INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

IN THE KIDNEY, prostaglandins (PGs) are important mediators of vascular tone and salt and water homeostasis and are involved in the mediation and/or modulation of hormonal action (37). PGs arise from enzymatic metabolism of free arachidonic acid by cyclooxygenase (COX), the enzyme responsible for the initial rate-limiting metabolism of arachidonic acid to PGG₂ and subsequently to PGH₂, followed by further metabolism by specific synthetases. Two isoforms of COX have been identified, COX-1 and COX-2. COX-1 is expressed constitutively in the kidney and is localized predominantly in mesangial cells, arteriolar endothelial cells, parietal epithelial cells of Bowman's capsule, and cortical and medullary collecting ducts (21, 44). It has been determined that COX-1 is not primarily responsible for the increased prostanoid production in inflammatory states nor is it glucocorticoid sensitive (35). In the kidney, COX-2, the "inducible" COX, is localized in macula densa cells and surrounding cortical thick ascending limb (cTAL) cells and in a subset of medullary interstitial cells near the papillary tip in normal adult kidney (21, 44). Macula densa/cTAL expression increases in high renin states, such as dietary salt depletion (21, 25), angiotensin-converting enzyme (ACE) inhibitor treatment (11, 50), diuretic administration (34), or experimental renovascular hypertension (22, 48).

Prostanoids act locally via specific transmembrane G protein-coupled receptors. In the kidney cortex, PGE₂ is the major PG produced (7). Four EP receptor subtypes have been cloned (3, 31). EP₁ and EP₃ receptor mRNA expression are detected in the collecting duct and thick limb, respectively (17, 45), whereas the EP₄ receptor is widely expressed compared with other EP receptors, with the highest concentrations in the glomerulus (6, 8). The human EP₂ receptor subtype protein was detectable only in the media of arteries and arterioles. Although the EP₂ receptor exhibits a low level expression in the kidney, recent studies demonstrated that the EP₂ receptor may be a mediator of arterial dilatation and salt-sensitive hypertension (30).

In previous studies, we determined that administration of a selective COX-2 inhibitor prevented increases in ACE inhibitor (ACEI)-mediated renin release (12). We also found that ACEI failed to increase renin in mice with genetic deletion of COX-2. Others also determined that COX-2 inhibition blunted renin expression in response to dietary sodium deprivation (19, 20, 28, 51). However, this issue remains controversial because other groups do not report changes in renin expression in response to COX-2 inhibitors (23, 29, 34, 40). To further investigate the role of COX metabolites in regulation of renin expression and release, the current studies used mice with genetic deletion of the COX-1 gene to investigate whether COX-1 is involved in mediation of ACEI-induced renin release. We also studied the potential role of the EP₂ receptor in mediation of renin release.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals and genotyping. Mice with genetic deletion of the COX-1 gene (COX-1 /) and maintained on a mixed B6/129 background were a generous gift from Robert Langenbach (33). Mice were genotyped by PCR. The employed primer pairs were IMR013 (CTTGGGTGGAGAGGCTATTC) and IMR014 (AGGTGAGATGACAGGAGATC), and the specific mouse COX-1 primers were (GAGAAGGAGATGGCTGCTGAGTTG) and (AGGTGAGATGACAGGAGATC). PCR was performed for 35 cycles at 95°C for 30 s, 62°C for 45 s, and 72°C for 90 s, followed by a 10-min extension at 72°C. Because the neocassette replaced a portion of the 5' exon 11 and surrounding introns, the neoprimers 0IMR13 and 0IMR14 amplified a 280-bp product in homozygous or heterozygous mice, whereas the specific COX-1 primers generated a 1,016-bp product from the wild-type and heterozygous mice (Fig. 1A). The EP₂ receptor-deleted mice were maintained on a BALB/c background as previously described (30). Mice were genotyped by Southern hybridization with a specific 3' probe. For Southern analysis, DNA samples were digested with XbaI, electrophoresed in agarose gels, and transferred to nylon transfer membranes. The blots were hybridized as described (12). Wild-type (+/+), heterozygous (+/), and homozygous (/) knockout offspring are indicated in Fig. 1B.

View larger version (36K): [in this window] [in a new window]   Fig. 1.   Genotyping of cyclooxygenase (COX)-1 and EP2 receptor / mice. A: PCR screen for COX-1 / mice. Genomic DNA was amplified as described in MATERIALS AND METHODS. PCR was performed for 35 cycles at 95°C for 30 s, 62°C for 45 s, and 72°C for 90 s, followed by a 10-min extension at 72°C. Because the neocassette replaced some of the 5' exon 11 and surrounding introns, the neo primers 0IMR13 and 0IMR14 amplified a 280-bp product in homozygous or heterozygous mice, whereas the specific COX-1 primers generated a 1,016-bp product from the wild-type and heterozygous mice. B: Southern blotting for EP2 receptor / mice. DNA samples were digested with XbaI and then hybridized with 3' probe. A 7.3-kb band was recognized in wild-type (+/+) mice, whereas a 4.5-kb band was recognized in / mice.

1

1

RNA extraction and Northern blotting. Kidney RNA was extracted by the acid guanidium thiocyanate-phenol chloroform method (11). RNA samples were electrophoresed in denatured agarose gel, transferred to nitrocellulose membranes, and hybridized with a 1.4-kb ³²P-labeled cDNA fragment of rat renin (48). The membranes were then stripped and rehybridized with glyceraldehyde-3-phosphate dehydrogenase (GAPDH).

Immunohistochemistry. Under deep anesthesia with Nembutal (70 mg/kg ip), mice were perfused in situ with saline containing 0.02% sodium nitrite and heparin (10 U/ml) and fixed by perfusion for renin immunohistochemistry with 3.7% formaldehyde, 1.4% lysine, 0.01 M sodium metaperiodate, 0.04 M sodium phosphate, and 1% acetic acid. The kidneys were then dehydrated with a graded series of ethanols and embedded in paraffin. Sections (4-µm thick) were mounted on glass slides and immunostained with polyclonal rabbit anti-renin antiserum (1:6,000 dilution), a generous gift from Prof. T. Inagami, Vanderbilt University. Vectastain ABC-Elite (Vector, Burlingame, CA) was used to localize the primary antibody with a chromogen of oxidized diaminobenzidine, followed by a light counterstain with toluidine blue.

Renin activity. Animals were killed by inhalation anesthesia with isoflurane (Baxter Pharmaceutical Products, Deerfield, IL). Blood was collected in EDTA (1 mg/ml blood) on ice at the time of death. The plasma was separated and frozen at 20°C until assay. For measurement of renal tissue renin concentration, the kidneys were homogenized in 0.1 M Tris · HCl, pH 7.4, containing (in mM) 3.4 8-hydroxyquinolone sulfate, 0.25 EDTA, 0.1 phenylmethysulfonyl fluoride, 1.6 dimercaprol, 5 sodium tetrathionate, and 0.1% Triton X-100. The concentration of protein was determined with a bicinchoninic acid protein assay kit (Pierce, Rockford, IL). After centrifugation of the homogenates, the supernatant was incubated for 1 h with excess exogenous renin substrate (plasma obtained from mice nephrectomized 48 h before collection). Renin was analyzed by radioimmunoassay with an ¹²⁵I-ANG I kit (New England Nuclear).

Statistical analysis. All values are presented as means ± SE. ANOVA and Bonferroni t-tests were used for statistical analysis, and differences were considered significant when P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Plasma renin activity (PRA) was measured at the time of death (±7 days of captopril treatment). Captopril significantly increased PRA in COX-1 +/+ (9.3 ± 2.2 vs. 50.1 ± 10.9 ng ANG I · ml¹ · h¹), +/ (6.6 ± 1.4 to 37.2 ± 2.1 ng ANG I · ml¹ · h¹), and / (13.7 ± 1.5 vs. 43.9 ± 6.6 ng ANG I · ml¹ · h¹) mice (n = 6; P < 0.05 for all groups comparing with and without captopril administration; Fig. 2A). The COX-2-specific inhibitor SC-58236 did not alter basal cortical COX-2 mRNA expression in control wild-type mice (Fig. 2B). However, addition of SC-58236 to the ACEI-treated animals led to comparable inhibition of the increases in PRA in all three groups (+/+: 7.3 ± 0.6, +/: 7.2 ± 1.2, /: 8.0 ± 0.9 ng ANG I · ml¹ · h¹; Fig. 2A).

View larger version (30K): [in this window] [in a new window]   Fig. 2.   COX-1 knockout mice. A: plasma renin activity (PRA) in response to captopril treatment. Animals were administered captopril (400 mg/l) in drinking water for 7 days (n = 6; ** P < 0.01). B: similar expression of COX-2 mRNA in renal cortex of wild-type B/6/129 mice with or without 7-day treatment with the COX-2 inhibitor SC-58236. Lanes 1 and 3: control; lanes 2 and 4: SC-58236. C: effect of the captopril on renal renin mRNA expression. Animals were treated as described in A. Renin mRNA was detected by Northern blotting and normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA expression. Inset: representative experiment. Lanes 1, 4, and 7: without captopril treatment; lanes 2, 5, and 8: with captopril treatment; lanes 3, 6, and 9: with combined captopril and SC-58236 treatment. +/+: Lanes 1-3; +/: lanes 4-6; /: lanes 7-9 (n = 6; ** P < 0.01). D: renal renin concentration (RRC) in response to captopril treatment. Animals were treated as described in A. Renal renin activity was determined as described in MATERIALS AND METHODS (n = 6; ** P < 0.01).

Renal renin mRNA was normalized with GAPDH and expressed as fold of control (+/+ mice without treatment). In studies investigating COX-1 knockouts, there were no significant differences in the basal level of renin mRNA among the three genotypes (1.00 ± 0.03 fold of control in +/ and 1.00 ± 0.06 fold of control in / mice, respectively; n = 6; not significant). Captopril treatment increased renin mRNA significantly in all of the three genotypes (2.4 ± 0.2 fold of control in +/+, 2.4 ± 0.2 fold of control in +/, and 2.1 ± 0.2 fold of control in /; n = 9; P < 0.01), and no significant differences were found among the groups (Fig. 2C). SC-58236 prevented increases in renal renin mRNA expression in all three groups (1.2 ± 0.1 fold of control in +/+, 1.2 ± 0.2 fold of control in +/, and 1.2 ± 0.2 fold of control in /; n = 5; P < 0.01; Fig. 2C).

Renal renin concentration (RRC) was also significantly elevated by the administration of captopril (+/+: 11.8 ± 1.7 to 35.3 ± 3.9, +/: 7.6 ± 0.9 to 27.6 ± 3.2, and / 13.0 ± 3.1 vs. 27.8 ± 2.7 ng ANG I · mg protein¹ · h¹; n = 6; P < 0.05 with or without captopril). There were no significant differences in either PRA or RRC among the three phenotypes (Fig. 2C). SC-58236 prevented increases in RRC (+/+: 9.1 ± 0.9, +/: 8.7 ± 0.6, and /: 9.6 ± 0.5 ng ANG I · mg protein¹ · h¹; Fig. 2C).

Immunoreactive (ir) renin was localized to the juxtaglomerular (JG) cells in all three groups of mice (Fig. 3). In agreement with the above results, captopril administration increased ir-renin expression (Fig. 3, B, E, and H) and SC-58236 equally blunted ACEI-induced increases in renin expression in wild-type heterozygotes and COX-1 knockout mice (Fig. 3, C, F, and I).

View larger version (178K): [in this window] [in a new window]   Fig. 3.   Immunoreactive renin expression in the kidney of COX-1 / mice. A, B, C: +/+ without (A), with (B) captopril, or with both captopril and SC-58236 treatment (C). D, E, F: +/ without (D), with (E) captopril, or with the combined captopril and SC-58236 treatment (F). G, H, I: / without (G), with (H) captopril, or with both (I). Arrows in B, E, and H indicate afferent arteriole proximal to juxtaglomerular region, with evidence of renin upregulation in all 3 genotypes. SC-58236 reversed the elevated renin expression (image width = 700 µm).

Determination of renin expression was also performed in EP₂ +/+ and / mice. Captopril led to similar increases of PRA in EP₂ +/+ and EP₂ / (14.8 ± 1.4 to 39.0 ± 6.7 ng ANG I · ml¹ · h¹; n = 5; P < 0.05 vs. 13.9 ± 2.9 to 38.3 ± 7.5 ng ANG I · ml¹ · h¹; n = 5; P < 0.05). The COX-2 inhibitor SC-58236 blocked ACEI-induced elevations in PRA in EP₂ / mice (+/+: 13.1 ± 2.9 ng ANG I · ml¹ · h¹; /: 13.5 ± 2.5 ng ANG I · ml¹ · h¹; n = 5; P < 0.05; Fig. 4A). Renal renin mRNA levels were also upregulated to the same extent in both groups by the administration of captopril (+/+: 2.6 ± 0.3 fold control; /: 2.7 ± 0.6 fold control; n = 4; P < 0.05 with or without captopril), and this increase was prevented by the addition of SC-58236 (+/+: 1.1 ± 0.1 fold control; /: 1.3 ± 0.3 fold control; n = 4; P < 0.05; Fig. 4B). Figure 4C indicates the results of RRC (basal level/captopril/additional SC-58236 in +/+: 11.0 ± 0.8/24.7 ± 1.7/9.8 ± 0.4 ng ANG I · mg protein¹ · h¹; /: 9.8 ± 0.7/21.1 ± 3.2/9.3 ± 0.4 ng ANG I · mg protein¹ · h¹; n = 5; P < 0.05, captopril group compared with control or additional SC-58236-treated group). Correspondingly, immunohistochemistry indicated that captopril administration led to comparable increases in renin expression in JG cells of wild-type and EP-2 knockout mice (Fig. 5, B and E), and SC-58236 blunted this increase in both groups (Fig. 5, C and F).

View larger version (24K): [in this window] [in a new window]   Fig. 4.   EP2 receptor / mice. A: PRA in response to captopril treatment. Animals were treated as described in Fig. 2 (n = 5; ** P < 0.01). B: effect of the captopril on renal renin mRNA expression. Renin mRNA was detected by Northern blotting and normalized to GAPDH mRNA expression (n = 4; * P < 0.05). Inset: representative experiment. Lane 1: +/+ without captopril treatment; lane 2: +/+ with captopril treatment; lane 3: +/+ with combined captopril and SC-58236 treatment; lane 4: / without captopril; lane 5: / with combined captopril and SC-58236 treatment; and lane 6: / with captopril. C: RRC in response to captopril treatment. Animals were treated as described in Fig. 2A. Renal renin activity was determined as described in MATERIALS AND METHODS (n = 5; ** P < 0.01).

View larger version (174K): [in this window] [in a new window]   Fig. 5.   Immunoreactive renin expression in the kidney of EP2 receptor / mice. A, B, C: +/+ without (A), with (B) captopril, or with both captopril and SC-58236 treatment (C). D, E, F: / without (D), with (E) captopril, or with both (F). Arrows in B and E indicate afferent arteriole proximal to juxtaglomerular region, with evidence of renin upregulation in both wild-type and null mice (image width = 700 µm).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

In the adult mammalian kidney, renin is synthesized by the JG cells, a group of myoepithelioid cells located in the afferent arteriole at the entrance to the glomerulus. The macula densa, a specialized segment of thick ascending loop of Henle located between the afferent and efferent arterioles, participates in regulation of JG renin secretion as well as tubuloglomerular feedback. Previous studies using nonselective COX inhibitors demonstrated a role for macula densa/cTAL-derived prostanoids in regulation of renin secretion and expression (16, 26, 41, 42, 47).

Administration of either ACEIs or AT₁-receptor antagonists results in increases in both renin mRNA and ir protein in the kidney in the absence of any detectable alteration in intravascular volume or renal hemodynamics (9, 15, 18). In previous studies, we determined that administration of an ACEI or an AT₁-receptor blocker led to significant increases in macula densa/cTAL COX-2 expression (11, 12). Furthermore, in rats treated with ACEIs, elevations in plasma and kidney renin were significantly inhibited by simultaneous treatment with a selective COX-2 inhibitor (11). In additional studies, we found that increases in renal renin expression in response to ACEI were significantly blunted in COX-2 knockout mice.

Studies in mice on sodium-deficient diets also indicated that the COX-2-selective inhibitor NS-398 would prevent renal renin expression (20); similar findings were observed in COX-2 / mice (51). Similarly, in humans on a sodium-deficient diet, PRA was significantly blunted by addition of the COX-2 inhibitor rofecoxib (28).

However, in contrast to the above studies, other investigators reported that treatment with COX-2-selective inhibitors did not affect renal renin expression and/or secretion. Harding et al. (19) reported that in rats, NS-398 decreased low salt-induced PRA but did not affect renal renin content. Kammerl et al. (29) also reported that rofecoxib decreased PRA induced by low salt but not low salt plus an ACEI, and COX-2 inhibition did not alter renal renin content in any condition. This same group of Kurtz and co-workers (23) also reported that celecoxib did not affect increased renal renin mRNA levels induced by the AT₁-receptor antagonist candesartan. Furthermore, in isolated perfused kidneys, they reported that NS-398 blocked increased renin secretion induced by bumetanide but not low-salt or ACE inhibition (10).

These divergent findings with COX-2 inhibitors coupled with previous reports by Kurtz and co-workers demonstrating that the nonselective, nonsteroidal anti-inflammatory drug indomethacin blunted increases in renal renin expression by either bumetanide or ACEIs (41, 43) raised the question of whether COX-1-derived prostanoids might be involved in regulation of renin. SC-58236 is a highly selective COX-2 inhibitor (39), but we could not absolutely rule out the possibility that there was also some inhibition of renal COX-1 in our previous studies. Furthermore, because COX-2 knockout mice have known renal developmental abnormalities (13, 32, 36), we could not completely rule out the possibility that the observed failure to increase renin production in response to ACEI was due to structural or functional abnormalities not directly related to the absence of COX-2 expression in macula densa/cTAL. To test definitively a potential role for COX-1 in ACEI-mediated increases in renin, we used mice with genetic deletion of the COX-1 gene. The fact that captopril increased plasma renin and renal renin mRNA and activity levels to the same extent in the COX-1 knockout and wild-type mice and the increased renin expression was effectively blocked by the selective COX-2 inhibitor demonstrates unequivocally that ACEI-mediated increases in renin expression are mediated by prostanoids derived from COX-2 rather than COX-1.

In agreement with the current results, Traynor et al. (46) found that in the isolated perfused glomerular preparation, the presence of the COX-2-selective inhibitor NS-398 but not the COX-1-selective inhibitor valerylsalicylate inhibited macula densa-mediated stimulation of renin secretion in response to a reduction in tubular perfusate NaCl concentrations. Furthermore, Athirakul et al. (1) also recently reported that renin expression was increased in COX-1 / mice in response to a sodium-deficient diet.

Although we previously noted that SC-58236 decreased cortical COX-2 expression in models of renovascular hypertension (48) and renal ablation (49), administration of this long-acting COX-2 inhibitor for 7 days to normal mice did not measurably affect basal renal cortical COX-2 mRNA expression. This suggests that the observed effects on COX-2 expression in the pathophysiological conditions may have been indirect, i.e., secondary to alteration of systemic or intrarenal stimuli for increased cortical COX-2 expression. However, further studies will be required to assess this hypothesis.

PGE₂ and PGI₂ are well-described mediators of renal renin expression and secretion (14, 24, 27). PG-mediated stimulation of JG renin expression and secretion has been shown to be due to activation of adenylate cyclase and an increase in intracellular cAMP (2, 27). PGE₂ and PGI₂ are both able to stimulate renin release (27). Of the four EP receptor subtypes, both EP₂ and EP₄ have been shown to be coupled to G_s and to stimulate adenylate cyclase activity (3, 31). Because the EP₂ receptor regulates vascular reactivity, we tested whether this receptor subtype might be involved in ACEI-mediated increases in renin release. However, our studies indicate that regulation of renin production and release was not altered in the EP₂ knockout mice, indicating that this EP subtype is not involved in this process. EP₄ mRNA is expressed in the glomerulus and collecting duct and may regulate glomerular tone (4, 5, 8). The current studies suggest that EP₄ may also be involved in regulation of renal renin release. In this regard, EP₄ transcripts in glomeruli were increased twofold by salt deprivation (26). In situ hybridization detected IP receptor mRNA in the interlobular arteries and glomerular arterioles but not in the JG cells, which suggests that the action of prostacyclin on renin release may be indirect (38). Confirmation of the effect of EP₄ receptor and IP receptor on regulation of JG cell renin production and release await further studies.

In conclusion, the current study provides definitive evidence that metabolites of COX-2 rather than COX-1 can mediate or modulate ACEI-induced renin production and release. The fact that EP₂ null mice also demonstrated no alterations in ACEI-induced renin production and release suggests involvement of EP₄ and/or IP.