The discovery of aquaporin-1 (AQP1) answered the long-standing biophysical question of how water specifically crosses biological membranes. In the kidney, at least seven aquaporins are expressed at distinct sites. AQP1 is extremely abundant in the proximal tubule and descending thin limb and is essential for urinary concentration. AQP2 is exclusively expressed in the principal cells of the connecting tubule and collecting duct and is the predominant vasopressin-regulated water channel. AQP3 and AQP4 are both present in the basolateral plasma membrane of collecting duct principal cells and represent exit pathways for water reabsorbed apically via AQP2. Studies in patients and transgenic mice have demonstrated that both AQP2 and AQP3 are essential for urinary concentration. Three additional aquaporins are present in the kidney. AQP6 is present in intracellular vesicles in collecting duct intercalated cells, and AQP8 is present intracellularly at low abundance in proximal tubules and collecting duct principal cells, but the physiological function of these two channels remains undefined. AQP7 is abundant in the brush border of proximal tubule cells and is likely to be involved in proximal tubule water reabsorption. Body water balance is tightly regulated by vasopressin, and multiple studies now have underscored the essential roles of AQP2 in this. Vasopressin regulates acutely the water permeability of the kidney collecting duct by trafficking of AQP2 from intracellular vesicles to the apical plasma membrane. The long-term adaptational changes in body water balance are controlled in part by regulated changes in AQP2 and AQP3 expression levels. Lack of functional AQP2 is seen in primary forms of diabetes insipidus, and reduced expression and targeting are seen in several diseases associated with urinary concentrating defects such as acquired nephrogenic diabetes insipidus, postobstructive polyuria, as well as acute and chronic renal failure. In contrast, in conditions with water retention such as severe congestive heart failure, pregnancy, and syndrome of inappropriate antidiuretic hormone secretion, both AQP2 expression levels and apical plasma membrane targetting are increased, suggesting a role for AQP2 in the development of water retention. Continued analysis of the aquaporins is providing detailed molecular insight into the fundamental physiology and pathophysiology of water balance and water balance disorders.

Water excretion by the kidney is regulated by the peptide hormone vasopressin. Vasopressin increases the water permeability of the renal collecting duct cells, allowing more water to be reabsorbed from collecting duct urine to blood. Despite long-standing interest in this process, the mechanism of the water permeability increase has remained undetermined. Recently, a molecular water channel (AQP-CD) has been cloned whose expression appears to be limited to the collecting duct. Previously, we immunolocalized this water channel to the apical plasma membrane (APM) and to intracellular vesicles (IVs) of collecting duct cells. Here, we test the hypothesis that vasopressin increases cellular water permeability by inducing exocytosis of AQP-CD-laden vesicles, transferring water channels from IVs to APM. Rat collecting ducts were perfused in vitro to determine water permeability and subcellular distribution of AQP-CD in the same tubules. The collecting ducts were fixed for immunoelectron microscopy before, during, and after exposure to vasopressin. Vasopressin exposure induced increases in water permeability and the absolute labeling density of AQP-CD in the APM. In parallel, the APM:IV labeling ratio increased. Furthermore, in response to vasopressin withdrawal, AQP-CD labeling density in the APM and the APM:IV labeling ratio decreased in parallel to a measured decrease in osmotic water permeability. We conclude that vasopressin increases the water permeability of collecting duct cells by inducing a reversible translocation of AQP-CD water channels from IVs to the APM.Vasopressin (the antidiuretic hormone) is a 9-amino acid peptide hormone, secreted by the neurohypophysis, which acts on the kidney via the adenylyl cyclase-coupled vasopressin receptor (V2 receptor) to regulate water excretion. Vasopressin reduces urinary water excretion in part by increasing the water permeability of the renal collecting duct, thereby accelerating the osmotically driven absorption of water from the collecting duct lumen to the blood. Although the mechanism by which the water permeability increases is controversial, it presumably involves an increase in the number or unit conductance of water channels in the apical plasma membrane (APM), the rate-limiting barrier for net transepithelial water transport (1, 2). Several molecular water channels of the aquaporin (AQP) family are expressed in the kidney (3-5). One of these, AQP-CD (also called WCH-CD or AQP-2), appears to be expressed exclusively in the renal collecting duct (6, 7) and has been documented to be the major pathway responsible for vasopressin-regulated water transport across collecting duct cells (8,9). Several years ago, it was hypothesized (10) that vasopressin induces translocation of water channels from intracellular vesicles (IVs) to the APM by exocytosis (the "shuttle hypothesis"). However, direct experimental evidence for translocation ofwater channels is lacking. We recently demonstrated that the AQP-CD water channel is present in the APM and IVs of collecting d...

Lithium, a widely used treatment for bipolar affective disorders, often. causes nephrogenic diabetes insipidus. The effect of chronic lithium therapy on the expression of the vasopressin-regulated water channel Aquaporin-2 (AQP2) in rat kidney was examined. Membranes were prepared from inner medulla of one kidney from each rat, while the contralateral one was fixed for iminunofluorescence and immunoelectromnicroscopy. Immunoblotting revealed that lithium treatment reduced AQP2 expression dramatically, to 31±8% after 10 d and to 4±1% after 25 d, coincident with development of severe polyuria. Immunofluorescence and immunogold quantitation confirmed the lithium-induced decrease in AQP2 expression (from 11.2±1.0 to 1.1±0.2 particles/,um2). The downregulation was only partly reversed by return to lithium-free diet for 1 wk (40±8% of control). Furthermore, immunoblotting and immunogold quantitation revealed that 2 d of thirsting or 7 d of dDAVP treatment, in the continued presence of lithium, increased AQP2 expression by six-and threefold, respectively, coincident with increased urinary osmolality. Thirsting increased AQP2 immunolabeling mainly of vesicles, whereas dDAVP caused accumulation of AQP2 predominantly in the subapical region and plasma membrane. Thus, lithium causes marked downregulation of AQP2 expression, only partially reversed by cessation of therapy, thirsting or dDAVP treatment, consistent with clinical observations of slow recovery from lithium-induced urinary concentrating defects. (J. Clin. Invest. 1995. 95:1838-1845

The aquaporins are a family of water channels expressed in several water-transporting tissues, including the kidney. We have used a peptide-derived, affinity-purified polyclonal antibody to aquaporin-3 (AQP-3) to investigate its localization and regulation in the kidney. Immunoblotting experiments showed expression in both renal cortex and medulla, with greatest expression in the base of the inner medulla. Subcellular fractionation of membranes, using progressively higher centrifugation speeds, revealed that AQP-3 is present predominantly in the 4,000 and 17,000 g pellets and, in contrast to AQP-2, is virtually absent in the high-speed (200,000 g) pellet that contains small intracellular vesicles. Immunocytochemistry and immunofluorescence studies revealed that labeling is restricted to the cortical, outer medullary, and inner medullary collecting ducts. Within the collecting duct, principal cells were labeled, whereas intercalated cells were unlabeled. Consistent with previous immunofluorescence studies (K. Ishibashi, S. Sasaki, K. Fushimi, S. Uchida, M. Kuwahara, H. Saito, T. Furukawa, K. Nakajima, Y. Yamaguchi, T. Gojobori, and F. Marumo. Proc. Natl. Acad. Sci. USA 91: 6269-6273, 1994; T. Ma, A. Frigeri, H. Hasegawa, and A. S. Verkman. J. Biol. Chem. 269: 21845-21849, 1994), the labeling was confined to the basolateral domain. Immunoelectron microscopy, using the immunogold technique in ultrathin cryosections, demonstrated a predominant labeling of the basolateral plasma membranes. In contrast to previous findings with AQP-2, there was only limited AQP-3 labeling of intracellular vesicles, suggesting that this water channel is not regulated acutely through vesicular trafficking. Immunoblotting studies revealed that thirsting of rats for 48 h approximately doubled the amount of AQP-3 protein in the inner medulla. These studies are consistent with a role for AQP-3 in osmotically driven water absorption across the collecting duct epithelium and suggest that the expression of AQP-3 is regulated on a long-term basis.

Aquaporin-2 (AQP2) is the predominant vasopressin-regulated water channel of the renal collecting duct. We tested whether vasopressin induces translocation of AQP2 from intracellular vesicles into the apical plasma membrane. AQP2 was quantitated in plasma membrane and intracellular vesicle fractions prepared from the inner medulla of one kidney from each rat before or 20 min after intravenous 1-desamino-8-D-arginine vasopressin (DDAVP) treatment, using immunoblotting and densitometry. Contralateral kidneys were prepared for immunofluorescence and immunoelectron microscopy. Immunoblotting revealed that, compared with untreated controls, DDAVP treatment significantly increased the fraction of AQP2 protein associated with the plasma membrane fraction relative to intracellular vesicles. This increase averaged 2.0-fold in untreated rats and 2.9-fold in rats water loaded for 12 h. Water loading, presumably by suppressing circulating vasopressin levels, decreased the fraction of AQP2 associated with the plasma membrane by 55%, suggesting retrieval of AQP2 from the plasma membrane. In rats sequentially thirsted for 48 h to increase expression and then water loaded for 72 h to minimize plasma membrane labeling, DDAVP caused a 12-fold increase in the plasma membrane to intracellular vesicle labeling ratio. The accentuation of the DDAVP response seen after water loading is consistent with the observed increase in the fraction of AQP2 in the intracellular pool available for insertion. Immunofluorescence confirmed a marked DDAVP-induced redistribution of AQP2 from intracellular to plasma membrane domains. Furthermore, quantitative immunoelectron microscopy demonstrated a 3.4-fold increase in apical plasma membrane to intracellular vesicle labeling ratio. These results provide a direct in vivo demonstration of vasopressin-induced translocation of AQP2 into the apical plasma membrane.

The aquaporins are a family of transmembrane proteins that function as molecular water channels. Recently, a mercurial-insensitive water channel [MIWC or aquaporin-4 (AQP4)] has been cloned, and its mRNA was found to be expressed strongly in kidney inner medulla and several nonrenal tissues. We prepared affinity-purified polyclonal antipeptide antibodies to AQP4 to define the regional distribution and cellular location of this water channel within the kidney. Immunoblotting of membrane fractions from different regions of the kidney revealed strongest expression in the base of the renal inner medulla, with detectable levels also in the inner medullary tip, but little or no expression in the outer medulla or cortex. Immunocytochemistry (light microscopy) revealed renal AQP4 labeling exclusively in the collecting duct principal cells, chiefly in the proximal two-thirds of the inner medullary collecting duct (IMCD). Little or no expression was seen in the outer medullary and cortical collecting ducts. Immunoelectron microscopy demonstrated AQP4 labeling of the basolateral membrane of IMCD cells, with relatively little labeling of intracellular vesicles. Differential centrifugation of inner medullary homogenates also revealed a lack of distribution to the vesicle-enriched fraction, which contains the vasopressin-regulated water channel, aquaporin-2. In contrast to aquaporin-2 and aquaporin-3, water restriction of rats did not increase the level of AQP4 expression. These results suggest a possible role for AQP4 in the basolateral exit of water from the IMCD.

Lithium (Li) treatment is often associated with nephrogenic diabetes insipidus (NDI). The changes in whole kidney expression of aquaporin-1 (AQP1), -2, and -3 as well as Na-K-ATPase, type 3 Na/H exchanger (NHE3), type 2 Na-Pi cotransporter (NaPi-2), type 1 bumetanide-sensitive Na-K-2Cl cotransporter (BSC-1), and thiazide-sensitive Na-Cl cotransporter (TSC) were examined in rats treated with Li orally for 4 wk: protocol 1, high doses of Li (high Na(+) intake), and protocol 2, low doses of Li (identical food and normal Na(+) intake in Li-treated and control rats). Both protocols resulted in severe polyuria. Semiquantitative immunoblotting revealed that whole kidney abundance of AQP2 was dramatically reduced to 6% (protocol 1) and 27% (protocol 2) of control levels. In contrast, the abundance of AQP1 was not decreased. Immunoelectron microscopy confirmed the dramatic downregulation of AQP2 and AQP3, whereas AQP4 labeling was not reduced. Li-treated rats had a marked increase in urinary Na(+) excretion in both protocols. However, the expression of several major Na(+) transporters in the proximal tubule, loop of Henle, and distal convoluted tubule was unchanged in protocol 2, whereas in protocol 1 significantly increased NHE3 and BSC-1 expression or reduced NaPi-2 expression was associated with chronic Li treatment. In conclusion, severe downregulation of AQP2 and AQP3 appears to be important for the development of Li-induced polyuria. In contrast, the increased or unchanged expression of NHE3, BSC-1, Na-K-ATPase, and TSC indicates that these Na(+) transporters do not participate in the development of Li-induced polyuria.

Prolonged hypokalemia causes vasopressin-resistant polyuria. We have recently shown that another cause of severe polyuria, chronic lithium therapy, is associated with decreased aquaporin-2 (AQP2) water channel expression (Marples, D., S. Christensen, E.I. Christensen, P.D. Ottosen, and S. Nielsen, 1995. J. Clin. Invest. , 95: 1838-1845). Consequently, we studied the effect in rats of 11 days' potassium deprivation on urine production and AQP2 expression and distribution. Membrane fractions were prepared from one kidney, while the contralateral kidney was perfusion-fixed for immunocytochemistry. Immunoblotting and densitometry revealed a decrease in AQP2 levels to 27 Ϯ 3.4% of control levels ( n ϭ 11, P Ͻ 0.001) in inner medulla, and 34 Ϯ 15% of controls ( n ϭ 5, P Ͻ 0.05) in cortex. Urine production increased in parallel, from 11 Ϯ 1.4 to 30 Ϯ 4.4 ml/day ( n ϭ 11, P Ͻ 0.01). After return to a potassium-containing diet both urine output and AQP2 levels normalized within 7 d. Immunocytochemistry confirmed decreased AQP2 labeling in principal cells of both inner medullary and cortical collecting ducts. AQP2 labeling was predominantly associated with the apical plasma membrane and intracellular vesicles. Lithium treatment for 24 d caused a more extensive reduction of AQP2 levels, to 4 Ϯ 1% of control levels in the inner medulla and 4 Ϯ 2% in cortex, in association with severe polyuria. The similar degree of downregulation in medulla and cortex suggests that interstitial tonicity is not the major factor in the regulation of AQP2 expression. Consistent with this furosemide treatment did not alter AQP2 levels. In summary, hypokalemia, like lithium treatment, results in a decrease in AQP2 expression in rat collecting ducts, in parallel with the development of polyuria, and the degree of downregulation is consistent with the level of polyuria induced, supporting the view that there is a causative link. ( J. Clin. Invest.