点击显示 收起
【关键词】 polyuria
Division of Nephrology, Department of Medicine, Center for Health Service Research, Vanderbilt University Medical Center, and Veterans Administration Medical Center, Nashville, Tennessee
ABSTRACT
The use of LiCl in clinical psychiatry is routinely complicated by overt nephrogenic diabetes insipidus (NDI), the mechanism of which is incompletely understood. In vitro studies indicate that lithium can induce renal medullary interstitial cell cyclooxygenase 2 (COX2) protein expression via inhibition of glycogen synthase kinase-3 (GSK-3). Both COX1 and COX2 are expressed in the kidney. Renal prostaglandins have been suggested to play an important role in lithium-induced polyuria. The present studies examined whether induction of the COX2 isoform contributes to LiCl-induced polyuria. Four days after initiation of lithium treatment in C57 BL/6J mice, urine volume increased in LiCl-treated mice by fourfold compared with controls (P < 0.0001) and was accompanied by decreased urine osmolality. This was temporally associated with increased renal COX2 protein expression and increased urinary PGE2 excretion, whereas COX1 levels remained unchanged. COX2 inhibition significantly blunted lithium-induced polyuria (P < 0.0001) and reduced urinary PGE2 levels. Lithium-associated polyuria was also seen in COX1/ mice and was associated with increased urinary PGE2. COX2 inhibition completely prevented polyuria and PGE2 excretion in COX1/ mice, suggesting that COX2, but not COX1, plays a critical role in lithium-induced polyuria. Lithium also induced renal medullary COX2 protein expression in congenitally polyuric antidiuretic hormone (AHD)-deficient rats, demonstrating that lithium-induced COX2 protein expression is not secondary to altered ADH levels or polyuria. Lithium also decreased renal medullary GSK-3 activity, and this was temporally related to increased COX2 expression in the kidney from lithium-treated mice, consistent with a tonic in vivo suppression of COX2 expression by GSK-3 activity. In conclusion, these findings temporally link decreased GSK-3 activity to enhanced renal COX2 expression and COX2-derived urine PGE2 excretion. Suppression of COX2-derived PGE2 blunts lithium-associated polyuria.
prostaglandin E2; urine osmolality
LITHIUM, A METALLIC MONOVALENT cation, has been used therapeutically for more than 150 years and remains an important pharmacotherapeutic agent for treatment of bipolar disorder and Alzheimer's disease (10, 24, 34). It is estimated that 1 of every 1,000 individuals in the population is on lithium treatment (33). The use of lithium is frequently complicated by impaired renal water reabsorption, resulting in nephrogenic diabetes insipidus (NDI) (3). The mechanism by which lithium induces diabetes insipidus is incompletely understood. Sugawara et al. (31) reported that after LiCl treatment, rats exhibit increased urinary PGE2 excretion. They also found the nonselective cyclooxygenase-inhibiting NSAID indomethacin reduced lithium-induced PGE2 excretion and polyuria, suggesting that a diuretic cyclooxygenase product may be involved in lithium-associated diabetes insipidus. COX2, an inducible isoform of cyclooxygenase, is expressed in the kidney together with COX1, a constitutively expressed isoform. Furthermore, recent studies demonstrate that lithium can increase COX2 in cultured cells via inhibition of glycogen synthase kinase-3 (GSK-3) (27). The present studies were designed to test the possibility that COX2 activity rather than COX1 contributes to lithium-induced polyuria and determine whether inhibition of GSK-3 is involved.
METHODS
Experimental animals and treatment. Male C57BL/6J mice were purchased from Jackson Laboratories (Bar Harbor, ME) and individually housed in metabolic cages specially designed for mice. Mice were allowed free access to water and standard rodent chow placed in containers outside the cage to avoid urine contamination. To minimize stress related to the new environment, mice were housed in metabolic cages for at least 3 days before initiation of the studies.
Eighteen mice were divided into three experimental groups consisting of six mice each. Two groups were administered LiCl (4 mmol/kg body wt ip) daily for 16 days. On day 5 of LiCl injection, one of the groups was provided with a COX2-specific inhibitor (SC-58236, 6 μg/ml) (22) in the drinking water. Another group was provided with vehicle. The third group was injected with saline and used as a control. Daily water intake and urine output were monitored. Urine osmolality was measured using a microosmometer (μOsmette, Precision Systems, Natick, MA). At the end of the experiment, mice from the LiCl- and vehicle-treated groups were killed, and kidney microsomal proteins were extracted to assess COX2 expression by immunoblotting. Another group of mice (n = 15) were treated with LiCl (4 mmol?kg1?day1) for 0, 3, 4, 5, or 16 days, respectively, and kidney COX2 expression was determined by immunoblotting.
Vasopressin-deficient Brattleboro rats were injected with LiCl (4 mmol/kg ip, n = 4) or saline (n = 4) for 5 days. Expression of COX2 and COX1 was analyzed by immunoblotting. Distribution of COX2 within the kidney was examined by immunohistochemistry.
The COX1 homozygous null mice (COX1/) were provided by Dr. R. Langenbach and maintained as described (22). A total of seven COX1/ mice were housed individually in metabolic cages and treated with LiCl for 5 days as described before for C57BL/6J mice. Four of these mice were concurrently treated with MF-tricyclic (another COX2 inhibitor) (21), beginning on day 1 of a 5-day LiCl treatment. The remaining three mice were treated with vehicle. Urine volume, urinary osmolality, and urine PGE2 were determined. All protocols were approved through the institutional animal care and use committee of Vanderbilt University.
Isolation of microsomes. Since COX2 is mainly localized to the endoplasmic reticulum-associated plasma membrane (30), microsomes were isolated to determine COX2 protein expression as previously reported (36). Briefly, kidneys were homogenized in buffer (30 mM Tris, pH 8, 1 mM EDTA, and 2 mM EGTA) using a homogenizer (PowerGen 700, Fisher Scientific) and then centrifuged at 10,000 g. The supernatant was spun again at 150,000 g for 60 min. The microsomal pellet was resuspended in RIPA buffer [1% Triton X-100, 0.5% deoxycholate, 1% SDS, 150 mM NaCl, 50 mM Tris?HCl (pH 8), 2 mM EDTA, and 1 mM sodium orthovanadate], and protein levels were measured with subsequent analysis by immunoblotting for COX2 and COX1 protein expression.
Cell culture. Primary cultures of medullary interstitial cells from C57BL/6J mice were established as reported (12). Cells were cultured in DMEM supplemented with 10% (vol/vol) FBS at 37°C in 95% air-5% CO2. Subconfluent cultures in passages 1015 were used for the study.
Immunoblotting. Cultured renal medullary interstitial cells (RMICs) were washed with PBS and lysed in RIPA buffer followed by sonication for 12 s. Protein concentration of kidney microsomes and whole cell lysate was determined using a bicinochoninic acid protein assay (Sigma, St. Louis, MO). Thirty micrograms of protein extract in SDS Laemmli buffer were loaded in each lane of a 10% SDS-PAGE minigel and separated at 120 V. Proteins were transferred to a nitrocellulose membrane at 22 V overnight at 4°C. The membrane was washed three times with TBST (50 mM Tris, pH 7.5, 150 mM NaCl, 0.05% Tween 20) and then incubated in blocking buffer (150 mM NaCl, 50 mM Tris, 0.05% Tween 20, and 5% Carnation nonfat dry milk, pH 7.5) for 1 h at room temperature. The membrane was then incubated with anti-GSK-3 (1:2500, BD Transduction Laboratories), anti-pGSK (ser9) (1:1,000, Cell Signaling), anti-COX1 (1:300, Santa Cruz Biotechnology), or anti-COX2 (1:1,000, Cayman Chemicals, Ann Arbor, MI) antibodies in blocking buffer overnight at 4°C. After being washed three times, the membranes were incubated with horseradish peroxidase-conjugated secondary antibody (1:5,000 for anti-mouse and 1:10,000 for anti-rabbit) for 1 h at room temperature, followed by three 15-min washings. Antibody labeling was visualized by addition of chemiluminescence reagent (Renaissance, DuPont-New England Nuclear, Boston, MA), and the membrane was exposed to Kodak XAR-5 film and developed.
Immunohistochemistry. In general, at the termination of an experiment, one kidney was removed from each rat for Western blot analysis, and the other was perfused with fixative for histological analysis. Under deep anesthesia with Nembutal (70 mg/kg ip), rats were exsanguinated with 50 ml/100 g heparinized saline (0.9% NaCl, 2 U/ml heparin, 0.02% sodium nitrite) through a transcardial aortic cannula and fixed with 3.7% formaldehyde in an acidic solution (pH 4.5) containing phosphate, periodate, acetate, and sodium chloride as described (25). The acidified aldehyde fixatives were crucial for reliable preservation of inner medulla/papillary structure and COX2 antigenicity. The fixed kidney was dehydrated through a graded series of ethanols, embedded in paraffin, sectioned (4 μm), and mounted on glass slides. Internal controls and comparisons were facilitated by creating compound blocks with multiple specimens that were sectioned and stained together. COX2 was immunolocalized with affinity-purified rabbit polyclonal anti-murine COX2 peptide (residues 570598) antibody no. 160126 (Cayman Chemicals) at a dilution of 0.1 μg IgG/ml (37). Staining was localized using a biotinylated anti-rabbit secondary antibody. Biotin was identified using streptavidin coupled to horseradish peroxidase and visualized with diaminiobenzidine (Vector Vectastain ABC kit). Sections were viewed and imaged with a Zeiss Axioskop and Spot-Cam digital camera (Diagnostic Instruments).
GSK-3 kinase activity. GSK-3 activity was determined in the renal medulla from LiCl-injected mice as well as in cultured cells. After treatment with LiCl (30 mM) for 8 h, cultured RMICs were lysed on ice in freshly prepared lysis buffer (20 mM Tris?HCl, pH 7.5, 137 mM NaCl, 1 mM EDTA, 20 mM NaF, 1% Triton X-100, 10% glycerol, 1 mM DTT, 1 mM sodium vanadate, 5 μg of aprotinin/ml, 5 μg of leupeptin/ml, 1 mM phenylmethylsulfonyl fluoride, and 1 μM microcystin LR). Cell lysates were cleared by centrifugation for 10 min at 14,000 rpm in a microcentrifuge at 4°C and precleared with protein G-Sepharose. To determine GSK3- activity in the kidney, mouse kidneys were homogenized, lysed, and cleared using the same protocol as described for cultured cells.
GSK-3 activity was measured in vitro. GSK-3 was immunoprecipitated by incubating 100 μg of total cellular protein diluted in lysis buffer with 1 μg of anti-GSK-3 antibody overnight at 4°C. Protein G-Sepharose was added and incubated for another hour. Immunoprecipitates were washed twice with lysis buffer, twice with wash buffer (10 mM Tris?HCl, pH 7.4, 100 mM NaCl, 1 mM EDTA, 0.2 mM sodium vanadate, and 1 μM microcystin LR), and twice with kinase reaction buffer (20 mM HEPES, pH 7.4, 10 mM MgCl2, 1 mM DTT, and 0.2 mM EGTA). Kinase activity of the immunoprecipitated GSK-3 was assayed in a total volume of 40 μl of kinase buffer containing 3.75 μg of phospho-glycogen synthase peptide 2, 15 μM of cold ATP, and 10 μCi of [32P]ATP. After 20 min of incubation at 30°C, 20 μl of SDS lysis buffer were added and incubated at 70°C to stop the reaction. Reaction mixtures were centrifuged, and 15 μl of the supernatant were spotted onto Whatman P81 phosphocellulose paper. Filters were washed in three changes of 0.75% phosphoric acid, rinsed in acetone, dried, and 32P incorporation was measured in a liquid scintillation counter. Nonphosphorylated glycogen synthase peptide was used as a negative control. Nonspecific 32P incorporation was subtracted from values obtained using the phospho-glycogen synthase peptide.
Measurement of urinary PGE2 levels. Total PGE2 excreted in the urine (24-h collection) of lithium-, COX2 inhibitor+lithium-, or vehicle-treated mice were determined using an EIA kit (Cayman Chemicals) according to the manufacturer's protocol (2).
Statistical analysis. Values are presented as means ± SE. Comparison between groups was made by repeated-measures ANOVA using SAS 8.02 (proc mixed). P values <0.05 were considered significant.
RESULTS
Effect of COX2 inhibition on lithium-induced polyuria. Within 5 days of LiCl injection, daily urine output in C57BL/6 mice increased by fourfold above baseline values (urine volume: 0.58 ± 0.05 ml/24 h before LiCl injection vs. 2.58 ± 0.15 ml/24 h on day 5, P < 0.0001). Urine output in lithium-treated mice was also significantly greater than that observed in saline-injected mice (0.56 ± 0.18 ml, day 5, n = 6, P < 0.0001) (Fig. 1A). LiCl-associated polyuria was significantly reduced by addition of the COX2 inhibitor SC-58236, beginning on day 5. Urine volume remained reduced throughout the period of SC-58236 treatment compared with LiCl alone (P < 0.0001, by repeated-measures ANOVA). The urine volume on individual days was also different between COX2 inhibitor-treated and vehicle-treated mice. For example, on day 10, urine volume in lithium-treated mice was reduced by 50% from 3.22 ± 0.24 (lithium alone) to 1.85 ± 0.52 ml (lithium+SC-58236-treated mice). Thus, while SC-58236 blunted lithium-induced polyuria, it did not completely restore urine volume to control values (P < 0.0001, SC-58236 vs. vehicle-treated control group) (Fig. 1A).
Urinary osmolality also decreased following LiCl treatment (P < 0.001 vs. vehicle control) (Fig. 1B) from 4,562 ± 230 mosmol/kgH2O before LiCl to 3,341 ± 737 and 1,476 ± 135 mosmol/kgH2O by days 4 and 8 of LiCl injection, respectively. Coadministration of SC-58236 with LiCl increased urine osmolality compared with mice treated with LiCl alone (P < 0.001). For instance, on day 10 of lithium treatment, urine osmolality was 1,378 ± 402 mosmol/kgH2O in LiCl-treated mice vs. 2,359 ± 492 mosmol/kgH2O in mice treated with SC-58236+LiCl. Despite this increase, urine osmolality in COX2 inhibitor-treated mice remained significantly lower compared with vehicle-treated mice (P < 0.001).
Lithium increases COX2-derived urine PGE2 excretion. Since PGE2 is an endogenous diuretic autocoid derived from COX1 or COX2, urinary PGE2 excretion following LiCl treatment was measured in the presence or absence of a COX2 inhibitor. LiCl treatment increased urinary PGE2 excretion from 2.96 ± 0.08 (day 0) to 26.7 ± 0.2 pg/24 h by day 4 (P < 0.0001). Urine PGE2 was also significantly higher in lithium-treated mice than in saline-injected mice (2.98 ± 0.7 pg/24 h, P < 0.0001) (Fig. 2). The increase in urinary PGE2 associated with LiCl treatment temporally correlated with increased COX2 protein expression (see below). Furthermore, this lithium-associated increase in PGE2 excretion was halved when mice were treated with the COX2 inhibitor SC-58236 (P < 0.0001 by repeated-measures ANOVA). On day 10, the urine PGE2 excretion rate was 39.57 ± 2.1 (LiCl alone) vs. 19.97 ± 3.4 pg/24 h (LiCl+SC-58236).
Effect of lithium treatment on COX1-deficient mice. To more definitively exclude a role of COX1 in lithium-induced polyuria, the effect of LiCl on mice deficient in COX1 was examined. LiCl increased urinary PGE2 excretion by sixfold above basal levels in COX1-deficient mice (14 ± 6 on day 0 to 88 ± 15 pg/24 h on day 5, P < 0.0001). Importantly, when COX1/ mice were treated with a COX2 inhibitor (MF-tricyclic), lithium failed to increase urinary PGE2 levels (13 ± 4 pg/24 h) (Fig. 3C). Furthermore, while LiCl treatment also increased urine output in COX1-deficient mice from 0.75 ± 0.1 ml on day 0 to 3.2 ± 0.4 ml by day 5 (P < 0.01), coadministration of LiCl with the COX2 inhibitor dramatically suppressed lithium-induced diuresis. Following 5 days of SC-58236 treatment, urine output was only 1 ± 0.1 vs. 3.2 ± 0.04 ml with LiCl alone (P < 0.01) (Fig. 3A). Similarly, while lithium decreased the urine osmolality significantly in the COX1-deficient mice (4,287 ± 414 on day 0 vs. 1,231 ± 176 mosmol/kgH2O on day 5, P < 0.001), urinary osmolality of the COX2 inhibitor-treated COX1/ mice remained unchanged from control values and were significantly greater than with lithium alone (3,569 ± 213 mosmol/kgH2O at day 5 with lithium+COX2 inhibitor vs. 1,231 ± 176 mosmol/kgH2O with LiCl alone) (P < 0.0005) (Fig. 3B).
Lithium increases COX2 expression in RMICs. LiCl also increased COX2 protein expression on immunoblots of kidney microsomes isolated from mice killed on days 3, 4, 5, or 16 of LiCl treatment. Renal COX2 expression increased beginning on day 4 (the same day that urine volume increased) and remained elevated throughout the period of LiCl-associated polyuria (day 16) (Fig. 4A). COX1 levels remained unchanged. To further examine whether LiCl directly induced COX2 expression, mouse RMICs were cultured. As previously observed in rabbit RMICs, LiCl treatment (30 mM) markedly induced COX2 protein expression in cultured mouse RMICs (Fig. 4B), consistent with a direct effect of LiCl on interstitial cell COX2 expression.
It has been reported that lithium can elevate circulating antidiuretic hormone (ADH) levels, which can lead to increased urinary PGE2 excretion (31). To further determine whether increased medullary COX2 expression is caused by a direct cellular effect of LiCl rather than an effect secondary to altered ADH levels, rats with a hereditary ADH deficiency (Brattleboro rats) were studied. LiCl treatment significantly increased renal medullary COX2 protein expression compared with vehicle treatment in Brattleboro rats (Fig. 5). Because the Brattleboro rats are already in a state of maximum polyuria, no further increase in urine volume was observed in these rats during lithium treatment. From this we infer that lithium can stimulate COX2 expression, independently of changes in circulating ADH or as a consequence of altered urine volume. Immunohistochemical staining of kidneys from LiCl-injected rats showed perinuclear COX2 staining in medullary interstitial cells. COX2 immunoreactivity was not detected in the collecting ducts (Fig. 6). These studies are consistent with lithium-associated upregulation of COX2 observed on immunoblots (Fig. 4) being restricted to RMICs.
Lithium inhibits renal GSK-3 activity. Recent studies indicate that lithium acts as an ATP-noncompetitive inhibitor of GSK-3 (28). GSK-3 is a constitutively active kinase whose activity can be inhibited by phosphorylation of the Ser9 residue. Constitutive GSK-3 activity has recently been shown to tonically suppress COX2 expression in cultured renal interstitial cells (27). To test whether LiCl-induced COX2 expression is associated with reduced renal GSK-3 activity in vivo, the effect of LiCl treatment on renal GSK-3 activity (32P incorporation) was determined using primed glycogen synthase as a substrate. LiCl treatment significantly decreased renal GSK-3 kinase activity over the same time frame of increase in COX2 expression (3-fold by day 5 and 10-fold by day 7 of lithium treatment) (Fig. 7). Total GSK-3 protein level remained unaltered. Serum Li levels were 2.14 ± 0.13 mmol/l in these mice.
Similarly, treatment of cultured mouse RMICs with LiCl for 8 h decreased GSK-3 activity by 45% (Fig. 8A). Increased phosph- GSK-3 Ser9, an inactive form of GSK-3, was also observed in these cells, further supporting the inhibition of GSK-3 in RMICs by LiCl. Total GSK-3 remained unaltered (Fig. 8B).
DISCUSSION
Lithium is commonly prescribed for treatment of bipolar illness and Alzheimer's disease (10, 24, 34). Its use is routinely associated with polyuria due to NDI. Several studies have addressed the potential mechanisms underlying lithium-induced NDI. Previous studies have focused on direct effects of lithium on the collecting duct, demonstrating impaired V2-receptor responsiveness of the collecting duct associated with decreased vasopressin-stimulated cAMP generation (6, 7). Lithium-treated animals also show reduced expression of the water channels aquaporin-2 and -3, which are critical mediators of epithelial water permeability in the collecting ducts (5, 20, 23). Other studies have demonstrated reduced protein abundance of collecting duct urea transporters A1 and B and reduced vasopressin-stimulated phosphorylation of urea transporter A1 following LiCl treatment (18). The preceding findings demonstrate that lithium treatment alters collecting duct function in vivo.
Collecting duct function may be significantly affected by paracrine factors elaborated by adjacent RMICs, which are a major site of COX2 expression in the medulla (12, 13). The present studies show that lithium increases both total renal medullary COX2 expression and medullary interstitial cell COX2 both in vivo and in vitro. Importantly, COX2 inhibition significantly reduced polyuria following lithium treatment, consistent with a role for interstitial cell COX2-derived prostaglandins in the pathogenesis of lithium-induced NDI.
The potential for a paracrine effect of interstitial cell-derived PGE2 is supported by several previous studies. Cultured interstitial cells produce abundant PGE2 via COX2 (12). Furthermore, basolateral PGE2 potently suppresses ADH-stimulated hydraulic conductivity in isolated, perfused renal collecting ducts (16), mainly through increased intracellular Ca2+ and reduced cAMP levels (16, 17). Previous studies in rats indicate that urinary PGE2 is increased by lithium treatment and that indomethacin, a nonselective COX inhibitor, reduced both urine PGE2 production and urine volume by 80% in lithium-induced NDI (31). These findings are consistent with a role for increased renal PGE2 production in promoting lithium-induced diuresis.
Both COX1 and COX2 prostaglandin synthases are present in the kidney and contribute to renal PGE2 production (26, 29, 32). Of the two COX isoforms identified, COX1 is regulated by a housekeeping-like promoter (19). It is constitutively expressed in most tissues and is particularly abundant in the renal collecting duct (11, 14). In the kidney, the inducible COX2 isoform is also highly expressed but is restricted to the macula densa and RMICs (11, 15, 35). The present studies demonstrate that LiCl treatment markedly increased renal medullary interstitial COX2 expression but did not alter COX1 expression. Furthermore, the present studies of COX1/ mice definitively exclude a role for COX1 in lithium-induced polyuria and increased urinary PGE2 excretion. While COX1 deficiency failed to prevent LiCl-induced polyuria, COX2-selective inhibition prevented polyuria induced by LiCl. LiCl also increased urinary PGE2 excretion in COX1-deficient mice, and this was abolished by a COX2 inhibitor. These studies suggest that increased RMIC COX2 expression is a major source of lithium-induced urinary PGE2.
LiCl administration is reported to stimulate ADH release from the central nervous system and thus increase plasma ADH levels (31). In turn, ADH may directly stimulate renal prostaglandin release (1) or increase medullary tonicity and secondarily increase renal medullary COX2 expression (37). To examine whether increased COX2 expression following lithium treatment is mediated by an indirect effect of ADH or from a direct intrarenal effect of lithium, rats genetically deficient in ADH (Brattleboro rats) were studied. The results indicate that LiCl treatment also increased renal medullary COX2 expression in polyuric Brattleboro rats (Fig. 4). These results show that lithium induced COX2 expression in the absence of ADH. Furthermore, LiCl treatment did not further increase daily urine output in Brattleboro rats, so the lithium-associated increase in COX2 expression is unlikely to be secondary to the polyuric state. The present studies are consistent with increased COX2 expression resulting from a direct effect of LiCl on RMICs. This conclusion is further supported by the cell culture studies showing that lithium treatment of cultured mouse RMICs increases COX2 expression.
In vitro studies indicate that COX2 expression in cultured rabbit RMICs is suppressed by GSK-3 activity (16), and lithium is known to be a potent GSK-3 inhibitor (17). In the present studies, lithium was demonstrated to reduce renal GSK-3 activity both in the mouse kidney and in cultured RMICs. Suppression of renal GSK-3 activity temporally correlated with increased renal COX2 expression. Thus constitutive GSK-3 activity may tonically suppress renal COX2 expression in vivo as well as in vitro. Since LiCl did not change renal COX1 expression and COX1 gene disruption did not prevent lithium-induced polyuria, COX2 appears to be a primary cause of polyuria and source of increased urinary PGE2 production following LiCl treatment.
It is well documented that prostaglandins exert biological effects via cell-surface G protein-coupled receptors (9). PGE2 has four cell-surface receptors, including EP1, EP2, EP3, and EP4. Importantly, both basolateral EP1 and EP3 receptors have been detected in renal collecting ducts, where they suppress ADH-stimulated water absorption (4). These findings are consistent with a model whereby increased renal interstitial cell COX2 produces increased PGE2, activating adjacent basolateral collecting duct EP1 and EP3 receptors, thus reducing water reabsorption by the collecting duct. This action may contribute to inhibition of collecting duct cAMP generation and decreased aquaporin-2 expression (12, 14), resulting in polyuria (Fig. 9).
LiCl likely has additional effects on collecting duct function, since delayed treatment with the COX2 inhibitor SC-58236 failed to completely restore urine volume and urine osmolality to control levels. Lithium has been previously been shown to directly inhibit ADH action in isolated, perfused collecting duct tubules (8). The mechanism of action could involve collecting duct GSK-3, since this enzyme is also expressed in cultured cells from this segment (unpublished observations).
In summary, LiCl-induced polyuria in mice is accompanied by increased renal COX2 expression and increased urine PGE2 excretion. Renal GSK activity progressively decreased in these animals over a time course matching increased COX2 expression (Fig. 8). Since COX2 selective inhibition significantly reduced this polyuria, these findings support an important paracrine role for interstitial cell-associated COX2 activity in lithium-associated polyuria. We further conclude that COX2 inhibition could provide a therapeutic option in the treatment of LiCl-induced diabetes insipidus in clinical practice.
GRANTS
These studies were supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-065024 and a Discovery Grant from Vanderbilt University to C.-M. Hao, a Veterans Merit Award and 2P50-DK-39261 to M. D. Breyer and R. C. Harris, and a South East Affiliate Postdoctoral fellowship grant from the American Heart Association to R. Rao.
ACKNOWLEDGMENTS
We thank Drs. Jianping Jin and Zhonghua Qi for help with COX1/ mice. We also thank Drs. Li Zhang and Gilbert Moeckel for assistance with the mouse medullary interstitial cells.
FOOTNOTES
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
REFERENCES
Beck TR, Hassid A, and Dunn MJ. The effect of arginine vasopressin and its analogs on the synthesis of prostaglandin E2 by rat renal medullary interstitial cells in culture. J Pharmacol Exp Ther 215: 1519, 1980.
Bosetti F and Weerasinghe GR. The expression of brain cyclooxygenase-2 is down-regulated in the cytosolic phospholipase A2 knockout mouse. J Neurochem 87: 14711477, 2003.
Boton R, Gaviria M, and Batlle DC. Prevalence, pathogenesis, and treatment of renal dysfunction associated with chronic lithium therapy. Am J Kidney Dis 10: 329345, 1987.
Breyer MD and Breyer RM. Prostaglandin E receptors and the kidney. Am J Physiol Renal Physiol 279: F12F23, 2000.
Christensen BM, Marples D, Kim YH, Wang W, Frkir J, and Nielsen S. Changes in cellular composition of kidney collecting duct cells in rats with lithium-induced NDI. Am J Physiol Cell Physiol 286: C952C964, 2004.
Christensen S, Kusano E, Yusufi AN, Murayama N, and Dousa TP. Pathogenesis of nephrogenic diabetes insipidus due to chronic administration of lithium in rats. J Clin Invest 75: 18691879, 1985.
Cogan E, Nortier J, and Abramow M. Impaired hydroosmotic response to vasopressin of cortical collecting tubules from lithium-treated rabbits. Pflügers Arch 416: 694703, 1990.
Cogan E, Svoboda M, and Abramow M. Mechanisms of lithium-vasopressin interaction in rabbit cortical collecting tubule. Am J Physiol Renal Fluid Electrolyte Physiol 252: F1080F1087, 1987.
Coleman RA, Smith WL, and Narumiya S. International Union of Pharmacology classification of prostanoid receptors: properties, distribution, and structure of the receptors and their subtypes. Pharmacol Rev 46: 205229, 1994.
Eldar-Finkelman H. Glycogen synthase kinase 3: an emerging therapeutic target. Trends Mol Med 8: 126132, 2002.
Guan Y, Chang M, Cho W, Zhang Y, Redha R, Davis L, Chang S, DuBois RN, Hao CM, and Breyer M. Cloning, expression, and regulation of rabbit cyclooxygenase-2 in renal medullary interstitial cells. Am J Physiol Renal Physiol 273: F18F26, 1997.
Hao CM, Komhoff M, Guan Y, Redha R, and Breyer MD. Selective targeting of cyclooxygenase-2 reveals its role in renal medullary interstitial cell survival. Am J Physiol Renal Physiol 277: F352F359, 1999.
Hao CM, Yull F, Blackwell T, Komhoff M, Davis LS, and Breyer MD. Dehydration activates an NF-kappaB-driven, COX2-dependent survival mechanism in renal medullary interstitial cells. J Clin Invest 106: 973982, 2000.
Harris RC and Breyer MD. Physiological regulation of cyclooxygenase-2 in the kidney. Am J Physiol Renal Physiol 281: F1F11, 2001.
Harris RC, McKanna JA, Akai Y, Jacobson HR, Dubois RN, and Breyer MD. Cyclooxygenase-2 is associated with the macula densa of rat kidney and increases with salt restriction. J Clin Invest 94: 25042510, 1994.
Hebert RL, Jacobson HR, and Breyer MD. PGE2 inhibits AVP-induced water flow in cortical collecting ducts by protein kinase C activation. Am J Physiol Renal Fluid Electrolyte Physiol 259: F318F325, 1990.
Hebert RL, Jacobson HR, and Breyer MD. Prostaglandin E2 inhibits sodium transport in rabbit cortical collecting duct by increasing intracellular calcium. J Clin Invest 87: 19921998, 1991.
Klein JD, Gunn RB, Roberts BR, and Sands JM. Down-regulation of urea transporters in the renal inner medulla of lithium-fed rats. Kidney Int 61: 9951002, 2002.
Kraemer SA, Meade EA, and DeWitt DL. Prostaglandin endoperoxide synthase gene structure: identification of the transcriptional start site and 5'-flanking regulatory sequences. Arch Biochem Biophys 293: 391400, 1992.
Kwon TH, Laursen UH, Marples D, Maunsbach AB, Knepper MA, Frkir J, and Nielsen S. Altered expression of renal AQPs and Na+ transporters in rats with lithium-induced NDI. Am J Physiol Renal Physiol 279: F552F564, 2000.
Lal G, Ash C, Hay K, Redston M, Kwong E, Hancock B, Mak T, Kargman S, Evans JF, and Gallinger S. Suppression of intestinal polyps in Msh2-deficient and non-Msh2-deficient multiple intestinal neoplasia mice by a specific cyclooxygenase-2 inhibitor and by a dual cyclooxygenase-1/2 inhibitor. Cancer Res 61: 61316136, 2001.
Langenbach R, Morham SG, Tiano HF, Loftin CD, Ghanayem BI, Chulada PC, Mahler JF, Lee CA, Goulding EH, and Kluckman KD. Prostaglandin synthase 1 gene disruption in mice reduces arachidonic acid-induced inflammation and indomethacin-induced gastric ulceration. Cell 83: 483492, 1995.
Marples D, Christensen S, Christensen EI, Ottosen PD, and Nielsen S. Lithium-induced downregulation of aquaporin-2 water channel expression in rat kidney medulla. J Clin Invest 95: 18381845, 1995.
Martinez A, Castro A, Dorronsoro I, and Alonso M. Glycogen synthase kinase 3 (GSK-3) inhibitors as new promising drugs for diabetes, neurodegeneration, cancer, and inflammation. Med Res Rev 22: 373384, 2002.
McKanna JA and Zhang MZ. Immunohistochemical localization of lipocortin 1 in rat brain is sensitive to pH, freezing, and dehydration. J Histochem Cytochem 45: 527538, 1997.
Murakami M, Naraba H, Tanioka T, Semmyo N, Nakatani Y, Kojima F, Ikeda T, Fueki M, Ueno A, Oh S, and Kudo I. Regulation of prostaglandin E2 biosynthesis by inducible membrane-associated prostaglandin E2 synthase that acts in concert with cyclooxygenase-2. J Biol Chem 275: 3278332792, 2000.
Rao R, Hao CM, and Breyer MD. Hypertonic stress activates glycogen synthase kinase 3-mediated apoptosis of renal medullary interstitial cells, suppressing an NFB-driven cyclooxygenase-2-dependent survival pathway. J Biol Chem 279: 39493955, 2004.
Ryves WJ and Harwood AJ. Lithium inhibits glycogen synthase kinase-3 by competition for magnesium. Biochem Biophys Res Commun 280: 720725, 2001.
Schneider A, Zhang Y, Zhang M, Lu WJ, Rao R, Fan X, Redha R, Davis L, Breyer RM, Harris R, Guan Y, and Breyer MD. Membrane-associated PGE synthase-1 (mPGES-1) is coexpressed with both COX-1 and COX-2 in the kidney. Kidney Int 65: 12051213, 2004.
Spencer AG, Woods JW, Arakawa T, Singer II, and Smith WL. Subcellular localization of prostaglandin endoperoxide H synthases-1 and -2 by immunoelectron microscopy. J Biol Chem 273: 98869893, 1998.
Sugawara M, Hashimoto K, and Ota Z. Involvement of prostaglandin E2, cAMP, and vasopressin in lithium-induced polyuria. Am J Physiol Regul Integr Comp Physiol 254: R863R869, 1988.
Tanioka T, Nakatani Y, Semmyo N, Murakami M, and Kudo I. Molecular identification of cytosolic prostaglandin E2 synthase that is functionally coupled with cyclooxygenase-1 in immediate prostaglandin E2 biosynthesis. J Biol Chem 275: 3277532782, 2000.
Walker RG. Lithium nephrotoxicity. Kidney Int Suppl 42: S93S98, 1993.
Woodgett JR. Judging a protein by more than its name: GSK-3. Sci STKE 2001: RE12, 2001.
Yang T, Singh I, Pham H, Sun D, Smart A, Schnermann JB, and Briggs JP. Regulation of cyclooxygenase expression in the kidney by dietary salt intake. Am J Physiol Renal Physiol 274: F481F489, 1998.
Zhang MZ, Harris RC, and McKanna JA. Regulation of cyclooxygenase-2 (COX-2) in rat renal cortex by adrenal glucocorticoids and mineralocorticoids. Proc Natl Acad Sci USA 96: 1528015285, 1999.
Zhang MZ, Sanchez Lopez P, McKanna JA, and Harris RC. Regulation of cyclooxygenase expression by vasopressin in rat renal medulla. Endocrinology 145: 14021409, 2004.