Body Water Homeostasis: Clinical Disorders of Urinary Dilution and Concentration

Review

Robert W. Schrier Department of Medicine, University of Colorado School of Medicine, Denver, Colorado J Am Soc Nephrol 17: 1820–1832, 2006. doi: 10.1681/ASN.2006030240

he discovery of the aquaporin-1 (AQP1) water channel AQP deletion/mutation are of importance not only for kidney by Agre and colleagues (1,2), which led to the Nobel function but also for other epithelia (14). T Prize in 2003, has revolutionized the understanding of AVP regulates AQP2 in both an acute (short term) (15) and body fluid water regulation by the kidney. Moreover, the iden- chronic (long term) manner (16,17). The short-term regulation tification of other water channels in the kidney, namely AQP2, by AVP involves trafficking of vesicles that contain AQP2 to the 3, and 4, along with urea and ion transporters, has allowed a apical membrane with resultant increased water permeability. much improved understanding of urinary dilution and concen- Suppression of plasma AVP reverses this exocytosis of AQP2, tration in health and disease at the cellular and molecular levels and the vesicles are retrieved from the membrane (endocytosis) (3–8). into the cytoplasm. Both the short- and long-term regulation of The AQP have provided a pathway for water movement AQP2 by AVP is initiated by activation of the V2 receptor on across cellular membranes that could not be explained by sim- the basolateral membrane of the collecting duct. The binding of ple diffusion through the lipid bilayers of cell membranes. AVP to the V2 receptor, which is coupled to a guanine-nucle- AQP1 has been found to be expressed constitutively on both otide–binding protein Gs, results in activation of adenylyl cy- the apical and the basolateral membranes of the proximal tu- clase. This results in an increase in intracellular cAMP concen- bule and descending limb of Henle’s loop. This water channel tration that mediates the activation of protein kinase A (PKA). is not under control of vasopressin but is important in urinary Activated PKA phosphorylates the serine 256 residue at the concentration. Water efflux through these channels in the de- C-terminus of AQP2 protein (18,19). The phosphorylated AQP2 scending limb is an important factor in the countercurrent then is translocated to the collecting duct apical membrane, concentrating mechanism, and diminished maximal urinary thus constituting short-term regulation of AQP2. The long-term osmolality has been shown in AQP1 knockout mice (9) and regulation of AQP2 is mediated by the cAMP response element humans without the AQP1 gene (10). in the 5Ј flanking region of the AQP2 gene in response to AVP AQP2, 3, and 4 are expressed in the cortical and medullary stimulation (20,21). The AQP2 transcription and protein expres- collecting duct (Figure 1) (11). AQP2 is found exclusively in the sion involves the phosphorylation of the cAMP response el- principal cells of the collecting tubule and collecting duct and is ement–binding protein. In contrast to the involvement of PKA known to be regulated by arginine vasopressin (AVP). AQP3 activation in the short-term regulation of AQP2, there is some and 4 are located on the basolateral membrane of the principal in vivo evidence for PKA-independent long-term regulation of cells in the collecting duct. AQP3 knockout mice exhibit sub- AQP2 expression by AVP (22). There also is recent in vitro (23) stantial polyuria secondary to vasopressin-resistant nephro- and in vivo (24) evidence for AQP2 regulation by hyperosmo- genic diabetes insipidus (NDI) (12). AQP3 is regulated by AVP. lality independent of AVP. The in vivo experiments were per- AQP4 predominates on the basolateral membrane of the inner formed in Brattleboro rats, which have no detectable circulatory medulla and is not regulated by AVP. AQP4 knockout mice AVP. also exhibit an NDI that is less severe than that observed in the Disorders of water balance can lead to either clinically rele- AQP3 knockout mice (13). Whereas AQP3 and AQP4 constitute vant hyponatremia or hypernatremia. Hyponatremia is much the exit channels for water movement across the basolateral more common and is the most frequent fluid and electrolyte membrane of the collecting duct, AQP2 is the water channel for disturbance in hospitalized patients. When defined as plasma Ͻ water reabsorption across the apical membrane of the principal sodium concentration 135 mEq/L, the prevalence of hypona- cells of the collecting duct. These transgenic mouse models of tremia in hospitalized patients may be as high as 15 to 30%. Studies indicate that 40 to 75% of these cases are hospital acquired (25,26). It is clear that hyponatremia may occur in the presence of an increase in total body sodium, a decrease in total Published online ahead of print. Publication date available at www.jasn.org. body sodium, or a near-normal total body sodium (Figure 2). In Address correspondence to: Dr. Robert W. Schrier, Department of Medicine, any of these three circumstances, hyponatremia has occurred University of Colorado School of Medicine, 4200 East Ninth Avenue B-173, Denver, CO 80262. Phone: 303-315-8059; Fax: 303-315-2685; E-mail: because of a relatively greater amount of total body water as [email protected] compared with total body solute. Recent advances of the mech-

water excretion. However, these defects generally are not of sufficient magnitude to lead to hyponatremia in association with a normal fluid intake. With normal fluid intake, hyponatremia generally results from the inability to decrease urine osmolality below plasma rather than a defect in maximal solute-free water excretion. The failure to dilute the urine below plasma generally is due to a failure to suppress maximally the antidiuretic hormone AVP (27). The failure of some hypothyroid patients to suppress plasma AVP during an acute water load was reported initially by Skowsky et al. (28). Recently, experimental studies in rats with advanced hypothyroidism (serum total T4 0.6 Ϯ 0.02 ng/dl) further examined urinary dilution and maximal solute-free water excretion (29). There was a significant decrease in cardiac index and heart rate in these hypothyroid animals that was normalized with l- thyroxine replacement. During an acute water load, these hypothyroid animals exhibited a significantly lower urine flow Figure 1. Nephron sites of aquaporin (AQP) water channels and higher urinary osmolality than either control animals or (blue), ion transporters (black), and urea transporters (green). hypothyroid animals that were treated with l-thyroxine. The ROMK, renal outer medullary potassium channel; UT, urea hypothyroid animals also had significantly higher plasma AVP transporter. concentrations despite a lower plasma osmolality that normalized with l-thyroxine treatment. The impaired water excretion anisms of impaired urinary diluting capacity that lead to these also was associated with an upregulation of AVP-mediated hyponatremic disorders are discussed. AQP2 expression and membrane trafficking in the inner medulla (Figure 3). Taken together, these results suggested a major Hypothyroidism role of nonosmotic AVP release in the hyponatremia associated Advanced primary hypothyroidism can lead to hyponatre- with advanced hypothyroidism. Studies therefore were under- mia, most frequently associated with myxedema. Milder forms taken with a vasopressin V2 receptor antagonist (OPC-31260) in of hypothyroidism can lead to a decrease in renal blood flow, these animals with advanced hypothyroidism. Minimal urinary glomerular filtration, and a decrease in maximal solute-free osmolality in hypothyroid rats that were treated with the V2

Figure 2. Diagnostic and therapeutic approach to the hypovolemic, euvolemic, and hypervolemic hyponatremic patient. 1822 Journal of the American Society of Nephrology J Am Soc Nephrol 17: 1820–1832, 2006

Figure 3. Effects in control (CTL) and hypothyroid (HT) and thyroid (T) replaced rats on urine flow (A), urine osmolality (B), plasma vasopressin (AVP; C), and inner medullary AQP2 plasma membrane (PM)/intracellular vesical (ICV) densitometry ratio (D). Reprinted from reference (27), with permission. receptor antagonist was normalized as compared with un- fect in hypothyroidism would be of clinical relevance only in Ͻ treated hypothyroid rats (97 versus 430 mOsm/kg H2O; P patients who were undergoing excessive extrarenal fluid losses 0.0001). Maximal solute-free water excretion in the hypothyroid (e.g., diarrhea). rats significantly increased but did not reach euthyroid values. These experimental results therefore support the pivotal role of Addison’s Disease and Hypopituitarism the baroreceptor-mediated nonosmotic AVP release in the hy- Primary adrenal insufficiency (Addison’s disease) involves ponatremia associated with advanced hypothyroidism. The deficiency of glucocorticoid and mineralocorticoid hormones, submaximal solute-free water excretion would not be of clinical both of which can be associated with hyponatremia. Hypopi- significance unless other factors, such as diuretics or large fluid tuitarism is associated with only glucocorticoid deficiency, be- intakes, intervene. cause the renin-angiotensin-aldosterone system is intact. Al- Of interest, a defect in maximal urinary concentration was though secondary hypothyroidism is associated with observed at a somewhat less prolonged stage of experimental hypopituitarism, in contrast to primary hypothyroidism, the hypothyroidism (30). Maximal urinary osmolality with fluid severity generally is not sufficient to be associated with hypo- deprivation was associated with decreased Na-K-2Cl expres- natremia. Therefore, the hyponatremia that is related to hypop- sion and diminished medullary osmolality. Even though AQP2 ituitarism generally is due to glucocorticoid deficiency rather expression was decreased, the failure to exhibit a difference than thyroid deficiency. between the diminished urinary and medullary osmolality sug- Mineralocorticoid deficiency that is associated with primary gested that the primary mediator of the concentrating defect adrenal insufficiency exhibits a different pathophysiology than was a decrease in the Na-K-2Cl co-transporter expression and glucocorticoid deficiency in causing hyponatremia. Because al- therefore impaired countercurrent concentration rather than dosterone increases potassium and hydrogen ion secretion, the decreased AQP2 expression. This urine-concentrating de- hyperkalemia and nonanion gap metabolic acidosis is charac- J Am Soc Nephrol 17: 1820–1832, 2006 Body Water Homeostasis 1823 teristic of primary but not secondary adrenal insufficiency as a with glucocorticoid deficiency was undertaken in adrenalecto- result of hypopituitarism. The mineralocorticoid deficiency also mized rats that received replacement physiologic concentra- is responsible for the sodium chloride wasting and extracellular tions of mineralocorticoid hormone (35). As compared with fluid volume depletion in primary adrenal insufficiency. Selec- adrenalectomized rats that received replacement of both min- tive mineralocorticoid deficiency has been studied in adrena- eralocorticoid and glucocorticoid hormone, during an acute lectomized animals that received replacement glucocorticoid water load, the glucocorticoid-deficient animals exhibited hormone (31–33). With isolated mineralocorticoid hormone de- higher plasma AVP concentrations, diminished solute-free wa- ficiency and hyponatremia, plasma AVP concentrations were ter excretion, and increased protein expression in the inner not suppressed and collecting duct AQP2 and 3 expressions medulla of AQP2, phosphorylated AQP2, and apical mem- were upregulated. Outer medullary Na-K-2Cl co-transporter brane trafficking of AQP2. The administration of a nonpeptide and Na-K-ATPase were decreased in mineralocorticoid animals vasopressin V2 receptor antagonist reversed these events (Fig- (32). Administration of AVP V2 antagonist in these animals ure 4). significantly improved urinary dilution, but a modest defect in There also was insight into AVP-independent effects on wa- maximal solute-free water excretion remained (33). This latter ter excretion during glucocorticoid deficiency. Pair feeding dur- defect relates to effects of extracellular fluid volume depletion, ing glucocorticoid deficiency was associated with an upregula- ϩ ϩ Ϫ ϩ ϩ including decreased GFR and increased proximal tubular so- tion of the Na -K -2Cl co-transporter, Na /H exchanger dium reabsorption with resultant diminished fluid delivery to isoform 3, and cortical ␤ and ␥ subunits of the epithelial sodium the distal diluting segment of the nephron. Avoidance of neg- channel and sodium retention (35). These events, along with the ative sodium balance in mineralocorticoid-deficient rats nor- effect of a decrease in renal perfusion pressure, would lead to malized Na-K-2Cl co-transporter, Na-K-ATPase, and collecting decreased fluid delivery to the distal nephron diluting seg- duct AQP2 and 3 (32). It also is of interest that high sodium ments and attenuate maximal solute-free water excretion. This chloride intake may correct not only the hyponatremia but also AVP-independent effect on maximal water excretion probably the hyperkalemia and metabolic acidosis in Addison’s disease, would not be of clinical significance except in the circumstance thereby supporting an important role for mineralocorticoid of large increases in water intake. deficiency. Of interest, maximal urinary concentration also was dimin- If hyponatremia persists in primary adrenal insufficiency ished in glucocorticoid-deficient rats after 36 h of fluid depri- despite avoiding negative sodium balance, then the hyponatre- vation (34). The concentrating defect involved primarily an mia is due to glucocorticoid deficiency. The mechanisms of impairment of the countercurrent concentrating mechanism hyponatremia with glucocorticoid deficiency are different than with comparable diminutions in medullary and urinary osmo- with mineralocorticoid deficiency. Glucocorticoid deficiency lalities. With the fluid deprivation, lower expression of the urea does not cause a negative sodium balance, and in fact a positive transporter and ion transporters (e.g., Na-K-2Cl) in the outer sodium balance may occur. The absence of glucocorticoid hor- medulla seemed to account for the diminished medullary os- mone, however, has major effects on systemic hemodynamics, motic gradient. including a decrease in cardiac index with an inadequate response of systemic vascular resistance to maintain mean arte- Primary Polydipsia or Compulsive Water rial pressure (34). These systemic effects result in several con- Drinking sequences. The resultant decrease in stretch on the arterial The renal capacity to excrete solute-free water is substantial baroreceptors in the carotid sinus and aortic arch removes the (27), yet the average daily fluid intake for hyponatremic pa- tonic vagal and glossopharyngeal inhibition on the central re- tients is only 2.4 L (39). The delivery of tubular fluid to the lease of AVP. Elevated plasma AVP concentrations have been distal diluting segment of the nephron is estimated to be ap- demonstrated in glucocorticoid-deficient animals (33,35) and proximately 20% of glomerular filtrate. The diluting segment patients with hypopituitarism (36) despite a degree of hypos- begins with the water-impermeable ascending limb of Henle’s molality that would maximally suppress AVP in normal indi- loop and in the absence of vasopressin includes virtually the viduals. The messenger RNA for AVP in the hypothalamus also remainder of the nephron, i.e., connecting tubule, collecting has been shown to be increased during glucocorticoid defi- tubule, and cortical and medullary collecting ducts. With a GFR ciency (37). AVP-synthesizing neurons terminate in the median of 100 ml/min, 140 L of filtrate occurs daily. If 20% of this eminence of the hypothalamus; therefore, a role for ACTH also filtrate reaches the distal diluting segment and plasma AVP is could be involved in the nonosmotic AVP stimulation that is maximally suppressed, then 28 L of solute-free water theoreti- associated with glucocorticoid deficiency. There also is evi- cally could be excreted. Therefore, the normal capacity to ex- dence of a central effect of glucocorticoid hormone in the hy- crete solute-free water may approximate 1 L/h if administered pothalamus whereby hypo-osmolality does not maximally sup- consistently over 24 h. Patients with primary polydipsia, how- press AVP synthesis during glucocorticoid deficiency (38). In ever, ingest their water intake mostly during their awake hours either case, the importance of this nonosmotic stimulation of of the day. Even so, individuals who have primary polydipsia AVP in glucocorticoid deficiency was documented using pep- and normal renal, endocrine, cardiac, and hepatic function and tide and nonpeptide vasopressin V2 receptor antagonists, are not volume depleted or taking diuretics do not become which profoundly reversed the water retention (33,35). hyponatremic from drinking up to 10 to 12 L of water per day. A recent molecular analysis of the impaired diluting capacity Their plasma osmolalities, however, may be in the lower nor- 1824 Journal of the American Society of Nephrology J Am Soc Nephrol 17: 1820–1832, 2006

Figure 4. Effect of V2 receptor antagonist (OPC) in glucocorticoid-deficient rats to reverse the impaired water excretion (A) and urinary dilution (B) as well as the increased AQP2 (C) and phosphorylated AQP2 expression (D). Reprinted from reference (33), with permission.

mal range (275 to 285 mOsm/kg H2O). If any of the above unique rat model of primary polydipsia in which the same circumstances intervene, then patients with primary polydipsia daily food intake was ingested in control animals with ad are in danger of developing “water intoxication” with severe libitum water intake and polyuric rats that drank 100 ml/d (44). central nervous system symptoms, including headache, nausea, This model therefore allowed assessment of the renal effects of vomiting, decreased mentation, confusion, obtundation, sei- polyuria at comparable electrolyte, caloric, and protein intakes. zures, and even cerebral hernia, cardiopulmonary arrest, and As compared with controls, serum osmolality was lower in the Ͻ death. Pure water intoxication rarely can occur without an polydipsic rats (293 versus 277 mOsm/kg H2O; P 0.04) as was Ͻ intervening event with massive water intake, e.g., drinking urine osmolality (1365 versus 139 mOsm/kg H2O; P 0.001). directly from a faucet or with an open hose in the stomach. As occurs in humans after 10 d of increased water intake (45), Psychiatric patients in hospital frequently have polydipsia, a form of AVP-resistant NDI emerged after 10 d in the poly- because of dry mouth from medications with anticholinergic dipsic rats. With 36 h of fluid deprivation, the polydipsic ani- properties and/or delusions (e.g., “water cleanses the soul”). In mals exhibited significantly higher urine output associated fact, another term that has been used for primary polydipsia with a lower urine osmolality despite significant higher plasma has been psychogenic water drinking. In 1973, acute psychoses AVP concentrations (Figure 5). The ion and urea transporters with hyponatremia mimicking the syndrome of antidiuretic with potential to alter the integrity of the countercurrent con- syndrome first was described (40). Subsequently, hyponatremia centrating mechanism were unaltered, including outer medul- and water intoxication with failure to suppress plasma AVP lary Na-K-2Cl, Na-K-ATPase, and inner medullary urea trans- have been described with psychotic disorders, such as schizo- porter, in polydipsic as compared with control rats. The AQP1 phrenia (41). Psychiatric patients also have a high incidence of water channel expression, which if mutated can cause a con- smoking, and nicotine is a very potent acute stimulus for AVP centrating defect (9,10), also was no different in the polydipsic release (42,43). animals. What was remarkably different was a highly signifi- Recently, molecular and cellular events were analyzed in a cant suppression of AQP2 protein expression in the outer and J Am Soc Nephrol 17: 1820–1832, 2006 Body Water Homeostasis 1825

Figure 5. During 36 h of water deprivation, rats with polydipsia (POLY) demonstrated increased urine output (A), decreased urine osmolality (B), increased plasma AVP (C), and decreased AQP2 expression as compared with control (CTL) rats (D). Reprinted from reference (44), with permission.

inner medulla. The AQP2 expression in the membrane fraction, lation also occur in pregnancy. There is a compensatory rise in as an index of trafficking to the collecting duct membrane, was cardiac output and stimulation of the renin-angiotensin-aldo- decreased. The AQP3 protein abundance in the outer medulla sterone system that attenuates the vasodilation-mediated fall in also was significantly diminished in the polydipsic rats. After BP during the first trimester (47). In other circumstances of the 36-h fluid deprivation, the medullary osmolalities in the arterial vasodilation, such as cirrhosis or a large arteriovenous control and polydipsic rats were no different, even though the fistula, the activation of arterial baroreceptors is accompanied polydipsic rats demonstrated a significant decrease in maximal by the nonosmotic release of AVP (50,51). In pregnancy, plasma urinary osmolality. This failure of osmotic equilibration be- AVP concentrations are still detectable in the presence of a tween the medullary interstitium and urine in the polydipsic degree of hypo-osmolality that normally would suppress AVP rats supported a critical role of the downregulation of AQP2 release, i.e., 1 to 2% decrease in plasma osmolality. The hypo- and, possibly, AQP3 in the impaired urine-concentrating capac- osmolality of pregnancy has been termed a “resetting” of the ity associated with polydipsia. hypothalamic osmoreceptor threshold to a lower level (52) but may be due merely to the nonosmotic release of AVP that Pregnancy occurs with arterial vasodilation (50,51). It also is of interest that

A decrease in plasma osmolality of 8 to 10 mOsm/kg H2Ois early in pregnancy, there is an increase in thirst and fluid intake an early occurrence in normal pregnancy. Sodium and water despite a fall in plasma osmolality, which normally suppresses retention occurs during pregnancy with an expansion of extra- thirst. This also is compatible with a nonosmotic stimulation of cellular fluid ranging from 30 to 50% (46). Water retention thirst as a result of arterial underfilling. exceeds the sodium retention in normal pregnancy and thus the Arterial underfilling as a result of arterial vasodilation, as fall in plasma osmolality and sodium concentration. Recent occurs in cirrhosis and pregnancy, is associated with compen- studies have shown that systemic arterial vasodilation occurs satory increases in total blood volume and cardiac output. early in the first trimester of pregnancy (47). The mediator(s) of Moreover, arterial vasodilation stimulates the nonosmotic AVP this decrease in vascular resistance is(are) not well defined, but release and activates the renin-angiotensin-aldosterone axis, as some experimental results indicate a role of estrogen-mediated occurs in cirrhosis and pregnancy. These hormonal responses nitric oxide (48) and relaxin (49). In any case, the standard also compensate for arterial underfilling, at the expense of responses to arterial underfilling secondary to arterial vasodi- hyponatremia and edema (50,51,53). 1826 Journal of the American Society of Nephrology J Am Soc Nephrol 17: 1820–1832, 2006

With the discovery of the renal water channels, i.e., AQP, be remembered, however, that the antidiuretic effect of oxy- body water regulation in pregnancy could be investigated in toxin also is mediated via the V2 receptor on the basolateral more depth. Pregnancy in the rat exhibits most of the charac- membrane of the collecting duct and therefore could be a teristics of human pregnancy and therefore has been the stan- contributing factor in the water retention of pregnancy (57). It dard experimental model. Studies therefore were undertaken in also has been shown that circulating vasopressinase from the pregnant rats to examine the effect on renal AQP (54). If the placenta, which increases AVP degradation, can uncover or nonosmotic release of AVP secondary to arterial vasodilation is cause diabetes insipidus in pregnancy (58). involved in the water retention in pregnancy (55), then vasopressin-mediated upregulation and trafficking of AQP2 should Cardiac Failure be demonstrable. Alternatively, if pregnancy “resets” the os- Advanced heart failure frequently is associated with hypo- motic threshold for AVP release, then the regulation of the natremia. In fact, hyponatremia has been found to be a risk AQP2 around this reset osmostat should mimic the nonpreg- factor for poor survival for patients with congestive heart fail- nant state. As compared with nonpregnant littermates, preg- ure (59). The pathogenesis of the hyponatremia of cardiac fail- nant rats were found to have in the first trimester a profound ure initially seemed not to involve AVP, primarily because the upregulation of inner medullary AQP2 mRNA and protein bioassay for antidiuretic hormone was relatively insensitive. expression that persisted throughout the pregnancy (Figure 6). With the development of the RIA to measure plasma AVP, In the pregnant rats, there also was an increase in AQP2 in the studies were undertaken to examine the cause of hyponatremia membrane fraction, indicating increased AQP2 trafficking to in patients with heart failure. In the first study, the hypo- the apical membrane of the collecting duct. It is known that the osmolality in patients with heart failure was of a degree that effect of AVP on AQP2 protein expression and trafficking is would maximally suppress plasma AVP in normal individuals, mediated by vasopressin V2 receptor. Studies therefore were yet 30 (81%) of 37 patients had detectable plasma AVP by RIA undertaken with a nonpeptide V2 receptor antagonist in preg- (60). Therefore, the term nonosmotic AVP release emerged to nant and nonpregnant animals. The AQP2 protein expression describe hyponatremic patients with heart failure and other and apical membrane location by immunofluorescence were edematous disorders. There also is evidence for increased AVP returned to the nonpregnant state during the V2 receptor an- synthesis in the hypothalamus in experimental heart failure tagonist administration to pregnant rats (53). These experimen- (61). Hyponatremia may occur both in low-output cardiac fail- tal results therefore strongly support a role of the nonosmotic ure (e.g., ischemic or nonischemic cardiomyopathy) and in AVP release in regulation AQP2 expression and trafficking in high-output cardiac failure (e.g., thyrotoxicosis, beriberi, large pregnancy. This was supported further by the observation that arteriovenous fistula) (62). This apparent dilemma is under- increased urinary AQP2 occurs during pregnancy (56). It must standable because arterial underfilling with baroreceptor-mediated AVP stimulation can occur with either a decrease in cardiac output or systemic arterial vasodilation as occurs with high-output cardiac failure (50,51,53). With the discovery of water channels, investigations were undertaken to examine whether increased plasma AVP in cardiac failure would be associated with increased renal AQP2 protein expression and trafficking to the apical membrane of the collecting duct (63,64). As in humans, plasma AVP increased in advanced cardiac failure in rats despite hypo-osmolality (63). In these animals with heart failure secondary to coronary artery ligation, AQP2 expression and membrane trafficking were increased in the inner medulla of the kidney (63,64). Administration of a nonpeptide V2 receptor antagonist to these animals with heart failure reversed the increased expression and trafficking of AQP2 (63). Support for this relationship between nonosmotic AVP and AQP2 expression has been demonstrated by studies in patients with heart failure by measurement of urinary AQP2 (65). Approximately 3 to 6% of AQP2 can be measured by RIA or Western immunoblotting in the urine. In patients with New York Heart Association class II or III heart failure, a V2 receptor antagonist caused a solute-free water diuresis and increased plasma sodium concentration. Figure 6. Inner medulla AQP2 protein expression (densitometry above and immunoblots below) in nonpregnant (NP) rats and Urinary AQP2 decreased in these patients during the V2 recep- first-trimester (P7), second-trimester (P14), and third-trimester tor antagonist administration (Figure 7) (65), thus indicating (P20) pregnant rats. Glycosylated (36 to 45 kD) and nonglyco- that less of this water channel reached the apical membrane of sylated (29 kD) expression is shown. Reprinted from reference the collecting duct. An orally active, nonpeptide vasopressin V2 (54), with permission. receptor antagonist was shown recently in patients with cardiac J Am Soc Nephrol 17: 1820–1832, 2006 Body Water Homeostasis 1827

ute-free water excretion in patients with cirrhosis (79) and experimental cirrhosis (80). These effects on the nonosmotic stimulation of AVP and AQP2 seem to be manifest with more severe cirrhosis, because milder experimental forms of cirrhosis (e.g., bile duct ligation, inhalation of carbon tetrachloride) did not show these changes (81,82). There are several candidates as mediators of the splanchnic vasodilation and arterial underfilling in cirrhosis. Recent evidence, however, indicates a prominent role of endothelial and inducible nitric oxide synthesis (83). In this regard,7dof treatment with a nonspecific nitric oxide synthase inhibitor at a dose to reverse the arterial vasodilation and thus the hyperdy- namic circulation of experimental cirrhosis was found to suppress plasma AVP, increase solute-free water diuresis, and correct the hyponatremia (84). Figure 7. Effect of V2 antagonist (VPA-985) in patients with chronic heart failure (New York Heart Association II and III) to NDI: Genetic and Acquired decrease urinary AQP2 excretion in a dose-dependent manner NDI is when the kidney does not respond normally to the ϩ Ϫ as compared with placebo during day 1. Day 1 was the antidiuretic effect of AVP. NDI can be due to genetic or ac- control day. T, time in hours. Reprinted from reference (65), quired causes. The cellular and molecular defects that cause the with permission. absolute or relative renal unresponsiveness to vasopressin may involve impairment of the water permeability across the collecting duct, the generation of the medullary osmotic gradient failure to cause a substantial loss of body weight over 30 d of for water transport, or both. treatment (66). Acute reversal of water retention with another nonpeptide V2 antagonist also was shown recently (67). Car- Congenital NDI diac afterload reduction with either hydralazine or an angio- The genetic or congenital NDI can be due primarily to mu- tensin-converting enzyme inhibitor in patients with cardiac tations in two areas. The most common defect relates to muta- failure increased cardiac output, thereby attenuating arterial tions of the vasopressin V2 receptor on the basolateral mem- underfilling. This was associated with an increase in solute-free brane of the collecting duct. More than 180 mutations of the V2 water excretion in association with a decrease in plasma and receptor have been found in chromosome region Xq28 (Figure platelet AVP concentration (68). 8), and together these abnormalities account for 90% of patients

Cirrhosis Hyponatremia frequently occurs in decompensated patients with cirrhosis and ascites (69), whereas patients with compen- sated cirrhosis and no ascites rarely are hyponatremic (70). Hyponatremia also is a risk factor for poor survival in patients with cirrhosis (71). As with heart failure, not until measurement of plasma AVP by RIA was the nonosmotic AVP incriminated in the hyponatremia of cirrhosis (72). Early in cirrhosis, portal hypertension is associated with an increased splanchnic blood flow (73). The associated decrease in systemic vascular resistance causes arterial underfilling and baroreceptor-mediated increase in AVP release (74). An increase in AVP synthesis in the hypothalamus also has been found in experimental cirrhosis (75). The increase in plasma AVP and water excretion in cirrhosis were shown to correlate directly with plasma norepi- nephrine and renin activity (76). This observation provides evidence that cirrhosis activates the sympathetic and renin- Figure 8. Diagram showing the type 2 vasopressin receptor (V2 angiotensin systems, as well as the nonosmotic stimulation of receptor) protein and identifying putative disease-causing mu- AVP, during arterial underfilling as a result of arterial vasodi- tations in the arginine vasopressin receptor 2 gene (circles). lation. Increased AQP2 expression and trafficking also have Mutations are present in the extracellular, transmembrane, and been shown to be present in experimental cirrhosis (77), a cytoplasmic domains of the V2 receptor. The numbers 1 and finding that is compatible with the observed increased urinary 371 refer to amino acids 1 and 371, respectively. COOH, car- excretion of AQP2 in patients with cirrhosis (78). Vasopressin boxy-terminus of the protein; NH2, amino-terminus of the pro- V2 receptor antagonists also have been shown to increase sol- tein. Reprinted from reference (11), with permission. 1828 Journal of the American Society of Nephrology J Am Soc Nephrol 17: 1820–1832, 2006 with congenital NDI (85). Studies suggest that most mutations Acquired NDI involve protein misfolding, which traps the V2 receptor intra- Hypokalemia and hypercalcemia are electrolyte disorders cellularly, thereby not allowing the receptor to translocate to that may be associated with polyuria and polydipsia as a result the basolateral membrane of the collecting duct (86). Therefore, of a defect in the renal response of vasopressin to concentrate the adenylate cyclase-cAMP signaling pathway, which medi- the urine maximally. Experimental studies in rats have been ates AQP2 expression and trafficking to the apical membrane, undertaken to study the mechanisms involved. A potassium- is not activated. Recent in vivo studies indicate that nonpeptide deficient diet for 11 d caused hypokalemic-related NDI (90), V2, V1, and V1/V2 receptor antagonists may allow some of the and7dofvitamin D (dihydrotachysterol) caused hypercalce- mutant V2 receptors to be transported to the basolateral mem- mic-related NDI (91). In both experimental models, there was a brane and to function normally (87). These antagonists act as downregulation of AQP2 expression in the inner medulla. In chaperones for the misfolded V2 receptors to reach the mem- the hypercalcemic model, there also was a decrease in the brane. A recent study in patients with NDI demonstrated an Na-K-2Cl co-transporter in the outer medulla. As the initiator approximate decrease in daily urine flow from 12 to 8 L and an of the countercurrent concentrating mechanism for generation increase in urinary osmolality from 98 to 170 mOsm/kg using of the corticomedullary osmotic gradient, this decrease in the such a receptor antagonist (88). This variety of congenital NDI Na-K-2Cl co-transporter no doubt also contributed to the ac- has an X-linked recessive mode of inheritance, and affected quired NDI. male individuals cannot concentrate their urine in response to Polyuria secondary to NDI also may accompany bilateral vasopressin. Heterozygous female individuals generally are urinary tract obstruction, in part due to the resultant volume asymptomatic but may have a modest degree of polyuria and expansion and urea retention. Experimental studies, however, polydipsia. In contrast, male individuals may have up to 20 L of have shown that unilateral ureteral obstruction, which avoids urine every day. The need to drink sufficient water to maintain these sequelae of bilateral ureteral obstruction, also causes NDI water balance, therefore, is challenging and certainly affects the (92). As with the above electrolyte disorders, AQP2 downregu- quality of life of the individual. Before genetic screening of lation has been demonstrated to be associated with acquired newborns in affected families, dehydration and mental retar- NDI related to urinary tract obstruction. Lithium also has been dation frequently developed. Now affected individuals who demonstrated in experimental animals to be associated with receive adequate water intake progress to normal adulthood downregulation of AQP2 expression and trafficking (93). Lith- without any mental or physical retardation. ium therapy has been a worldwide, effective treatment for Another cause of congenital NDI involves mutations of the bipolar affective disorder for many years. Therefore, under- AQP2 gene in chromosome region 12q13 (89). This accounts for standing the mechanism for the NDI that is associated with this 10% of families with congenital NDI. Currently, more than 30 medication is important, particularly because it occurs in ap- such mutations have been identified (Figure 9). In the autoso- proximately 20% of treated patients. In this regard, recent in mal recessive variety, the mutant AQP2 proteins are trapped in vitro and in vivo experiments suggest that the effect of lithium to the endoplasmic reticulum, whereas the autosomal dominant downregulate AQP2 may occur independent of adenylyl cy- varieties are mutations of the carboxy terminus. clase activity (94,95). A vasopressin-resistant urine-concentrating defect is one of the earliest abnormalities associated with acute renal failure. This variety of acquired NDI also seems to involve an inability to establish a high medullary solute content, i.e., the osmotic driving force for water reabsorption. Moreover, AQP1, 2, and 3 expression has been shown to be reduced in rats with both oliguric and nonoliguric ischemic acute renal failure (96). Sim- ilar results have been found in the 5/6 nephrectomy–induced model of chronic renal failure (97). In patients with advanced chronic renal failure, vasopressin-resistant hypotonic urine has been described (98). This finding suggests a defect in water transport across the medullary collecting duct, because absolute impairment of the countercurrent concentrating mechanism still would be associated with an isotonic, not hypotonic, medullary interstitium. In contrast to congenital NDI, the polyuria of acquired NDI generally is of a moderate degree (e.g.,3 to 4 L/24 h). The thirst mechanism generally protects against hypernatremia in NDI states unless age (newborn, elderly) or Figure 9. Diagram showing the AQP2 protein and identifying illness restricts availability of fluid intake. putative disease-causing mutations in AQP2 (circles). Muta- tions are present in the extracellular, transmembrane, and cytoplasmic domains of AQP2. The numbers 1 and 271 refer to Conclusion amino acids 1 and 271, respectively. Reprinted from reference The ability to analyze the renal water channels and ion and (11), with permission. urea transporters has allowed for the better understanding of J Am Soc Nephrol 17: 1820–1832, 2006 Body Water Homeostasis 1829 several clinical disorders with impaired urinary dilution 16. DiGiovanni S, Nielsen S, Christensen E, Knepper M: Reg- and/or concentration. ulation of collecting duct water channel expression by vasopressin in Brattleboro rats. Proc Natl Acad Sci U S A 91: 8984–8988, 1994 Acknowledgments 17. Wade J, Nielsen S, Coleman R, Knepper M: Long-term This work was supported by National Institutes of Health grant DK regulation of collecting duct water permeability: Freeze- 19928. fracture analysis of isolated perfused tubules. Am J Physiol I appreciate the assistance of Jan Darling in the preparation of the Renal Fluid Electrolyte Physiol 166: F723–F730, 1994 manuscript. Most important, I acknowledge Peter Agre, MD, and his 18. Fushimi K, Sasaki S, Marumo F: Phosphorylation of serine associates, whose discovery of the first mammalian water channel has 256 is required for cAMP-dependent regulatory exocytosis allowed much of this work to be undertaken. of the aquaporin-2 water channel. J Biol Chem 272: 14800– 14804, 1997 19. Katsura T, Gustafson C, Ausiello A, Brown D: Protein References kinase A phosphorylation is involved in regulated exocy- 1. Agre P, Saboori AM, Asimos A, Smith BL: Purification and tosis of aquaporin-2 transfected LLC-PK1 cells. Am J partial characterization of the M 30,000 integral membrane Physiol 272: F817–F822, 1997 protein associated with the erythrocyte Rh (D) antigen. 20. Hozawa S, Holtzman E, Ausiello D: cAMP motifs regulat- J Biol Chem 262: 17497–17503, 1987 ing transcription in the aquaporin-2 gene. Am J Physiol 270: 2. Agre P, Preston GM, Smith BL, Jung JS, Raina S, Moon C, C1695–C1702, 1996 Guggino WB, Nielsen S: Aquaporin CHIP: The archetypal 21. Matsumura Y, Uchida S, Rai T, Sasaki S, Marumo F: Tran- molecular water channel. Am J Physiol Renal Fluid Electro- scriptional regulation of aquaporin-2 water channel gene lyte Physiol 265: F463–F476, 1993 by cAMP. J Am Soc Nephrol 8: 861–867, 1997 3. Nielsen S, Frokiaer J, Marples D, Kwon T-H, Agre P, Knep- 22. Umenishi F, Narikiyo T, Vandewalle A, Schrier RW: cAMP per M: Aquaporins in the kidney: From molecules to med- regulates vasopressin-induced AQP-2 expression via pro- icine. Physiol Rev 82: 205–244, 2002 tein kinase A-independent pathway. BBA Biomembranes 4. Schrier RW, Cadnapaphornchai MA: Renal aquaporin wa- 2006, in press ter channels: From molecules to human disease. Prog Bio- 23. Umenishi F, Narikiyo T, Schrier RW: Effect on stability, phys Mol Biol 81: 117–131, 2003 degradation, expression, and targeting of aquaporin-2 wa- 5. Schrier RW, Chen Y-C, Cadnapaphornchai MA: From finch ter channel by hyperosmolality in renal epithelial cells. to fish to man: Role of aquaporins in body fluid and brain Biochm Biophys Res Commun 338: 1593–1599, 2005 water regulation. Neuroscience 129: 897–904, 2004 24. Li C, Wang W, Summer S, Cadnapaphornchai MA, Falk S, 6. Chen Y-C, Cadnapaphornchai MA, Schrier RW: Clinical Umenishi F, Schrier RW: Hyperosmolality in vivo upregu- update on renal aquaporins. Biol Cell 97: 357–371, 2005 lates aquaporin 2 water channel and Na-K-2Cl cotrans- 7. Ishikawa S, Schrier RW: Pathophysiological roles of argi- porter in Brattleboro rats. J Am Soc Nephrol 17: 1657–1664, nine vasopressin and aquaporin-2 in impaired water ex- 2006 cretion. Clin Endocrinol 58: 1–17, 2003 25. Anderson R, Chung H, Kluge R, Schrier RW: Hyponatre- 8. Sands J: Mammalian urea transporters. Annu Rev Physiol mia: A prospective analysis of its epidemiology and the 65: 543–566, 2003 pathogenetic role of vasopressin. Ann Intern Med 102: 164– 9. Ma T, Yang B, Gillespie A, Carlson E, Epstein GC, Verk- 168, 1985 man A: Severely impaired urinary concentrating ability in 26. Hoorn EJ, Lindemans J, Zietse R: Development of severe transgenic mice lacking aquaporin-1 water channels. J Biol hyponatraemia in hospitalized patients: Treatment-related Chem 273: 4296–4299, 1998 risk factors and inadequate management. Nephrol Dial 10. King LS, Choi M, Ferenandez PC, Cartron J-P, Agre P: Transplant 21: 70–76, 2006 Defective urinary concentrating ability due to a complete 27. Berl T, Schrier RW: Disorders of water metabolism. In: deficiency of aquaporin-1. N Engl J Med 345: 175–179, 2001 Renal and Electrolyte Disorders, 6th Ed., edited by Schrier 11. Sands J, Bichet D: Nephrogenic diabetes insipidus. Ann RW, Philadelphia, Lippincott Williams & Wilkins, 2002, pp Intern Med 144: 186–194, 2006 1–63 12. Ma T, Song Y, Yang B, Gillespie A, Carlson E, Epstein C, 28. Skowsky W, Kikuchi T: The role of vasopressin in the Verkman A: Nephrogenic diabetes insipidus in mice lack- impaired water excretion of myxedema. Am J Med 64: ing aquaporin-3 water channels. Proc Natl Acad Sci U S A 613–621, 1978 97: 4386–4391, 2000 29. Chen Y-C, Cadnapaphornchai MA, Yang J, Summer S, Falk 13. Ma T, Yang B, Gillespie A, Carlson E, Epstein C, Verkman S, Li C, Wang W, Schrier RW: Nonosmotic release of va- A: Generation and phenotype of a transgenic knockout sopressin and renal aquaporins in impaired urinary dilu- mouse lacking the mercurial-insensitive water channel tion in hypothyroidism. Am J Physiol Renal Physiol 289: aquaporin-4. J Clin Invest 100: 957–962, 1997 F672–F678, 2005 14. Verkman AS: Novel roles of aquaporins revealed by phe- 30. Cadnapaphornchai MA, Kim Y-W, Gurevich A, Summer S, notype analysis of knockout mice. Rev Physiol Biochem Falk S, Thurman J, Schrier RW: Urinary concentrating de- Pharmacol 155: 31–55, 2005 fect in hypothyroid rats: Role of sodium, potassium, 15. Nielsen S, Chou C, Marples D, Christensen E, Knepper M, 2-chloride co-transporter and aquaporins. J Am Soc Nephrol Harris H: Cellular and subcellular immunolocalization of 14: 566–574, 2003 vasopressin-regulated water channel in rat kidney. Proc 31. Ufferman R, Schrier RW: Importance of sodium intake and Natl Acad Sci U S A 90: 11663–11667, 1993 mineralocorticoid hormone in the impaired water excre- 1830 Journal of the American Society of Nephrology J Am Soc Nephrol 17: 1820–1832, 2006

tion in adrenal insufficiency. J Clin Invest 51: 1639–1646, Rogachev B, Ladtkow T, Carter E, Schrier RW: Chronic 1972 nitric oxide synthase inhibition reverses systemic vasodi- 32. Ohara M, Cadnapaphornchai MA, Summer S, Falk S, Yang lation and glomerular hyperfiltration in pregnancy. Am J J, Togawa T, Schrier RW: Effect of mineralocorticoid defi- Physiol Renal Physiol 280: F592–F598, 2001 ciency on ion and urea transporters and aquaporin water 49. Conrad K, Jeyabalen A, Danielson L, Kerchner L, Novak J: channels in the rat. Biochem Biophys Res Commun 299: 285– Role of relaxin in maternal renal vasodilation of preg- 290, 2002 nancy. Ann N Y Acad Sci 1041: 147–154, 2005 33. Ishikawa S, Schrier RW: Effect of arginine vasopressin 50. Schrier RW: Pathogenesis of sodium and water retention in antagonist on renal water excretion in glucocorticoid and high and low output cardiac failure, cirrhosis, nephrotic mineralocorticoid deficient rats. Kidney Int 22: 587–593, syndrome, and pregnancy (1). N Engl J Med 319: 1065–1072, 1982 1988; published erratum appears in N Engl J Med 320: 676, 34. Chen Y-C, Cadnapaphornchai MA, Summer S, Falk S, Li C, 1989 Wang W, Schrier RW: Molecular mechanisms of impaired 51. Schrier RW: Pathogenesis of sodium and water retention in urinary concentrating ability in glucocorticoid-deficient high and low output cardiac failure, cirrhosis, nephrotic rats. J Am Soc Nephrol 16: 2864–2871, 2005 syndrome, and pregnancy (2). N Engl J Med 319: 319: 35. Wang W, Li C, Summer S, Falk S, Cadnapaphornchai MA, 1127–1134, 1988; published erratum appears in N Engl Chen Y-C, Schrier RW: Molecular analysis of impaired J Med 320: 676, 1989 urinary diluting capacity in glucocorticoid deficiency. Am J 52. Durr J, Stamoutsos B, Lindheimer M: Osmoregulation dur- Physiol Renal Physiol 290: F1135–F1142, 2006 ing pregnancy in the rat. Evidence for resetting of the 36. Celkjers W: Hyponatremia and inappropriate secretion of threshold for vasopressin secretion during gestation. J Clin vasopressin (antidiuretic hormone) in patients with hy- Invest 68: 337–346, 1981 popituitarism. N Engl J Med 321: 492–496, 1989 53. Schrier RW: Body fluid volume regulation in health and 37. Pyo H, Summer S, Kim J, Schrier RW: Vasopressin gene disease: A unifying hypothesis. Ann Intern Med 113: 155– expression in glucocorticoid hormone-deficient rats. Ann 159, 1990 N Y Acad Sci 689: 659–662, 1993 54. Ohara M, Martin P-Y, Xu D-L, St. John J, Pattison T, Kim J, 38. Kim J, Summer S, Wood W, Schrier RW: Role of glucocor- Schrier RW: Upregulation of aquaporin 2 water channel ticoid hormones in arginine vasopressin gene regulation. expression in pregnant rats. J Clin Invest 101: 1076–1083, Biochem Biophys Res Commun 289: 1252–1256, 2001 1998 39. Gross P, Pehrisch H, Rascher W, Schomig A, Hackenthal E, 55. Schrier RW, Briner VA: Peripheral arterial vasodilation Ritz E: Pathogenesis of clinical hyponatremia: Observation hypothesis of sodium and water retention in pregnancy: of vasopressin and fluid intake in 100 hyponatremic med- Implications for pathogenesis of preeclampsia-eclampsia ical patients. Eur J Clin Invest 17: 123–129, 1987 state. Obstet Gynecol 77: 632–639, 1991 40. Dubovsky S, Grabon S, Berl T, Schrier RW: Syndrome of 56. Buemi M, D’Anna R, DiPasquale G, Floccari F, Ruello A, inappropriate secretion of antidiuretic hormone with exac- Aloisi C, Leonardi I, Frisina N, Corica F: Urinary excretion erbated psychosis. Ann Intern Med 79: 551–554, 1973 of aquaporin-2 water channel during pregnancy. Cell 41. deLeon J, Verghese C, Tracy J, Josiassen R, Simpson G: Physiol Biochem 11: 203–208, 2001 Polydipsia and water intoxication in psychiatric patients: A 57. Chou C, DiGiovanni S, Luther A, Knepper M: Oxytocin as review of the epidemiological literature. Biol Psychiatry 35: an antidiuretic hormone. II. Role of V2 vasopressin recep- 408–419, 1994 tor. Am J Physiol Renal Fluid Electrolyte Physiol 269: F78–F85, 42. Blum A: The possible role of tobacco cigarette smoking in 1995 hyponatremia of long-term psychiatric patients. JAMA 252: 58. Durr J, Hoggard J, Hunt J, Schrier RW: Diabetes insipidus 2864–2865, 1984 due to abnormally high circulating vasopressinase activity 43. Allon M, Allen H, Deck L, Clark M: Role of cigarette use in in a pregnancy. N Engl J Med 316: 1070–1074, 1987 hyponatremia in schizophrenic patients. Am J Psychiatry 59. Lee W, Packer M: Prognostic importance of serum concen- 147: 1075–1077, 1990 tration and its modification by converting-enzyme inhibi- 44. Cadnapaphornchai MA, Summer S, Falk S, Thurman J, tion in patients with severe chronic heart failure. Circula- Knepper M, Schrier RW: Effect of primary polydipsia on tion 73: 257–267, 1986 aquaporin and sodium transporter abundance. Am J 60. Szatalowicz V, Arnold P, Chaimovitz C, Bichet D, Schrier Physiol Renal Physiol 285: F965–F971, 2003 RW: Radioimmunoassay of plasma arginine vasopressin in 45. DeWardener H, Herxheimer A: The effect of a high water hyponatremic patients with congestive heart failure. intake on the kidney’s ability to concentrate urine in man. N Engl J Med 305: 263–266, 1981 J Physiol 139: 42–52, 1957 61. Kim J, Michel J, Soubrier F, Durr J, Corvol P, Schrier RW: 46. Briner V, Cadnapaphornchai M, Schrier RW: Hypertension Arginine vasopressin gene expression in chronic cardiac and pregnancy. In: Diseases of the Kidney and Urinary Tract, failure in rats. Kidney Int 38: 818–822, 1990 8th Ed., edited by Schrier RW, Philadelphia, Lippincott 62. Schrier RW, Abraham WT: Hormones and hemodynamics Williams & Wilkins, 2006, in press in heart failure. N Engl J Med 341: 577–585, 1999 47. Chapman AB, Abraham WT, Zamudio S, Coffin C, Mer- 63. Xu D, Martin P-Y, Ohara M, St. John J, Pattison T, Meng X, ouani A, Young D, Johnson A, Osorio F, Goldberg C, Morris K, Kim J, Schrier RW: Upregulation of aquaporin-2 Moore LG, Dahms T, Schrier RW: Temporal relationships water channel expression in chronic heart failure. J Clin between hormonal and hemodynamic changes in early Invest 99: 1500–1505, 1997 human pregnancy. Kidney Int 54: 2056–2063, 1998 64. Nielsen S, Terris J, Andersen D, Ecelbarger C, Frokiaer J, 48. Cadnapaphornchai MA, Ohara M, Morris K, Knotek M, Jonassen T, Marples D, Knepper M, Petersen J: Congestive J Am Soc Nephrol 17: 1820–1832, 2006 Body Water Homeostasis 1831

heart failure in rats is associated with increased expression 79. Inoue T, Ohnishi A, Matsuo A, Kawai B, Kunihiro N, Tada and targeting of aquaporin-2 water channel in collecting Y, Koizumi F, Chau T, Okada Y, Yamamura Y, Tanaka T: duct. Proc Natl Acad Sci U S A 94: 5450–5455, 1997 Therapeutic and diagnostic potential of a vasopressin-2 65. Martin P-Y, Abraham W, Lieming X, Olson B, Oren R, antagonist for impaired water handling in cirrhosis. Clin Ohara M, Schrier RW: Selective V2-receptor vasopressin Pharmacol Ther 63: 561–570, 1998 antagonism decrease urinary aquaporin-2 excretion in pa- 80. Claria J, Jiminez W, Arroyo V, Guarner F, Lopez C, La Villa tients with chronic heart failure. J Am Soc Nephrol 10: G, Asbert M, Rivera F, Rodes J: Blockade of the hydroos- 2165–2170, 1999 motic effect of vasopressin normalizes water excretion in 66. Gheorghiade M, Niazi I, Ouyang J, Czerwiec F, Kamba- cirrhotic rats. Gastroenterology 97: 1294–1299, 1989 yashi J-I, Xampino M, Orlandi C: Vasopressin V2-receptor 81. Fernandez-Llama P, Turner R, Dibona G, Knepper M: Re- blockade with Tolvaptan in patients with chronic heart nal expression of aquaporins in liver cirrhosis induced by failure. Results from a double-blind, randomized trial. Cir- chronic common bile duct ligation in rats. J Am Soc Nephrol culation 107: 2690–2696, 2003 10: 195–1957, 1999 67. Abraham W, Shamshirsaz A, McFann K, Oren R, Schrier 82. Fernandez-Llama P, Jimenez W, Bosch-Marce M, Arroyo RW: Aquaretic effect of lixivaptan, an oral, non-peptide, V, Nielsen S, Knepper M: Dysregulation of renal aquapor- selective V2 receptor vasopressin antagonist, in New York ins and Na-Cl cotransporter in CC14-induced cirrhosis. Heart Association functional class II and III heart failure Kidney Int 58: 216–228, 2000 patients. J Am Coll Cardiol 47: 1615–1621, 2006 83. Martin PY, Gines P, Schrier RW: Nitric oxide as a mediator 68. Bichet D, Kortas C, Mettauer B, Manzini C, Marc-Aurele J, of hemodynamic abnormalities and sodium and water re- Rouleau J, Schrier RW: Modulation of plasma and platelet tention in cirrhosis. N Engl J Med 33: 533–541, 1998 vasopressin by cardiac function in patients with heart fail- 84. Martin PY, Ohara M, Gines P, Xu D, St. John J, Nieder- ure. Kidney Int 29: 1188–1196, 1986 berger M, Schrier RW: Nitric oxide synthase (NOS) inhibi- 69. Gines P, Berl T, Bernardi M, Bichet D, Hamon G, Jimenez tion for one week improves renal sodium and water excre- W, Liard J, Martin PY, Schrier RW: Hyponatremia in cir- tion in cirrhotic rats with ascites. J Clin Invest 101: 235–242, rhosis: From pathogenesis to treatment. Hepatology 28: 851– 1998 864, 1988 85. Morello J, Bichet D: Nephrogenic diabetes insipidus. Annu 70. Madsen M, Pedersen E, Danielsen H, Jensen L, Sorensen S: Rev Physiol 63: 607–630, 2001 Impaired renal water excretion in early hepatic cirrhosis. 86. Arid M, Balch W: Integration of endoplasmic reticulum Lack of relationship between renal water excretion and signaling in health and disease. Nat Med 5: 745–751, 1999 plasma levels of arginine vasopressin, angiotensin II, and 87. Fujiwara T, Bichet D: Molecular biology of hereditary di- aldosterone after water loading. Scand J Gastroenterol 21: abetes insipidus. J Am Soc Nephrol 16: 2836–2846, 2005 749–755, 1986 88. Bernier V, Morello J-P, Zarruk A, Debrand N, Salahpour A, 71. Gines A, Escorsell A, Gines P, Salo J, Jimenez W, Inglada L, Lonergan M, Arthus M-F, Laperriere A, Brouard R, Bou- Navasa M, Claria J, Rimola A, Arroyo V, et al.: Incidence, vier M, Bichet D: Pharmacologic chaperones as a potential predictive factors, and prognosis of the hepatorenal syn- treatment for x-linked nephrogenic diabetes insipidus. drome in cirrhosis with ascites. Gastroenterology 105: 229– J Am Soc Nephrol 17: 232–243, 2006 236, 1993 89. Deen P, Croes H, van Aubel R, Ginsel L, van Os C: Water 72. Bichet D, Szatalowicz V, Chaimovitz C, Schrier RW: Role of channels encoded by mutant aquaporin-2 genes in neph- vasopressin in abnormal water excretion in cirrhotic pa- rogenic diabetes insipidus are impaired in their cellular tients. Ann Intern Med 96: 413–417, 1982 routing. J Clin Invest 95: 2291–2296, 1995 73. Bosch J, Garcia-Pagan J: The splanchnic circulation in cir- 90. Marples D, Frokiaer J, Dorup J, Knepper M, Nielsen S: rhosis. In: Ascites and Renal Dysfunction in Liver Disease, 2nd Hypokalemia-induced downregulation of aquaporin-2 wa- Ed., edited by Gines P, Arroyo V, Rodes J, Schrier RW, ter channel expression in rat kidney medulla and cortex. Oxford, Blackwell Publishing, 2005, pp 156–163 J Clin Invest 97: 1960–1968, 1996 74. Schrier RW, Arroyo V, Bernardi M, Epstein M, Henricksen 91. Earm J, Christensen B, Frokiaer J, Marples D, Han J, Knep- J, Rodes J: Peripheral arterial vasodilation hypothesis: A per M, Nielsen S: Decreased aquaporin-2 expression and proposal for the initiation of renal sodium and water re- apical plasma membrane delivery in kidney collecting tention in cirrhosis. Hepatology 8: 1151–1157, 1988 ducts of polyuric hypercalcemic rats. J Am Soc Nephrol 9: 75. Kim J, Summer S, Howard R, Schrier RW: Vasopressin 2181–2193, 2998 gene expression in rats with experimental cirrhosis. Hepa- 92. Frokiaer J, Christensen B, Marples D, Djurhuus J, Jensen U, tology 17: 143–147, 1993 Knepper M, Nielsen S: Downregulation of aquaporin-2 76. Bichet D, Van Putten V, Schrier RW: Potential role of parallels changes in renal water excretion in unilateral increased sympathetic activity in impaired sodium and ureteral obstruction. Am J Physiol Renal Physiol 273: F213– water excretion in cirrhosis. N Engl J Med 397: 1552, 1982 F223, 1997 77. Fujita N, Ishikawa S, Sasaki S, Fujisawa G, Fushimi K, 93. Marples D, Christensen S, Christensen E, Ottosen P, Marumo F, Saito T: Role of water channel AQP-CD in Nielsen S: Lithium-induced downregulation of aqua- water retention in SIADH and cirrhotic rats. Am J Physiol porin-2 water channel expression in rat kidney medulla. 269: F926–F931, 1995 J Clin Invest 95: 1838–1845, 1995 78. Ivarsen P, Frokiaer J, Aagaard N, Hansen E, Bendtsen F, 94. Bichet DG: Lithium, cyclic AMP signaling, A-kinase an- Nielsen S, Vilstrup H: Increased urinary excretion of aqua- choring proteins, and aquaporin-2. J Am Soc Nephrol 17: porin 2 in patients with liver cirrhosis. Gut 52: 1194–1199, 920–922, 2006 2003 95. Li Y, Shaw S, Kamsteeg E-J, Vandewalle A, Deen P: De- 1832 Journal of the American Society of Nephrology J Am Soc Nephrol 17: 1820–1832, 2006

velopment of lithium-induced nephrogenic diabetes insip- 97. Kwon T, Frokiaer J, Knepper M, Nielsen S: Reduced AQP1, idus is dissociated from adenylyl cyclase activity. JAmSoc -2, and -3 levels in kidneys of rats with CRF induced by Nephrol 17: 1063–1072, 2006 surgical reduction in renal mass. Am J Physiol Renal Physiol 96. Fernandez-Llama P, Andrews P, Turner R, Saggi S, Dimari 275: F724–F741, 1998 J, Kwon T, Nielsen S, Safirstein R, Knepper M: Decreased 98. Tannen RL, Regal EM, Dunn MJ, Schrier RW: Vasopressin- abundance of collecting duct aquaporins in post-ischemic resistant hyposthenuria in advanced chronic renal disease. renal failure in rats. J Am Soc Nephrol 10: 1658–1668, 1999 N Engl J Med 280: 1135–1141, 1969

Access to UpToDate on-line is available for additional clinical information at http://www.jasn.org/