Two mechanisms have been proposed to account for solute permeation of lipid bilayers. Partitioning into the hydrophobic phase of the bilayer, followed by diffusion, is accepted by many for the permeation of water and other small neutral solutes, but transient pores have also been proposed to account for both water and ionic solute permeation. These two mechanisms make distinctively different predictions about the permeability coefficient as a function of bilayer thickness. Whereas the solubility-diffusion mechanism predicts only a modest variation related to bilayer thickness, the pore model predicts an exponential relationship. To test these models, we measured the permeability of phospholipid bilayers to protons, potassium ions, water, urea, and glycerol. Bilayers were prepared as liposomes, and thickness was varied systematically by using unsaturated lipids with chain lengths ranging from 14 to 24 carbon atoms. The permeability coefficient of water and neutral polar solutes displayed a modest dependence on bilayer thickness, with an approximately linear fivefold decrease as the carbon number varied from 14 to 24 atoms. In contrast, the permeability to protons and potassium ions decreased sharply by two orders of magnitude between 14 and 18 carbon atoms, and leveled off, when the chain length was further extended to 24 carbon atoms. The results for water and the neutral permeating solutes are best explained by the solubility-diffusion mechanism. The results for protons and potassium ions in shorter-chain lipids are consistent with the transient pore model, but better fit the theoretical line predicted by the solubility-diffusion model at longer chain lengths.

In all living cells, coordination of solute and water movement across cell membranes is of critical importance for osmotic balance. The current concept is that these processes are of distinct biophysical nature. Here we report the expression cloning of a liver cDNA encoding a unique promiscuous solute channel (AQP9) that confers high permeability for both solutes and water. AQP9 mediates passage of a wide variety of non-charged solutes including carbamides, polyols, purines, and pyrimidines in a phloretin-and mercury-sensitive manner, whereas amino acids, cyclic sugars, Na ؉ , K ؉ , Cl ؊ , and deprotonated monocarboxylates are excluded. The properties of AQP9 define a new evolutionary branch of the major intrinsic protein family of aquaporin proteins and describe a previously unknown mechanism by which a large variety of solutes and water can pass through a single pore, enabling rapid cellular uptake or exit of metabolites with minimal osmotic perturbation.Transport of solutes such as ions, nutrients, neurotransmitters, and metabolic waste products across cell membranes is of fundamental importance to all mammalian cells. Despite the identification of many selective solute transporters and water channels (1-4), it has remained unclear how transport of large amounts of solutes is coordinated with water movement in metabolically highly active cells such as hepatocytes, spermatocytes, neurons, and glia. The liver is a major site of production and elimination of metabolites such as urea, nucleotides, and ketone bodies, and substantial amounts of these solutes must rapidly cross the hepatocyte plasma membrane with minimal osmotic perturbation (5). In testis, a solute transport mechanism is presumably required to supply nutrients to rapidly growing spermatocytes and to provide an exit pathway for metabolites. In brain, regulation of solute transport is critical because osmolality changes in extracellular fluids can affect neuronal cell function (6).Among metabolically active tissues, liver was selected as a target for expression cloning of a new solute-transporting protein because a phloretin-sensitive urea exit mechanism had been described (7-9). MATERIALS AND METHODSExpression Cloning-Total RNA was extracted from rats fed a high protein diet (50%, w/w) for 2 weeks. Poly(A) ϩ RNA purified by oligo(dT) chromatography was size-fractionated by preparative agarose gel electrophoresis (30). Specific fractions were screened for 1 mM [14 C]urea uptake activity in RNA-injected Xenopus oocytes (4, 30). A directional cDNA library was constructed from the positive fraction by using the SuperScript Plasmid System (Life Technologies, Inc.), and cDNA clones were screened for urea uptake (4). Northern Analysis and in Situ Hybridization-Poly(A)ϩ RNA (3 g) from rat tissues was electrophoresed in a formaldehyde-agarose gel and transferred to a nylon membrane. The filter was probed with 32 Plabeled full-length AQP9 1 cDNA, hybridized at 42°C, and washed with 0.1% SDS, 0.1ϫ SSC, at 65°C. Autoradiography was performed at Ϫ80°C for 5 days...

Aquaporins (AQP) are members of the major intrinsic protein (MIP) superfamily of integral membrane proteins and facilitate water transport in various eukaryotes and prokaryotes. The archetypal aquaporin AQP1 is a partly glycosylated water-selective channel that is widely expressed in the plasma membranes of several water-permeable epithelial and endothelial cells. Here we report the three-dimensional structure of deglycosylated, human erythrocyte AQP1, determined at 7 A resolution in the membrane plane by electron crystallography of frozen-hydrated two-dimensional crystals. The structure has an inplane, intramolecular 2-fold axis of symmetry located in the hydrophobic core of the bilayer. The AQP1 monomer is composed of six membrane-spanning, tilted alpha-helices. These helices form a barrel that encloses a vestibular region leading to the water-selective channel, which is outlined by densities attributed to the functionally important NPA boxes and their bridges to the surrounding helices. The intramolecular symmetry within the AQP1 molecule represents a new motif for the topology and design of membrane protein channels, and is a simple and elegant solution to the problem of bidirectional transport across the bilayer.

This review summarizes recent progress in water-transporting mechanisms across cell membranes. Modern biophysical concepts of water transport and new measurement strategies are evaluated. A family of water-transporting proteins (water channels, aquaporins) has been identified, consisting of small hydrophobic proteins expressed widely in epithelial and nonepithelial tissues. The functional properties, genetics, and cellular distributions of these proteins are summarized. The majority of molecular-level information about water-transporting mechanisms comes from studies on CHIP28, a 28-kDa glycoprotein that forms tetramers in membranes; each monomer contains six putative helical domains surrounding a central aqueous pathway and functions independently as a water-selective channel. Only mutations in the vasopressin-sensitive water channel have been shown to cause human disease (non-X-linked congenital nephrogenic diabetes insipidus); the physiological significance of other water channels remains unproven. One mercurial-insensitive water channel has been identified, which has the unique feature of multiple overlapping transcriptional units. Systems for expression of water channel proteins are described, including Xenopus oocytes, mammalian and insect cells, and bacteria. Further work should be directed at elucidation of the role of water channels in normal physiology and disease, molecular analysis of regulatory mechanisms, and water channel structure determination at atomic resolution.

CHIP28 is an integral membrane protein that has been identified as the erythrocyte water channel and that is also expressed in the kidney. Antibodies against erythrocyte CHIP28 were used to localize this protein along the rat urinary tubule. By Western blotting, CHIP28 was detected in kidney plasma membrane and endosome fractions. With the use of immunocytochemistry, CHIP28 was located in brush-border and basolateral plasma membranes of the proximal tubule. The initial S1 segment was weakly stained, but the S2 and S3 segments were heavily labeled. Subapical vesicles were also positive. Apical and basolateral membranes of the long thin descending limb were strongly labeled, but ascending thin and thick limbs of Henle and distal convoluted tubules were negative. Some vasa recta profiles in the medulla were positive. CHIP28 is, therefore, present in membranes with a high constitutive water permeability, where it probably acts as a transmembrane water-conducting channel. Finally, a weak staining of apical and basolateral membranes of cortical collecting duct principal cells was detectable, suggesting a potential relationship of CHIP28 to the vasopressin-sensitive water channel.

Fluid movement across epithelia lining portions of the male reproductive tract is important for modulating the luminal environment in which sperm mature and reside, and for increasing sperm concentration. Some regions of the male reproductive tract express aquaporin (AQP) 1 and/or AQP2, but these transmembrane water channels are not detectable in the epididymis. Therefore, we used a specific antibody to map the cellular distribution of another AQP, AQP9 (which is permeable to water and to some solutes), in the male reproductive tract. AQP9 is enriched on the apical (but not basolateral) membrane of nonciliated cells in the efferent duct and principal cells of the epididymis (rat and human) and vas deferens, where it could play a role in fluid reabsorption. Western blotting revealed a strong 30-kDa band in brush-border membrane vesicles isolated from the epididymis. AQP9 is also expressed in epithelial cells of the prostate and coagulating gland where fluid transport across the epithelium is important for secretory activity. However, it was undetectable in the seminal vesicle, suggesting that an alternative fluid transport pathway may be present in this tissue. Intracellular vesicles in epithelial cells along the reproductive tract were generally poorly stained for AQP9. Furthermore, the apical membrane distribution of AQP9 was unaffected by microtubule disruption. These data suggest that AQP9 is a constitutively inserted apical membrane protein and that its cell-surface expression is not acutely regulated by vesicular trafficking. AQP9 was detectable in the epididymis and vas deferens of 1-wk postnatal rats, but its expression was comparable with adult rats only after 3--4 wk. AQP9 could provide a route via which apical fluid and solute transport occurs in several regions of the male reproductive tract. The heterogeneous and segment-specific expression of AQP9 and other aquaporins along the male reproductive tract shown in this and in our previous studies suggests that fluid reabsorption and secretion in these tissues could be locally modulated by physiological regulation of AQP expression and/or function.

Measurements of CO 2 permeability in oocytes and liposomes containing water channel aquaporin-1 (AQP1) have suggested that AQP1 is able to transport both water and CO 2 . We studied the physiological consequences of CO 2 transport by AQP1 by comparing CO 2 permeabilities in erythrocytes and intact lung of wild-type and AQP1 null mice. Erythrocytes from wild-type mice strongly expressed AQP1 protein and had 7-fold greater osmotic water permeability than did erythrocytes from null mice. CO 2 permeability was measured from the rate of intracellular acidification in response to addition of CO 2 /HCO 3 ؊ in a stopped-flow fluorometer using 2,7-bis-(2-carboxyethyl)-5-(and -6)-carboxyfluorescein (BCECF) as a cytoplasmic pH indicator. In erythrocytes from wild-type mice, acidification was rapid (t1 ⁄2 , 7.3 ؎ 0.4 ms, S.E., n ‫؍‬ 11 mice) and blocked by acetazolamide and increasing external pH (to decrease CO 2 /HCO 3 ؊ ratio). Apparent CO 2 permeability (P CO 2 ) was not different in erythrocytes from wild-type (0.012 ؎ 0.0008 cm/s) versus null (0.011 ؎ 0.001 cm/s) mice. Lung CO 2 transport was measured in anesthetized, ventilated mice subjected to a decrease in inspired CO 2 content from 5% to 0%, producing an average decrease in arterial blood pCO 2 from 77 ؎ 4 to 39 ؎ 3 mm Hg (14 mice) with a t1 ⁄2 of 1.4 min. The pCO 2 values and kinetics of decreasing pCO 2 were not different in wild-type versus null mice. Because AQP1 deletion did not affect CO 2 transport in erythrocytes and lung, we re-examined CO 2 permeability in AQP1-reconstituted liposomes containing carbonic anhydrase (CA) and a fluorescent pH indicator. Whereas osmotic water permeability in AQP1-reconstituted liposomes was >100-fold greater than that in control liposomes, apparent P CO 2 (ϳ10 ؊3 cm/s) did not differ. Measurements using different CA concentrations and HgCl 2 indicated that liposome P CO 2 is unstirred layer-limited and that HgCl 2 slows acidification because of inhibition of CA rather than AQP1. These results provide direct evidence against physiologically significant AQP1-mediated CO 2 transport and establish an upper limit to the CO 2 permeability through single AQP1 water channels.A family of related water-transporting proteins (aquaporins, AQP) 1 has been identified in which individual members are expressed in many fluid-transporting epithelia and endothelia. Many of the mammalian aquaporins appear to transport only water, whereas others (AQP3 and AQP7) transport small polar solutes such as glycerol and urea, and even larger solutes including monosaccharides (AQP9) (1-4). Structural information is available for AQP1, the water channel expressed in erythrocytes, kidney tubules and microvessels, alveolar endothelia, choroid plexus, ciliary body, and other tissues. AQP1 molecules associate in tetramers (5) in which the monomeric subunits are functionally independent with respect to water transport (6). Electron crystallography revealed that each AQP1 monomer contains six membrane-spanning, tilted helical domains (7-9); however, the resolut...

(AQP4) water channels exist as heterotetramers of M1 and M23 splice variants and appear to be present in orthogonal arrays of intramembraneous particles (OAPs) visualized by freeze-fracture microscopy. We report that AQP4 forms OAPs in rat gastric parietal cells but not in parietal cells from the mouse or kangaroo rat. Furthermore, the organization of principal cell OAPs in Brattleboro rat kidney is perturbed by vasopressin (arginine vasopressin). Membranes of LLC-PK1 cells expressing M23-AQP4 showed large, abundant OAPs, but none were detectable in cells expressing M1-AQP4. Measurements of osmotic swelling of transfected LLC-PK1 cells using videomicroscopy, gave osmotic water permeability coefficient (Pf) values (in cm/s) of 0.018 (M1-AQP4), 0.019 (M23-AQP4), and 0.003 (control). Quantitative immunoblot and immunofluorescence showed an eightfold greater expression of M1-over M23-AQP4 in the cell lines, suggesting that single-channel pf (cm 3 /s) is much greater for the M23 variant. Somatic fusion of M1-and M23-AQP4 cells (Pf ϭ 0.028 cm/s) yielded OAPs that were fewer and smaller than in M23 cells alone, and M1-to-M23 expression ratios (ϳ1:4) normalized to AQP4 in M1 or M23 cells indicated a reduced single-channel pf for the M23 variant. Expression of an M23-AQP4-Ser 111E mutant produced ϳ1.5-fold greater single-channel pf and OAPs that were up to 2.5-fold larger than wild-type M23-AQP4 OAPs, suggesting that a putative PKA phosphorylation site Ser 111 is involved in OAP formation. We conclude that the higher-order organization of AQP4 in OAPs increases single-channel osmotic water permeability by one order of magnitude and that differential cellular expression levels of the two isoforms could regulate this organization. water transport; freeze-fracture; LLC-PK1 cells; orthogonal arrays; intramembraneous particles THE MAMMALIAN FAMILY OF AQUAPORIN water channels consists of 12 known members, each with a specific tissue distribution and membrane localization pattern. However, the role of aquaporin-4 (AQP4) in water transport physiology is not well understood. AQP4 was the first aquaporin to be observed and identified in biological membranes, because when examined by freeze-fracture electron microscopy, it forms characteristic arrays of intramembraneous particles (IMPs) in the form of checkerboard aggregates or orthogonal arrays of intramembraneous particles (OAPs). These OAPs were described in various cell membranes long before a role in water permeability was suspected (3,18,27,31,32). The relationship of OAPs to water channels was first suggested by earlier studies showing that membranes that contained OAPs, including astrocytes, gastric parietal cells, and collecting duct principal cells were immunostained by an antibody raised against the whole AQP1 (then called CHIP28) protein (37). A different antibody raised against skeletal muscle OAP-containing membranes also recognized an ϳ30-kDa protein in these membranes (15,41). In this way, a protein initially called basolateral intrinsic membrane protein (BLIP) was ...