Aquaporin-2 inhibitors: fishing in the chemical pool.

The critical contribution of aquaporin-2 (AQP) to water homeostasis in mammals is highlighted by the severe concentrating defects seen in humans with mutations in the AQP2 gene and in mice that are AQP2 null or have loss-of-function mutations in AQP2.1,2 The role of the antidiuretic hormone vasopressin and its receptor, V2R, in the signaling cascade that leads to membrane accumulation of AQP2 in collecting-duct principal cells, and to an increase in renal water reabsorption, is well established.3 The involvement of VP-induced cAMP elevation by adenylyl cyclase activation, and subsequent protein kinase A–induced phosphorylation of AQP2, is also well known. Defects in any of themajor components of this signaling cascade have the potential to lead to aberrant AQP2 trafficking and function, with an associated loss of urine-concentrating ability or an inappropriate increase in collecting-duct water retention. Decreased water reabsorption leads to diabetes insipidus, either central (loss of vasopressin) or nephrogenic (loss of V2R or AQP2 function). Excessive AQP2 expression and plasma membrane localization are associated with such conditions as congestive heart failure, cirrhosis of the liver, and the syndrome of inappropriate antidiuretic hormone secretion. In such cases, inappropriate water re-absorption can lead to moderate or severe hyponatremia and its subsequent complications, sometimes fatal. Furthermore, many defects in water homeostasis result from acquired conditions, such as lithium-induced nephrogenic diabetes insipidus, a common consequence of the use of this compound as therapy for bipolar disorder. Considerable effort has therefore been expended over the past several years to identify components of the AQP2 trafficking pathway that could be targeted by drugs or other therapeutic strategies in order to alleviate the consequences of these diseases. On the basis of current knowledge of the signaling and trafficking pathways that have been elucidated by several groups, different strategies have been explored. These include bypassing the V2R-mediated signaling pathway using alternative cyclic nucleotide-generating agents, such as the cyclic guanosine monophopshate phosphodiesterase inhibitor sildenafil;4,5 activating alternative pathways, including prostaglandin signaling;6 and modulating trafficking with drugs, including statins, that affect the actin cytoskeleton,7,8 a key factor in endoand exocytotic events, as well as inhibiting endocytosis directly.9 Some of these interventions indeed result in increased membrane AQP2 expression, and they increase urinary-concentrating ability in animal models, including vasopressin-deficient Brattleboro rats and lithium-treated rats.2 In the current issue of JASN, Bogum et al.10 use a smallmolecule screening approach to identify chemical compounds that inhibit cAMP-induced AQP2 membrane accumulation. Such compounds would function as aquaretics and counteract the effects of elevated vasopressin/cAMP or increased AQP2 levels in conditions that might lead to excessive fluid reabsorption, hypertension, and hyponatremia. Although this approach is expected to rediscover compounds that are already known to have such an inhibitory effect, the most interesting and exciting outcome of such an unbiased screen would be to uncover new and unexpected active compounds that might not have been predicted with a directed approach to this problem. This could potentially lead to the discovery of new pathways of AQP2 trafficking and regulation, as well as alternative reagents for intervening in the trafficking of AQP2. The screening assay used by Bogum et al. is based on the known effect of forskolin (which maximally increases cAMP levels in cells) to induce translocation of AQP2 from a cytoplasmic vesicular location in epithelial cells to a predominantly plasma membrane localization. In this assay, MCD4 cells of collecting duct origin were used after stably transfecting them with human AQP2. An automated dispensing and washing system was used to immunostain cells for AQP2 after seeding them in 384-well plates and exposing them to forskolin in the presence or absence of almost 18,000 small chemical compounds from a commercially available library. An automated microscope system was used to image the AQP2 in cells from each well, and the distribution of AQP2 on the membrane or in the cytosol was expressed as a software-calculated ratio based on a comparisonwith the cellular localization of actin-phalloidin as a marker that reveals the position of the plasma membrane, and DAPI as a nuclear stain. The screening strategy identified 17 small molecules that inhibit cAMP-induced AQP2membrane localization in transfectedMCD4cells. These compoundswere subsequently tested using inner medullary collecting duct cells expressing endogenous AQP2, and five of them successfully inhibited forskolininduced AQP2 redistribution in these cells as well. One of Published online ahead of print. Publication date available at www.jasn.org.