other hand, elevation of bath potassium did cause a small alkalinization even though this maneuver should be far less effective at depolarizing the cell than either barium addition or chloride removal (^{Reddy and Quinton, 1991}). This may be due to the presence of an electroneutral potassium bicarbonate cotransporter as has been described in rat renal thick ascending limb (^{Leviel et al, 1992}). It is also conceivable that the effect of potassium elevation was a secondary effect of elevated potassium stimulating the sodium pump.

In summary, we have quantitatively estimated pH_i in the human reabsorptive duct to investigate basolateral membrane acid-base transport. Although ion substitution experiments were not supportive of either electrogenic or chloride-coupled transporters having a major role in basolateral acid-base transport, we have demonstrated the existence of sodium-coupled acid-base transport, which is presumably a form of a sodium:hydrogen exchanger whose role is most likely involved in the regulation of intracellular pH.

References

References