Sweating carries with it a potential loss of salt and as NaCl is physiologically important, it must, therefore, be conserved. Functionally, the human eccrine sweat gland has two regions: a secretory coil, which produces an isosmotic primary sweat by the electrogenic movement of Cl^– and Na⁺ ions, coupled osmotically to the movement of water, and a reabsorptive duct, which hypertonically reabsorbs NaCl from the primary sweat. The reabsorption of salt depends upon Na⁺ being pumped from the duct cells into the interstitial fluid by a Na/K-ATPase on the basolateral membranes, while Na⁺ moves passively into the cells via sodium channels sited only in the luminal membrane. ATP-dependent Cl^– (CFTR) channels allow Cl^– ions to follow down their electrochemical gradient. At low flow rates, sweat can be produced physiologically at concentrations as low as 5-15 mM NaCl. However, the electrochemical salt recovery mechanism cannot extract salt from the lumen when luminal Cl^– concentration falls below its electrochemical equilibrium across the apical membrane. Previous findings suggested that an additional mechanism for salt recovery at low flow rates may be driven by H⁺ pumps associated with a Cl ^–-HCO₃ ^– exchanger in the luminal membrane which could lower luminal Cl^– to the levels observed physiologically (Reddy & Quinton, 1994).

The presence of a H⁺ pump has been demonstrated (Bovell et al. 2000), while that of a Cl^–-HCO₃ ^– exchanger has not been established. A family of anion exchangers (AE1, AE2 and AE3) have been cloned that functionally mediate Cl^–-HCO₃ ^– exchange and which are expressed in a variety of tissues. AE2 is expressed in many epithelia including kidney, while AE3 has been detected in the nervous system (bAE3) and cardiac muscle (cAE3), as well as in the gut. This study investigated the presence of an AE in the luminal cells of the human eccrine gland reabsorptive duct.

Frozen sections were obtained from skin biopsies donated, with informed consent and ethics committee approval, by male volunteers (n = 6). Immunocytochemical techniques employed antibodies, raised against AE2 and AE3. Appropriate negative controls were performed using pre-absorbed sera.

In this study, immunoperoxidase labelling demonstrated the presence of anion exchangers and showed a predominant localisation of AE2 at the apical plasma membrane of the ductal cells, suggesting that the AE2 is involved in exchanging Cl^– and HCO₃ ^– ions at slow sweat flow rates through the duct. However, the human sweat gland results differ from those in rat salivary duct cells (Park et al. 1999), where basolaterally located AE2 may, together with a Na⁺ -H⁺ exchanger, control the intracellular pH and contribute to intracellular accumulation of Cl^–. Our results also demonstrate the localisation of AE3 in the cells of the reabsorptive duct, which is surprising. AE3 is normally found in either brain (AE3b) or cardiac tissue (AE3c), although it has been found in gut epithelia (Kudrycki et al. 1990). Further work is clearly required to clarify the role of these exchangers.

The authors wish to acknowledge the American Physiological Society for a Research Career Enhancement Award to D.L.B. and Professor S. Alper for generously providing the AE antibodies.

All procedures accord with current local guidelines and the Declaration of Helsinki.