Thomas Zeuthen
Nordic Centre for Water Imbalance Related Disorders, The Panum Institute, University of Copenhagen, Denmark

https://doi.org/10.36866/pn.68.18

In order to live on land we must closely control and maintain our internal aqueous milieu. Each day the small intestine receives a total of about 10 l of water which originates from glandular secretions plus what is being ingested. This water and its contents of sugars, amino acids, salts etc. have to be absorbed during its passage through the intestine. This is done efficiently – only 0.1 l (1%) is lost to the outside. Clearly, diseases which upset this balance lead to fatal dehydrations. In the kidney the situation is even more dramatic. Each day, 180 l of plasma (four times our total body water!) is being filtered in the nephrons. Fortunately, most of this water is reabsorbed and only about 2 l lost in the urine. At present, there is no adequate explanation of how water is absorbed in the intestine and the kidney. Most models presume the existence of an intraepithelial hyperosmolar compartment, such as in the lateral intercellular spaces. Despite numerous investigations no such hyperosmolarities have been found yet.

Uphill water transport in the small intestine

A major stumbling block for epithelial models has been to explain uphill water transport, i.e. transport against an osmotic gradient. This problem is illustrated most clearly in the small intestine just after a meal consisting of sugars.

In order to be absorbed, sugars are cleaved by local enzymes into monosaccharides such as glucose. In the process, the concentration of glucose becomes much higher and the concentration of water much lower in the lumen than in the blood plasma (Fig. 2, upper panel). Thus glucose could, in principle, be transported passively or downhill into the blood plasma. Water, in contrast, has to move actively or uphill in order to be absorbed. It has been established for more than a century that the small intestine is able to absorb significant amounts of water from a hyperosmolar solution. Indeed, the human small intestine can absorb water from luminal solutions that contain up to 250 mM of glucose in addition to the electrolytes (Pappenheimer, 1998) equivalent to uphill transport of water against an osmotic pressure difference of about five atmospheres. This ability is not unique for the small intestine. Even the very water-permeable epithelium of the kidney proximal tubule transports water uphill, although to a lesser degree. Key questions are therefore:

• why is water not lost into the intestinal lumen but instead transported into the plasma?
• what is energizing the water transport?
• how is the water transport linked to the glucose transport?

Molecular water pumps

To answer these questions, we have suggested that cotransporters function as molecular water pumps. In the human Na+/glucose cotransporter (hSGLT1), for example, a wide range of experiments has shown that 250 water molecules follow the transport of 2 Na+ ions and 1 glucose molecule in a strict stoichiometrical relationship; in the rabbit SGLT1 360 water molecules (Loo et al. 2002). As a result, these cotransporters can utilize the energy contained in the Na+ gradient for the uphill transport of water. The tightness of the coupling between ions and water is emphasized by the fact that it works both ways: a gradient in water chemical potential can drive an uphill flux of ions in the same stoichio-metrical ratio as above.

The idea of molecular water pumps originated from measurements in the choroid plexus epithelium where it was found that transport by the K+/Cl-cotransporter (KCC) was strictly coupled to transport of water (Zeuthen, 1991). Today, the idea is pursued by expression of co-transporters in frog eggs (Xenopus laevis oocytes), a technique that allows for the detection of minute water fluxes (Fig. 1). So far, we have found coupling between substrates and water in all Na+-coupled cotransporters of the symport type. Artefacts such as conventional unstirred layer effects can be excluded due to the relatively high rates of diffusion inside the oocytes. How could the structural features and conformational changes in co-transporters give rise to cotransport of water? One model is an extension of the mobile barrier model first suggested by Mitchell (1990) and explained in Fig. 3. On this model a shift between closed conformations and conformations open to the outside or the inside of the membrane gives rise to a coupling between the transport of water and substrates.

Interestingly, Mitchell did allude to a link between water movements and conformational changes: ‘If, as seems likely, the mobile barrier mechanism involves the opening and closing of a cleft on either side of a substrate-binding domain, one might expect considerable hydrodynamic action as aqueous medium was sucked in and squirted out by the crevices.’ Each step in the cycle suggested in Fig. 3 is well established for aqueous enzymes. Hexokinase, for example, looses and takes up 300 water molecules in the process of phosphorylating a glucose molecule. A clear picture of how the conformational changes are linked to water transport must await structural determination at the atomic level for each of the conformations shown in Fig. 3. So far, it is encouraging that a wide aqueous cavity has been found in the substrate binding conformation of a variety of cotransporters.

Molecular mechanism of water and glucose transport across epithelia

How could molecular water pumps explain the uphill transport of water and the downhill transport of glucose just after a meal (Fig. 2)? Glucose absorption into the cell takes place via two transporters, the SGLT1 and the luminal glucose monoport GLUT2. Interestingly, this GLUT2 is only active in connection with a meal (Kellett, 2001). In this location the GLUT2 is rather tight to water (Zeuthen et al. 2007). Glucose finally leaves the cell and enters the plasma via the ubiquitous, abluminal GLUT2. The water absorption requires energy. The influx of water takes place via the SGLT1 and is energized by the coupling to the influx of Na+, which, in turn, derives its energy from the Na+/K+ATPase. The final step of water transport, from the cell into the plasma, is energized by a coupling to the downhill efflux of K+ and Cl- by the K+/Cl- cotransporter (KCC), which is found abluminally in most epithelia. Again, the energy is provided indirectly by the Na+/K+ATPase. In addition, there is preliminary evidence for a coupling between glucose and water transport in the GLUT2 working in the efflux direction (Zeuthen et al. 2007). The model in Fig. 2 shows only the molecular water pumps; in a given epithelium there are, of course, plenty of other transporters. The kidney proximal tubules, for example, are richly provided with water channels (aquaporins) taking advantage of the downhill gradient for water absorption across this epithelium in vivo.

Cholera and oral rehydration therapy (ORT)

Cholera toxins give rise to major secretions of fluid into the intestinal lumen. The patients experience severe diarrhoea and may loose 3 l or more of water each day. The underlying molecular mechanism is the insertion of a Cl- channel in the luminal membrane. Fortunately, the situation can be countered by the stimulation of the absorptive mechanism outlined in Fig. 2. The two molecular mechanisms, the secretory and the absorptive, are, so to speak, matching each other while the infection is treated. The absorptive component is maximized by ORT: The patient drinks a solution containing 75 mM of NaCl, 75 mM of glucose, and water. This stimulates the uptake of water by the SGLT1 to a degree that compensates for the amount secreted; the uptake of 120 gram of NaCl and 200 gram of glucose would be linked to the uptake of 4 to 5 l of water. ORT is estimated to save more than half the children suffering from severe diarrhoea.

In conclusion, the molecular model of epithelial water transport gives a rational background for the understanding and treatment of life threatening dehydrations.

Acknowledgements

The author wishes to thank Nanna MacAulay, Emil Zeuthen, Svend Christoffersen and Magnus Bundgaard for their help with this manuscript.

References

Kellett GL (2001). The facilitated component of intestinal glucose absorbtion. J Physiol 531, 585-595.

Loo DDF, Wright EM & Zeuthen T (2002). Water Pumps. J Physiol 542, 53-60.

Mitchell P (1990). Osmochemistry of solute translocation. Res Microbiol 141, 286-289.

Pappenheimer JR (1998). Scaling of dimensions of small intestines in non-ruminant eutherian mammals and its significance for absorptive mechanisms. Comp Biochem Physiol 121, 45-58.

Zeuthen T (1991). Secondary active transport of water across ventricular cell membrane of choroid plexus epithelium of necturus maculosus. J Physiol 444, 153-173.

Zeuthen T, Zeuthen E & MacAulay N (2007). Water transport by GLUT2 expressed in Xenopus laevis oocytes. J Physiol 579, 345-361.