Molecular Mechanisms of Impaired Urinary Concentrating Ability in Glucocorticoid-Deficient Rats

Yung-Chang Chen,*† Melissa A. Cadnapaphornchai,*‡ Sandra N. Summer,* Sandor Falk,* Chunling Li,* Weidong Wang,* and Robert W. Schrier* Departments of *Medicine and ‡Pediatrics, University of Colorado Health Sciences Center, Denver, Colorado; and †Division of Critical Care Nephrology, Chang Gung Memorial Hospital, Taipei, Taiwan

The purpose of this study was to examine urinary concentrating ability and protein expression of renal aquaporins and ion transporters in glucocorticoid-deficient (GD) rats in response to water deprivation as compared with control rats. Rats underwent bilateral adrenalectomies, followed only by aldosterone replacement (GD) or both aldosterone and dexamethasone replacement (control). As compared with control rats, the GD rats demonstrated a decrease in cardiac output and mean arterial pressure. In response to 36-h water deprivation, GD rats demonstrated significantly greater urine flow rate and decreased urine osmolality as compared with control rats at comparable serum osmolality and plasma vasopressin concentrations. The initiator of the countercurrent concentrating mechanism, the sodium-potassium-2 chloride co-transporter, was significantly decreased, as was the medullary osmolality in the GD rats versus control rats. There was also a decrease in inner medulla aquaporin-2 (AQP2) and urea transporter A1 (UT-A1) in GD rats as compared with control rats. There was a decrease in outer medulla Gs␣ protein, an important factor in vasopressin-mediated regulation of AQP2. Immunohistochemistry studies confirmed the decreased expression of AQP2 and UT-A1 in kidneys of GD rats as compared with control. In summary, impairment in the urinary concentrating mechanism was documented in GD rats in association with impaired countercurrent multiplication, diminished osmotic equilibration via AQP2, and diminished urea equilibration via UT-A1. These events occurred primarily in the relatively oxygen-deficient medulla and may have been initiated, at least in part, by the decrease in mean arterial pressure and thus renal perfusion pressure in this area of the kidney. J Am Soc Nephrol 16: 2864–2871, 2005. doi: 10.1681/ASN.2004110944

he ability to conserve water during periods of fluid limb, creates the osmotic driving force for passive water reab- deprivation is an important function of the kidney. In sorption across the collecting duct. There are also roles for other T both humans and experimental animals, adrenal insuf- water channels, including aquaporins 1, 3, and 4, and for urea ficiency has been associated with several alterations in renal transporters in urinary concentration. This study was under- function, including impairment of urinary diluting and concen- taken to define the effect of glucocorticoid deficiency on these trating capacity (1–7). Isolated glucocorticoid deficiency has various molecular events during fluid deprivation in the rat. also been associated with impaired urinary concentration (8,9). However, the mechanisms of this defect at the cellular and Materials and Methods molecular levels have not been defined. Animal Model In this study, glucocorticoid-deficient (GD) rats were com- The study protocol was approved by the University of Colorado pared with glucocorticoid-replete rats with respect to their Institutional Animal Care and Use Committee. Male Sprague-Dawley capacities to concentrate the urine. Critical components of uri- rats that weighed 175 to 200 g were allowed to acclimate to Denver’s nary concentration in response to fluid deprivation include altitude (1500 m) for 1 wk before any experimental protocols. All release of the antidiuretic hormone arginine vasopressin (AVP) animals underwent acclimation to metabolic cages for a continuous 5-d and upregulation of the abundance of aquaporin-2 (AQP2) period before initiation of study. The animals were housed individually water channels in the principal cells of the collecting duct. in metabolic cages and exposed to a 12-h light-dark cycle and constant Moreover, activation of the countercurrent concentrating mech- ambient temperature. Eighteen rats were divided equally into each of anism, which is initiated by the sodium-potassium-2 chloride two study groups: GD and control. Under anesthesia with ketamine (40 (Na-K-2Cl) co-transporter in the water-impermeable ascending mg/kg body wt intraperitoneally) and xylazine (5 mg/kg body wt intraperitoneally), all animals were adrenalectomized through bilateral flank incisions. Simultaneously, osmotic minipumps (Alzet Osmotic Pump model 2ML4; Durect, Cupertino, CA) that contained aldosterone Received November 16, 2004. Accepted July 13, 2005. (Research Plus, Bayonne, NJ) at a dose calculated to deliver 17 ␮g/kg Published online ahead of print. Publication date available at www.jasn.org. per 24 h into the peritoneal cavity were implanted into GD rats (10). Control rats received combined treatment with osmotic minipumps Address correspondence to: Dr. Robert W. Schrier, Division of Renal Diseases that contained aldosterone plus subcutaneous injections of dexameth- and Hypertension, University of Colorado Health Sciences Center, 4200 East 9th ␮ Avenue, Box B173, Denver, CO 80262. Phone: 303-315-8059; Fax: 303-315-2685; asone (Research Plus) dissolved in peanut oil at a dose of 12 g/kg per E-mail: [email protected] d starting immediately after adrenalectomy. This dose of dexametha-

Copyright © 2005 by the American Society of Nephrology ISSN: 1046-6673/1610-2864 J Am Soc Nephrol 16: 2864–2871, 2005 Impaired Urinary Concentrating Ability in Glucocorticoid-Deficient Rats 2865 sone has been reported to maintain normal weight gain, GFR, and medulla; and UT-A1 and AQP4 in the inner medulla. SDS-PAGE was fasting plasma glucose and insulin levels in adrenalectomized rats (11). performed on 8% acrylamide gels for the Na-K-2Cl co-transporter, For this study, we elected to use hormone-replaced adrenalectomized Na-K-ATPase ␣1 subunit, and UTA and on 12% acrylamide gels for rats as controls, rather than intact (unaltered) animals. Our studies AQP, NHE3, Na-K-ATPase ␤1 subunit, and Gs␣ subunit proteins. After (data not shown) demonstrated that the protein abundance of inner transfer by electroelution to polyvinylidene difluoride membrane (Mil- medulla urea transporter A1 (UT-A1) was similarly diminished in the lipore, Bedford, MA), blots were blocked overnight with 5% nonfat intact rats and in our controls. dried milk in PBS(Ϫ) and then probed with the respective antibodies After adrenalectomies, all rats were pair-fed with plain powdered rat for 24 h at 4°C. After washing with buffer that containing PBS(Ϫ) with chow (Harlan Teklad Bioproducts, Indianapolis, IN) 15 g/d. Drinking 0.1% Tween 20 (J.T. Baker, Phillipsburg, NJ), the membranes were water was provided ad libitum. All animals were maintained in meta- exposed to secondary antibody for1hatroom temperature. Subse- bolic cages for the duration of the study to assess accurately daily food quent detection of the specific proteins was carried out by enhanced intake, water intake, and urine output. chemiluminescence (Amersham, Arlington Heights, IL) according to On day 7 after adrenalectomies, echocardiography was performed the manufacturer’s instructions. Prestained protein markers were used using a GE Vingmed System 5 imaging tool (GE, Horten, Norway) for for molecular mass determinations. Densitometric results were re- small rodents with a 10-MHz probe. The animals were anesthetized for ported as integrated values (area ϫ density of band) and expressed as echocardiography with ketamine and xylazine in doses as described a percentage compared with the mean value in controls (100%). Mem- above. Cardiac output was calculated via measurement of the diameter branes were stained with Coomassie blue to ensure equal loading. For of the left ventricular outflow tract (LVOT), the flow through the each gel, an identical gel was run in parallel and subjected to Coomas- outflow tract (VTI), and the heart rate (HR) by the formula, 0.785 ϫ sie staining to verify identical protein loading. Blots shown in the LVOT2 ϫ VTI ϫ HR. The right femoral artery then was catheterized Results section are representative of the results obtained from all sam- with a polyethylene tube (PE-50; Intramedic, Clays Adams, Parsippany, ples. Densitometry as shown in the Results section reflects means Ϯ NJ), and BP was measured using a Transpac disposable transducer SEM densitometry of all 18 samples. (Abbott Critical Care Systems, Salt Lake City, UT) connected to a Transonic Systems T106 BP monitor (Ithaca, NY). BP was analyzed Antibodies using WinDaq software (Dataq Instruments, Akron, OH). Cardiac out- Antibodies to AQP2, AQP3, AQP4, NHE3, UT-A1, and UT-A2 have put was factored by body weight and expressed as cardiac index (CI; been characterized previously (15–19). Anti–Na-K-ATPase ␣1 and ␤1 ml/min per 100 g). Total peripheral resistance (mmHg/min per ml/100 antibodies were obtained from Upstate Biotechnology (Lake Placid, g) was calculated by dividing mean arterial pressure (MAP) by CI. NY). Antibodies to AQP1 and the Na-K-2Cl co-transporter were ob- Stroke volume (ml/beat per 100 g) was obtained by dividing CI by HR tained from Chemicon International, Inc. (Temecula, CA). Anti-Gs␣ (12,13). The catheter was removed. Animals were allowed to recover antibody was obtained from Calbiochem-Novabiochem (San Diego, and then were returned to metabolic cages. CA). Two days after echocardiography, all animals were subjected to a 36-h period of water deprivation, during which time food intake and Immunohistochemical Studies urine output were recorded. Urine was collected under oil, and urine Tissue samples were fixed in 4% formaldehyde solution for 24 h, volume was measured every 12 h. In the final 12 h of the water dehydrated, embedded in paraffin, and cut into 2-␮m-thick slices. deprivation period, urine was collected for osmolality and creatinine Immunohistochemical staining was performed using the avidin-bioti- and urea concentrations. Animals then were killed by decapitation to nylated peroxidase method. After deparaffinization and rehydration, avoid any influence of anesthesia on plasma AVP concentration (14). sections were pretreated with 0.3% hydrogen peroxidase in 70% meth- Trunk blood was collected for plasma AVP concentration, blood urea anol to exhaust endogenous peroxidase activities. Sections were prein- nitrogen, serum glucose, serum osmolality, serum sodium, and serum cubated with 10% horse serum, then incubated with the antibodies to creatinine concentration. UT-A at a 1:50 dilution and AQP2 at a 1:200 dilution at 37°C for 1 h. Slides were washed and incubated with biotinylated secondary anti- body, goat anti-rabbit IgG, at a 1:400 dilution in PBS. After treatment of Protein Isolation the slides with an Elite ABC kit (Vector Laboratories, Burlingame, CA), After decapitation, kidneys were placed in ice-cold isolation solution antigens were visualized with the Sigma (St. Louis, MO) fast 3,3- that contained 250 mM sucrose, 25 mM imidazole, 1 mM EDTA (pH diaminobenzidine tablet system applied on the slides for 1 min (AQP2) 7.2), with 0.1% vol protease inhibitors (0.7 ␮g/ml pepstatin, 0.5 ␮g/ml and 5 min (UT-A). Counterstaining was performed with Mayers’ he- leupeptin, and 1 ␮g/ml aprotinin), and 200 ␮M PMSF. Kidneys were matoxylin (Fluka, Buchs, Switzerland) for 30 s. dissected on ice into cortex, outer medulla, and inner medulla regions. Tissue samples were homogenized immediately in a glass homogenizer Biochemical Measurements at 4°C. After homogenization, protein concentration was determined Plasma AVP concentration was assessed by RIA as described previ- for each sample by the Bradford method (Bio-Rad, Richmond, CA). ously (14). Serum and urine osmolality was measured by freezing point Tissue protein was used for immunoblotting for AQP water channels depression (Advanced Instruments, Inc., Norwood, MA). Serum and and sodium and urea transporters. urine urea nitrogen and creatinine were measured (Beckman Instru- ments, Inc., Fullerton, CA). Twenty-four-hour creatinine clearance was Western Blot Analysis used as an estimate of GFR. Serum sodium concentrations were mea- Western blot analysis was performed to examine expression of sured by flame photometry. Blood glucose level was measured using a AQP1, AQP2, and Gs␣ subunit proteins in the renal cortex, outer single-touch glucometer. medulla, and inner medulla; the Na-K-2Cl co-transporter, sodium- potassium-adenosine triphosphatase (Na-K-ATPase) ␣1 and ␤1 sub- Medullary Tonicity units, and sodium-hydrogen exchanger (NHE3) in the renal cortex and Samples of inner medulla were placed in a preweighed Eppendorf outer medulla; AQP3 and UT-A2 in the outer medulla and inner tube that contained 200 ␮l of deionized distilled water. The tissue was 2866 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 2864–2871, 2005 homogenized in a glass homogenizer at 4°C. Tissue osmolality was measured by freezing point depression (Advanced Instruments, Inc.). The original tissue osmolality was estimated on the basis of the nominal dilution factor and the assumption that 80% of the wet weight was water (20).

Statistical Analyses Statistical analysis of results was performed using the unpaired t test. Results were expressed as means Ϯ SEM with P Ͻ 0.05 considered significant.

Results Systemic Hemodynamics in GD and Control Rats Systemic hemodynamic studies revealed significantly de- creased MAP and cardiac index in GD rats (n ϭ 9) as compared with control rats (n ϭ 9; Table 1). Neither calculated total peripheral resistance nor HR increased in response to the car- diac output–mediated decrease in MAP.

Water Deprivation Studies All rats then were subjected to a 36-h period of water depri- vation. During this period, GD rats demonstrated significantly higher urine flow rate (Figure 1A) than control rats. In addition, maximal urine osmolality was significantly decreased in GD rats in response to water deprivation as compared with control rats (Figure 1B). The medullary osmolality was decreased to a similar degree as urinary osmolality in the GD rats (Figure 1C). After the 36-h water deprivation period, GD rats demonstrated a significantly lower serum glucose concentration than control rats. Fractional excretion of urea was significantly increased in GD rats after water deprivation as compared with control rats. Creatinine clearance, serum creatinine, sodium, and osmolality were similar between the study groups. No significant differ- Figure 1. Glucocorticoid deficient (GD) rats demonstrated ences in plasma AVP concentrations were found in response to increased urine flow rate (A), decreased maximal urinary water deprivation. These results are shown in Table 2. osmolality (B), and decreased medullary osmolality (C) in response to 36-h water deprivation as compared with control Alterations to Renal Cortex, Outer Medulla, and Inner (CTL) rats. Medulla with Fluid Deprivation Immunoblotting studies were performed for ion transporters and AQP water channels as described below for both GD (n ϭ 7) Renal Cortex and control (n ϭ 8) rats. Blots shown are representative of all In response to fluid deprivation, there were no significant samples. Densitometric results were reported as integrated values changes by immunoblotting in the renal cortex protein abun- (area ϫ density of band) and expressed as a percentage compared dance of NHE3, the Na-K-2Cl co-transporter, Na-K-ATPase ␣1 with the mean value in controls (100%) for each blot and are and ␤1 subunit and Gs␣ subunit, or AQP1 or 2 in the GD as representative of all samples from both GD and control rats. compared with the control rats (Table 3).

Table 1. Systemic hemodynamic data obtained on day 7 after adrenalectomies in GD and control ratsa

Control GD P

Heart rate (beats/min) 296 Ϯ 22 250 Ϯ 14 NS (0.10) Mean arterial pressure (mmHg) 93 Ϯ 773Ϯ 3 0.02 Cardiac index (ml/min per 100 g) 34.9 Ϯ 3.1 26.8 Ϯ 1.8 0.04 Total peripheral resistance (mmHg/min per ml/100 g) 2.7 Ϯ 0.4 2.7 Ϯ 0.2 NS (1.00)

aAll data are expressed as the mean Ϯ SEM; n ϭ 9 in each study group. GD, glucocorticoid-deficient. J Am Soc Nephrol 16: 2864–2871, 2005 Impaired Urinary Concentrating Ability in Glucocorticoid-Deficient Rats 2867

Table 2. Characteristics of study groups after 36-h period of water deprivationa

Control GD P Serum glucose (mg/dl) 114 Ϯ 581Ϯ 10 Ͻ0.01 Serum creatinine (mg/dl) 0.25 Ϯ 0.03 0.22 Ϯ 0.03 NS (0.49) Creatinine clearance (ml/min) 1.15 Ϯ 0.18 1.04 Ϯ 0.22 NS (0.70) Blood urea nitrogen (mg/dl) 21 Ϯ 222Ϯ 1 NS (0.66) Fractional excretion of urea (%) 12.7 Ϯ 4.9 27.6 Ϯ 4.1 0.03 Serum sodium (mmol/L) 143 Ϯ 1 143 Ϯ 1 NS (1.00) Ϯ Ϯ Serum osmolality (mOsm/kg H2O) 310 2 310 3 NS (1.00) Plasma AVP (pg/ml) 12.3 Ϯ 1.0 13.3 Ϯ 1.5 NS (0.59)

aAll data are expressed as the mean Ϯ SEM; n ϭ 9 for each study group. AVP, arginine vasopressin.

Table 3. Summary of densitometry of immunoblots from study groups with nonsignificant differences in response to 36-h period of water deprivationa

Control (n ϭ 8) GD (n ϭ 7) P Cortex Na-K-2Cl co-transporter 100 Ϯ 37 130 Ϯ 45 NS (0.61) NHE3 100 Ϯ 14 81 Ϯ 8 NS (0.28) Na-K-ATPase ␣1 subunit 100 Ϯ 6 119 Ϯ 9 NS (0.10) Na-K-ATPase ␤1 subunit 100 Ϯ 11 67 Ϯ 15 NS (0.09) Gs␣ subunit 100 Ϯ 10 86 Ϯ 9 NS (0.32) AQP1 100 Ϯ 12 103 Ϯ 21 NS (0.90) AQP2 100 Ϯ 789Ϯ 4 NS (0.21) Outer medulla AQP1 100 Ϯ 17 122 Ϯ 14 NS (0.34) AQP3 100 Ϯ 8 107 Ϯ 7 NS (0.53) UT-A2 100 Ϯ 15 92 Ϯ 24 NS (0.78) Inner medulla Gs␣ subunit 100 Ϯ 7 143 Ϯ 22 NS (0.07) AQP3 100 Ϯ 11 107 Ϯ 13 NS (0.69) AQP4 100 Ϯ 996Ϯ 6 NS (0.73) UT-A2 100 Ϯ 19 104 Ϯ 29 NS (0.91)

aValues are mean Ϯ SEM. Densitometric results were reported as integrated values (area ϫ density of band) and expressed as a percentage compared with the mean value in controls (100%). Na-K-2Cl , sodium-potassium-2 chloride; NHE3, sodium- hydrogen exchanger; Na-K-ATPase, sodium-potassium-adenosine triphosphatase; UT, urea transporter.

Outer Medulla compared with the control rats (Figure 3, A and B). There were, In the more hypoxic outer medullary region of the kidney, there however, no differences in AQP3 and 4 (Table 3). were several significant differences between the GD and control UT-A2 protein expression was no different between the outer rats with fluid deprivation. Protein abundance of the Na-K-2Cl and inner medulla (Table 3), but UT-A1 was significantly de- co-transporter, Na-K-ATPase ␣1 and ␤1 subunits, and NHE3 was creased in the inner medulla of the GD as compared with the significantly decreased in the GD as compared with control rats control rats (Figure 3C). In Figure 4 is shown the diminished UT-A (Figure 2). The Gs␣ protein expression was also significantly de- by immunohistochemistry in the base of the inner medulla in the creased in the outer medulla in the GD as compared with control GD rats. This difference was not evident in the tip of the inner rats (88 Ϯ 3 versus 100 Ϯ 1% control mean; P Ͻ 0.01) as was the medulla. In Figure 5 is demonstrated the diminished immunohis- AQP2 water channel (95 Ϯ 1 versus 100 Ϯ 1; P Ͻ 0.01). There were tochemistry staining for AQP2 in the inner medulla in GD rats. In no differences in AQP1 and 3 or UT-A2. fact, this decrease in AQP2 seems somewhat greater than that observed by Western analysis. We do not have an explanation for the modest quantitative difference. Inner Medulla In the inner medulla, there were also significant differences Discussion in expression of several proteins between GD and control rats. In our study, GD rats were demonstrated to have a signifi- AQP1 and 2 both were significantly decreased in the GD as cant diminution in maximal urinary osmolality during 36 h of 2868 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 2864–2871, 2005

Figure 2. GD rats demonstrated significant decreases in the outer medulla protein abundance of the Na-K-2Cl co-transporter (A), Na-K-ATPase ␣1 subunit (B), Na-K-ATPase ␤1 subunit (C), and NHE3 (D) as compared with CTL rats.

fluid deprivation as compared with the glucocorticoid-replete renal perfusion pressure in the GD animals could have contrib- animals. The systemic hemodynamic alterations that accompa- uted to the diminished expression of the medullary Na-K-2Cl nied this defect in maximal urine concentration included a co-transporter in addition to any direct nephron effect of the decrease in cardiac output and MAP. A decrease in myocardial absence of glucocorticoid hormone. Support for this possibility contractility has been reported in the presence of glucocorticoid is that the protein expression of several transporters, including deficiency (21–23). The response of the peripheral vasculature the Na-K-2Cl co-transporter, Na-K-ATPase ␣1 and ␤1 subunits, to catecholamines has also been demonstrated to be dependent and NHE3, were not altered during glucocorticoid deficiency in on glucocorticoid hormones (24). In that regard, the GD rats the oxygen-rich renal cortex. were shown in this study not to demonstrate the increase in The observed diminution in protein expression of the Na-K- peripheral vascular resistance that normally occurs with a de- ATPase ␣1 and ␤1 subunits in the outer medulla would be crease in cardiac output and a fall in MAP. Thus, the GD state expected to decrease vectorial sodium transport across the ba- was shown to exhibit an effect at both the cardiac and the solateral membrane of the water-impermeable segment of the peripheral vascular levels. Despite the lower MAP in the GD thick ascending limb and thus further impair the countercur- rats, the creatinine clearance, an assessment of GFR, was no rent concentrating mechanism. Some of the diminished NHE3 different from that observed in the glucocorticoid-replete ani- in the outer medulla with glucocorticoid deficiency no doubt mals. This observation was no doubt due to the preservation of occurred in the proximal tubule, but there is also histochemical autoregulation of GFR at the MAP of 73 mmHg observed in the evidence for expression of NHE3 in medullary thick ascending GD rats. limb of the rat (25). A decrease in sodium-hydrogen exchange Although a decline in GFR was not involved in the impair- in the thick ascending limb also could contribute to impairment ment of renal water conservation, there was considerable evi- of the countercurrent concentrating mechanism. dence that several nephron factors that are important to normal There are additional factors involved in maximal urinary urinary concentration were perturbed in the GD state. The concentration that were examined in our study. AQP1 knock- primary component of the countercurrent concentrating mech- out mice (26) and humans without AQP1 (27) are known to anism is the active reabsorption of sodium chloride in the exhibit defects in urinary concentration. AQP1 water channels water-impermeable portion of the medullary thick ascending in the thin descending limb of Henle play an important role in limb by the Na-K-2Cl co-transporter. In the GD animals, there water abstraction, and the resulting increase in luminal sodium was a significant downregulation of the protein expression of chloride concentration toward the tip of the inner medulla the Na-K-2Cl co-transporter in the outer medulla. The outer therefore is critical to the intact countercurrent concentrating medulla is a relatively hypoxic area of the nephron as com- mechanism. In our study, the expression of AQP1 water chan- pared with the renal cortex, and thus the observed decrease in nels was diminished in the inner medulla of the GD rats. J Am Soc Nephrol 16: 2864–2871, 2005 Impaired Urinary Concentrating Ability in Glucocorticoid-Deficient Rats 2869

Figure 4. Glucocorticoid deficiency was associated with dimin- ished UT-A as assessed by immunohistochemistry in inner medulla base (IM base; B versus D) but not in terminal inner medulla (IM tip; A versus C) as compared with controls (CTL). Magnification, ϫ400.

Figure 3. GD rats demonstrated significant decreases in the renal inner medulla protein abundance of aquaporin 1 (AQP1; A), AQP2 (B), and urea transporter A1 (UT-A1; C). Figure 5. GD rats (B) demonstrated diminished staining for AQP2 in inner medulla on immunohistochemical studies as compared with CTL (A). Negative control (NC; C) stained with secondary antibody only. Magnification, ϫ400. The interstitial osmotic gradient that is created in the medulla by the countercurrent concentrating mechanism involves not only sodium chloride but also accumulation of urea. Stress doses of exogenous glucocorticoid hormone (100 ␮g/kg per d rats that received mineralocorticoid but not glucocorticoid re- of dexamethasone), which exceed the physiologic doses (12 placement demonstrated a decrease in UT-A1 protein expres- ␮g/kg per d) used in this study, have been shown to down- sion in the inner medulla. UT-A2 expression was not altered in regulate the UT-A1 protein abundance in the inner medulla either the outer or the inner medulla. The glucocorticoid defi- (28). High endogenous levels of glucocorticoid hormone in the ciency was also associated with an increase in the fractional untreated diabetic rat have also been associated with a similar excretion of urea, a finding consonant with the decrease in UT-A1 protein repression (29). In this study, adrenalectomized UT-A1 protein expression. Immunohistochemistry studies 2870 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 2864–2871, 2005

demonstrated that the decrease in the UT-A1 protein was pri- dominant defect in renal water conservation during glucocor- marily in the basal portion of the inner medulla of the glucocor- ticoid deficiency. ticoid rat. This is particularly important because replacement of ␮ dexamethasone (100 g/kg per d) has been shown to decrease Acknowledgments the vasopressin-responsive urea transporter only in the termi- This work was supported by the National Institutes of Health nal portion of the inner medulla of adrenalectomized rats that (DK19928). Y.-C.C. was supported by a grant from Chang Gung Me- did not receive replacement of mineralocorticoid (28). In our morial Hospital (Taipei, Taiwan). study, the decrease in UT-A1 in the GD rats would be expected to diminish urea transport and recycling between vasa recta and the thin descending limb of Henle’s loop, an important References factor in the maximal generation of the medullary osmotic 1. Gill JR Jr, Gann DS, Bartter FC: Restoration of water diure- gradient. The decrease in inner medullary protein abundance sis in Addisonian patients by expansion of the volume of for UT-A1 was associated with diminished UTA by immuno- extracellular fluid. J Clin Invest 41: 1078–1085, 1962 histochemistry. 2. Green HH, Harrington AR, Valtin H: On the role of anti- Maximal urine concentration also involves the integrity of diuretic hormone in the inhibition of acute water diuresis membrane water channels. In addition to AQP1, the AQP2, 3, in adrenal insufficiency and the effects of gluco- and min- and 4 water channels in the mammalian collecting duct are eralocorticoids in reversing the inhibition. J Clin Invest 49: important in the concentration of urine. In our study, the ob- 1724–1736, 1970 3. Cooke CR, Steenberg RW: Effects of aldosterone and cor- served decrease in maximal urinary osmolality and increase in tisol on the renal concentrating mechanism. J Lab Clin Med urine flow in the GD rats during fluid deprivation are compat- 82: 784–792, 1973 ible with a downregulation or decreased trafficking of the 4. Boykin J, DeTorrente A, Erickson A, Robertson G, Schrier AQP2 water channels to the apical surface of the collecting RW: Role of plasma vasopressin in impaired water excre- duct, as has been shown in several models of acquired neph- tion of glucocorticoid deficiency. J Clin Invest 62: 738–744, rogenic diabetes insipidus (e.g., hypokalemia, hypercalcemia, 1978 lithium administration) (30–32). Although there were no dif- 5. Linas SL, Berl T, Robertson GL, Aisenbrey GA, Schrier RW, ferences in AQP3 and 4 between GD and control rats, there was Anderson RJ: Role of vasopressin in the impaired water- a decrease in AQP2 expression in both the outer and the inner excretion of glucocorticoid deficiency. Kidney Int 18: 58–67, medulla in the absence of glucocorticoid hormone. This alter- 1980 6. Ishikawa S, Schrier RW: Effect of arginine vasopressin ation in AQP2 between the GD and glucocorticoid-replete antagonist on renal water-excretion in glucocorticoid and states during fluid deprivation occurred in the absence of dif- mineralocorticoid deficient rats. Kidney Int 22: 587–593, ferences in plasma vasopressin. The action of vasopressin with 1982 glucocorticoid deficiency, however, could have been blunted as 7. Ohara M, Cadnapaphornchai MA, Summer SN, Falk S, ␣ a result of the observed decrease in the outer medulla of Gs , Yang JH, Togawa T, Schrier RW: Effect of mineralocorti- an important factor in vasopressin signaling. coid deficiency on ion and urea transporters and aqua- The medullary osmolality was diminished in the GD state, a porin water channels in the rat. Biochem Biophys Res Com- finding supportive of an impairment in the countercurrent mun 299: 285–290, 2002 concentrating mechanism. Although glucocorticoid deficiency 8. Schwartz MJ, Kokko JP: Urinary concentrating defect of was associated with a decrease in AQP2, there seemed to be adrenal insufficiency. Permissive role of adrenal steroids osmotic water equilibration between the terminal medullary on the hydroosmotic response across the rabbit cortical collecting tubule. J Clin Invest 66: 234–242, 1980 collecting duct fluid and the medullary interstitium, because no 9. Jackson BA, Braun-Werness JL, Kusano E, Dousa TP: Con- significant difference was observed between the medullary and centrating defect in the adrenalectomized rat. Abnormal urine osmolality. vasopressin-sensitive cyclic adenosine monophosphate In summary, an impairment in maximal urinary concentra- metabolism in the papillary collecting duct. J Clin Invest 72: tion was demonstrated in adrenalectomized rats that received 997–1004, 1983 replacement of aldosterone but not glucocorticoid. This defect 10. Yagil Y, Krakoff LR: The differential effect of aldosterone in urinary concentration in the GD state was associated with and dexamethasone on pressor responses in adrenalecto- diminished protein expression of critical factors in the counter- mized rats. Hypertension 11: 174–178, 1988 current concentrating mechanism, including the Na-K-2Cl co- 11. Stanton B, Giebisch G, Klein-Robbenhaar G, Wade J, De- transporter, Na-K-ATPase ␣1 and ␤1, AQP1, and UT-A1. These Fronzo RA: Effects of adrenalectomy and chronic adrenal defects associated with glucocorticoid deficiency occurred in corticosteroid replacement on potassium transport in rat kidney. J Clin Invest 75: 1317–1326, 1985 the relatively hypoxic medulla in association with a decrease in 12. Cadnapaphornchai MA, Kim YW, Gurevich AK, Summer cardiac output and renal perfusion pressure. There was also a SN, Falk S, Thurman JM, Schrier RW: Urinary concentrat- significant decrease in AQP2 but not AQP3 or 4 water channels ing defect in hypothyroid rats: Role of sodium, potassium, during glucocorticoid deficiency. However, because equilibra- 2-chloride co-transporter, and aquaporins. J Am Soc Neph- tion of fluid between medullary interstitium and urinary os- rol 14: 566–574, 2003 molality was observed, a multifactorial impairment in the 13. Chen YC, Gines P, Yang J, Summer SN, Falk S, Russell NS, countercurrent concentrating mechanism seems to be the pre- Schrier RW: Increased vascular heme oxygenase-1 expres- J Am Soc Nephrol 16: 2864–2871, 2005 Impaired Urinary Concentrating Ability in Glucocorticoid-Deficient Rats 2871

sion contributes to arterial vasodilation in experimental cardiac muscle in adrenal insufficiency. Am J Physiol 218: cirrhosis in rats. Hepatology 39: 1075–1087, 2004 1015–1019, 1970 14. Howard RL, Summer S, Rossi N, Schrier RW: Short-term 24. Cueto C Jr, Moran NC: The circulatory effects of cat- hypothyroidism and vasopressin gene expression in the echolamines and ouabain in glucocorticoid-deficient ani- rat. Am J Kidney Dis 19: 573–577, 1992 mals. J Pharmacol Exp Ther 164: 31–44, 1968 15. Terris J, Ecelbarger CA, Marples D, Knepper MA, Nielsen 25. Amemiya M, Loffing J, Lotscher M, Kaissling B, Alpern RJ, S: Distribution of aquaporin-4 water channel expression Moe OW: Expression of NHE-3 in the apical membrane of within rat kidney. Am J Physiol 269: F775–F785, 1995 rat renal proximal tubule and thick ascending limb. Kidney 16. Fernandez-Llama P, Andrews P, Ecelbarger CA, Nielsen S, Int 48: 1206–1215, 1995 Knepper M: Concentrating defect in experimental ne- 26. Ma T, Yang B, Gillespie A, Carlson EJ, Epstein CJ, Verkman phrotic syndrome: Altered expression of aquaporins and AS: Severely impaired urinary concentrating ability in ϩ thick ascending limb Na transporters. Kidney Int 54: 170– transgenic mice lacking aquaporin-1 water channels. J Biol 179, 1998 Chem 273: 4296–4299, 1998 17. Wade JB, Lee AJ, Liu J, Ecelbarger CA, Mitchell C, Brad- 27. King LS, Choi M, Fernandez PC, Cartron JP, Agre P: De- ford AD, Terris J, Kim GH, Knepper MA: UT-A2: A 55-kDa fective urinary-concentrating ability due to a complete de- urea transporter in thin descending limb whose abundance ficiency of aquaporin-1. N Engl J Med 345: 175–179, 2001 is regulated by vasopressin. Am J Physiol Renal Physiol 278: 28. Naruse M, Klein JD, Ashkar ZM, Jacobs JD, Sands JM: F52–F62, 2000 Glucocorticoids downregulate the vasopressin-regulated 18. Bradford AD, Terris JM, Ecelbarger CA, Klein JD, Sands urea transporter in rat terminal inner medullary collecting JM, Chou CL, Knepper MA: 97- and 117-kDa forms of ducts. J Am Soc Nephrol 8: 517–523, 1997 collecting duct urea transporter UT-A1 are due to different 29. Klein JD, Price SR, Bailey JL, Jacobs JD, Sands JM: Glu- states of glycosylation. Am J Physiol Renal Physiol 281: F133– cocorticoids mediate a decrease in AVP-regulated urea F143, 2001 19. Xu DL, Martin PY, Ohara M, St John J, Pattison T, Meng X, transporter in diabetic rat inner medulla. Am J Physiol 273: Morris K, Kim JK, Schrier RW: Upregulation of aqua- F949–F953, 1997 porin-2 water channel expression in chronic heart failure 30. Marples D, Frokiaer J, Dorup J, Knepper MA, Nielsen S: rat. J Clin Invest 99: 1500–1505, 1997 Hypokalemia-induced downregulation of aquaporin-2 wa- 20. Apostol E, Ecelbarger CA, Terris J, Bradford AD, Andrews ter channel expression in rat kidney medulla and cortex. P, Knepper MA: Reduced renal medullary water channel J Clin Invest 97: 1960–1968, 1996 expression in puromycin aminonucleoside-induced ne- 31. Earm JH, Christensen BM, Frokiaer J, Marples D, Han JS, phrotic syndrome. J Am Soc Nephrol 8: 15–24, 1997 Knepper MA, Nielsen S: Decreased aquaporin-2 expres- 21. Penefsky ZJ, Kahn M: Inotropic effects of dexamethasone sion and apical plasma membrane delivery in kidney col- in mammalian heart muscle. Eur J Pharmacol 15: 259–266, lecting ducts of polyuric hypercalcemic rats. JAmSoc 1971 Nephrol 9: 2181–2193, 1998 22. Lefer AM: Influence of corticosteroids on mechanical per- 32. Marples D, Christensen S, Christensen EI, Ottosen PD, formance of isolated rat papillary muscles. Am J Physiol Nielsen S: Lithium-induced downregulation of aqua- 214: 518–524, 1968 porin-2 water channel expression in rat kidney medulla. 23. Rovetto MJ, Lefer AM: Electrophysiologic properties of J Clin Invest 95: 1838–1845, 1995

See related editorial, “Molecular Physiology of Renal Aquaporins and Sodium Transporters: Exciting Approaches to Understand Regulation of Renal Water Handling,” on pages 2827–2829.

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