WATER TRANSPORT IN THE BRAIN: ROLE OF COTRANSPORTERS N. MACAULAY, S. HAMANN AND T. ZEUTHEN* The Panum Institute, Department of Medical Physiology, University of Copenhagen, Blegdamsvej 3C, DK 2200N, Copenhagen, Denmark Abstract—It is generally accepted that cotransporters transport water in addition to their normal substrates, although the pre- cise mechanism is debated; both active and passive modes of transport have been suggested. The magnitude of the water flux mediated by cotransporters may well be significant: both the number of cotransporters per cell and the unit water permeabil- ity are high. For example, the Na  -glutamate cotransporter (EAAT1) has a unit water permeability one tenth of that of aquaporin (AQP) 1. Cotransporters are widely distributed in the brain and participate in several vital functions: inorganic ions are transported by K  –Cl  and Na  –K  –Cl  cotransporters, neurotransmitters are reabsorbed from the synaptic cleft by Na  -dependent cotransporters located on glial cells and neu- rons, and metabolites such as lactate are removed from the extracellular space by means of H  -lactate cotransporters. We have previously determined water transport capacities for these cotransporters in model systems (Xenopus oocytes, cell cul- tures, and in vitro preparations), and will discuss their role in water homeostasis of the astroglial cell under both normo- and pathophysiologal situations. Astroglia is a polarized cell with EAAT localized at the end facing the neuropil while the end abutting the circulation is rich in AQP4. The water transport properties of EAAT suggest a new model for volume homeosta- sis of the extracellular space during neural activity. © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: cotransport, astrocytes, water, glutamate, NKCC1, extracellular space. This review is focused on cotransporters that are ex- pressed in the brain and have been found to transport water: the KCC (K + –Cl - ), the NKCC (Na + –K + –Cl - ), the MCT (H + -lactate), the GAT (Na + -GABA), and the EAAT (Na + -glutamate), for recent reviews see (Zeuthen and Mac- Aulay, 2002a,b). We will review the localization and phys- iological function of these cotransporters and estimate their contribution to the membrane water permeability con- sidering their high level of expression and relatively high unit water permeability. Specifically, we will discuss the role of the unilateral distribution of cotransporters toward the neuropil aspects of the astroglial cell. The transport properties of this part of the membrane have a central role for the function of the synapse, since they control the dimensions and composition of the extracellular space. The water transport properties of the astroglial cell mem- brane facing the neuropil are determined by cotransport- ers, in particular EAAT and NKCC, since no other water transporting proteins have been found here. Water crosses cell membranes by several routes (Fig. 1): across the lipid bilayers, through specific water chan- nel, and via proteins usually associated with other func- tions such as uniports and cotransporters. The precise mechanism of water transport in cotransporters is not en- tirely understood. It is generally accepted that cotransport- ers act as passive water channels but an active mode of transport has also been suggested. A transmembrane os- motic gradient induces passive water transport in cotrans- porters, as would be expected from a conventional water channel. But during cotransport another component of wa- ter transport emerges in parallel and independently of the passive water transport. By this mechanism, water can move uphill against the osmotic gradient since the energy contained in the substrate gradients is transferred to the transport of water. In short, cotransporters act as molecu- lar water pumps. We have investigated active water trans- port extensively in intact tissues by ionselective microelec- trodes and fluorescence methods as well as in Xenopus oocytes, in which relevant cotransporters were overex- pressed. In general, the stoichiometry between the co- transport of water and the non-aqueous substrates is fixed for a given transporter irrespective of transport conditions. Coupling ratios of 150 –500 water molecules per charge translocated by the protein has been determined for differ- ent cotransporters, which means that the tonicity of the transportate may be compatible to that of the surrounding tissues. Cotransport of water has been demonstrated in symports, but not in antiports such as the Na + /H + cotrans- porter. For a recent review, see Zeuthen and MacAulay (2002a). The concept of molecular water pumps is relevant for a number of well-established physiological phenomena, which cannot be explained by simple osmosis. In the small intestine, water is transported uphill from the lumen into the blood; during digestive processes, the lumen can attain hyperosmolarities of more than 100 mosm l -1 relative to plasma (Reid, 1892). Glandular secretion also proceeds against an osmotic gradient; secretion of saliva, for exam- ple, can proceed against hydrostatic pressures of more than 2 m of water (Ludwig, 1861). Finally in the eye, the retinal pigment epithelium transports fluid from the hyper- osmolar retinal space into the blood, a process which may be directly involved in retinal adhesion, for reviews see Hamann (2002) and Zeuthen (2002). Are molecular water pumps relevant for brain as well? The two most important changes in the extracellular space of the neuropil during neural activity is the increase in K + concentration and in osmolarity. The K + homeosta- sis has long been a major focus of research, but it is the *Corresponding author. Tel: +45-3532-7582; fax: +45-3532-7526. E-mail address: tzeuthen@mfi.ku.dk (T. Zeuthen). Abbreviations: EAAT, Na + -glutamate; ECS, extracellular space; GAT, Na + -GABA; KCC, K + –Cl - ; MCT, H + -lactate; NKCC, Na + –K + -Cl - ; RPE, retinal pigment epithelium. Neuroscience 129 (2004) 1031–1044 0306-4522/04$30.00+0.00 © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2004.06.045 1031