Jump to
Abstract
Prostaglandins contribute to the regulation of renin synthesis and secretion. We tested the hypothesis that the inducible isoform of prostaglandin G/H synthase, cyclooxygenase-2, contributes to the stimulation of renin synthesis in renovascular hypertension. The expression of cyclooxygenase-2 and renin was investigated in the kidneys of rats with two-kidney, one-clip renovascular hypertension or sham operation. Systolic blood pressure was increased 2 weeks after clipping (153±7 versus 112±4 mm Hg in controls, n=6 each, P<.05) and continued to rise until 4 weeks. Cyclooxygenase-2 mRNA levels were increased in clipped kidneys but remained unchanged or slightly decreased in nonclipped kidneys. Cyclooxygenase-2 protein was expressed mainly in the macula densa and occasionally in distal tubular cells not associated with the macula densa. Two weeks after clipping, the percentage of juxtaglomerular apparatus staining positive for cyclooxygenase-2 was 27.8±3.6 in clipped kidneys, 3.1±0.4 in contralateral kidneys, and 8.0±1.3 in controls; the percentages for immunoreactive renin staining in the afferent arteriole were 33.6±2 in clipped kidneys, 1.9±0.5 in contralateral kidneys, and 12.4±4.0 in controls, respectively. Similar parallel changes in renin and cyclooxygenase-2 staining were observed 4 weeks after clipping. The percentage of cyclooxygenase-2-positive juxtaglomerular apparatus correlated positively with the percentage that was renin positive (r=0.78, P<.05). Double immunostaining showed coexpression of cyclooxygenase-2 and renin protein in the same juxtaglomerular apparatus. Our data are consistent with a role for macula densa cyclooxygenase-2 in the regulation of renin in renovascular hypertension.
- COX = cyclooxygenase
- 2K1C = two-kidney, one-clip renovascular hypertension
- JGA = juxtaglomerular apparatus
Prostaglandin G/H synthase (COX) is a key enzyme in the generation of prostaglandins. There are two isoforms of COX that share the same enzymatic activity but differ markedly in their regulation. COX-1 is constitutively expressed in many tissues whereas COX-2 is considered an inducible isoform transcribed primarily during inflammatory reactions.1 In the kidney, however, COX-2 also exhibits constitutive expression. Gene targeting studies indicated that COX-2 expression is required for the normal development of the kidney.2,3⇓ Harris et al4 first described expression of COX-2 in the macula densa of the distal tubule of the rat. COX-2 expression increased with salt restriction,4 suggesting a role for COX-2 in the macula densa control of renin secretion. In support of this notion, selective inhibition of COX-2 in mice impaired the increased renin content induced by salt restriction.5
The importance of the macula densa mechanism for the control of renin synthesis and release during changes in salt intake is well established.6–9⇓⇓⇓ A different mechanism has been described for the control of renin by altered perfusion pressure, eg, in renovascular hypertension. The renal baroreceptor mechanism, which causes increased renin synthesis and release after renal artery constriction,10 may reside in endothelial cells11 or the renin-secreting cells,12 but the macula densa has usually not been implicated by most authors.6,8,9⇓⇓ Recent observations, however, suggested the possibility that the macula densa mechanism may contribute to the control of renin after renal artery stenosis.13,14⇓
Therefore, we hypothesized that COX-2 expression in the macula densa of rats might contribute to the differential regulation of renin in the clipped and nonclipped kidney in renovascular hypertension. We quantified COX-2 protein expression in the macula densa of rats with renal artery stenosis.
Methods
Induction of Renovascular Hypertension
Male Sprague-Dawley rats (Charles River) were housed in a room maintained at 22±2°C, exposed to a 12-hour dark/light cycle. The animals were allowed unlimited access to chow (no. 1320, Altromin) and tap water. All procedures performed on animals were done in accordance with guidelines of the American Physiological Society and were approved by the local government authorities (Regierung von Mittelfranken, AZ no. 621-2531.3-10/94).
2K1C was induced in rats weighing 150 to 170 g by placing a silver clip of 0.2-mm ID around the left renal artery through a flank incision under ether anesthesia, as previously described.15,16⇓ Control animals underwent the same procedure without placement of the clip. The animals were then followed by weekly measurements of weight and systolic blood pressure by tail-cuff plethysmography under light ether anesthesia. Animals were only included in the 2K1C group if systolic blood pressure was above 150 mm Hg. Animals that failed to thrive or lost weight were excluded.17 Rats were killed, and the kidneys were harvested 2 weeks after clipping (six 2K1C and six control animals) as well as 4 weeks after clipping (six 2K1C and six control animals). After kidney weight was measured, the organs were decapsulated. A part of each kidney was immediately snap frozen on liquid nitrogen for later protein and RNA extraction while a second part was put in methyl-Carnoy solution (60% methanol, 30% chloroform, and 10% glacial acetic acid) for fixation.
Northern Blot Detection of mRNA
Total RNA was extracted using the single step method of Chomczynski18 with the TRI reagent (Molecular Research Center Inc). For Northern blot analysis, 20 μg of total RNA was electrophoresed in a 1% agarose-1.8% formaldehyde gel and transferred to positively charged nylon membranes (Hybond N; Amersham) with 20× SSC overnight. Ethidium bromide staining of the gel and membrane confirmed equal loading and sufficient blotting of the RNA. The RNA was fixed by baking for 2 hours at 80°C.
The blots were probed with a full-length renin cDNA,19 an 1156-bp COX-2 cDNA,20 and a 500-bp GAPDH probe.21 DNA fragments were radiolabeled with [α32P]CTP by random prime labeling22 with a commercially available kit (Amersham). Prehybridization was performed at 42°C for 2 hours with a buffer containing 5× SSC, 0.1% SDS, 5× Denhardt’s solution, 50 mmol/L Na2HPO4 (pH 6.5), 50% deionized formamide, and 250 μg/ml denatured salmon sperm DNA. Hybridization was carried out under the same conditions for 24 hours. The membranes were washed twice with 2× SSC/0.1% SDS at 40°C for 15 minutes and twice with 0.2× SSC/0.1% SDS for 30 minutes at the same temperature. Blots were then exposed to Kodak x-ray films with intensifying screens, and quantitative analysis was performed by densitometric scanning (Bioprofil).
Immunohistochemistry
After overnight fixation in methyl-Carnoy solution, tissues were dehydrated by bathing in increasing concentrations of methanol, followed by 100% isopropanol. After embedding in paraffin, 4-μm sections were cut with a Leitz SM 2000 R microtome (Leica Instruments). After deparaffinization, endogenous peroxidase activity was blocked with 3% H2O2 in methanol for 20 minutes at room temperature. Sections were then layered with the primary antibody and incubated at 4°C overnight. After addition of the secondary antibody (dilution 1:500; biotin-conjugated, goat anti-rabbit immunoglobulin G or rabbit anti-goat immunoglobulin G, respectively), the sections were incubated with avidin-biotinylated horseradish peroxidase complex (Vectastain DAB kit, Vector Laboratory) and exposed to 0.1% diaminobenzidine tetrahydro-chloride and 0.02% H2O2 as a source of peroxidase substrate. Each slide was counterstained with hematoxylin. As a negative control, we used equimolar concentrations of preimmune rabbit or goat immunoglobulin G. For double stainings, serial sections were stained for COX-2 first, followed by blocking of the peroxidase activity and subsequent staining for renin. To detect renin staining, the Vector VIP substrate kit for peroxidase (Vectastain) was used, resulting in a purple staining.
Antibodies
The antibody used to detect rat renin was kindly supplied as a generous gift by Walter Fischli, PhD (Hofmann-La Roche, Basel, Switzerland). The polyclonal antiserum 8914 was raised in rabbit by immunization with recombinant human active renin. Western blot of rat kidney protein (extracted by the method of Chomczynski18 with the TRI reagent) with antiserum 8914 yielded a single band of 35 kD. For immunohistochemical studies in rat kidney sections, the antiserum was used at a dilution of 1:2000. Preliminary experiments demonstrated that the same dilution of the antiserum caused specific stainings of the JGA of mouse and human kidney sections (data not shown). COX-2 was stained with a 1:500 dilution of a commercially available antiserum (M-19, Santa Cruz Biotechnology). The affinity-purified polyclonal antibody was raised in goat. Western blot experiments in extracts from rat mesangial cells showed that the antiserum detects only the inducible 72 kD COX-2 but not COX-1 (data not shown).
Analysis of Data
Quantification of renin or COX-2 expression was performed by counting the number of glomeruli with an adjacent JGA staining positive for COX-2 or renin, respectively. In each kidney, 250 to 450 glomeruli were counted, and the number of positive glomeruli was expressed as a percentage of the total number of glomeruli counted. These percent values were used for statistical analysis. Two-way analysis of variance, followed by posthoc Newman-Keuls test, was used to test significance of differences between groups. A value of P<.05 was considered significant. Pearson correlations were used to test for a correlation between COX-2 and renin expression. The procedures were carried out using the Statistical software (StatSoft). Values are displayed as means±SEM.
Results
Systolic blood pressure as measured by tail-cuff plethysmography was increased 2 weeks after clipping in 2K1C (153±7 versus 112±4 mm Hg in controls, P<.05, n=6 each) and 4 weeks after clipping (190±13 versus 115±11 mm Hg in controls, P<.05, n=6 each). Fig 1 shows representative Northern blots for COX-2 and renin mRNA. As expected, renin mRNA was increased in clipped but decreased in nonclipped kidneys (Fig 1). COX-2 mRNA was barely detectable in control kidneys and in nonclipped kidneys of 2K1C animals. However, an increased COX-2 mRNA signal was detected in clipped kidneys of 2K1C animals (Fig 1).
Figure 1. Northern Blot showing the expression of COX-2 (left) and renin (right) mRNA in clipped (C) and nonclipped (N) kidneys of 2K1C rats and kidneys of sham operated rats (S), 4 weeks after clipping. Upper panels, autoradiographs after hybridization with the respective probes; lower panels, ethidium bromide-stained membrane to demonstrate comparable loading with RNA. The results are representative for five rats from each group.
By immunohistochemical analysis, COX-2 protein was expressed mainly in the macula densa cells of the distal tubule (Fig 2A through 2C). Occasionally, several distal tubular cells not associated with the macula densa also stained for COX-2 (Figs 2D and 3F⇓). Renin staining was confined to the juxtaglomerular cells of the afferent arteriole (Fig 3). In 2K1C animals, COX-2 staining was diminished in the contralateral, nonclipped kidneys but increased in the clipped kidneys. Counting the percentage of glomeruli with adjacent macula densa staining positive for COX-2 confirmed these observations (Fig 4). The altered expression of COX-2 paralleled the well-known pattern of renin staining in the clipped versus nonclipped kidneys (Fig 4). In clipped kidneys, more macula densa-associated cells of the distal tubule stained positive for COX-2 than in contralateral or control kidneys (Figs 2A, 2B, and Fig 3F).
Figure 2. Localization of COX-2 immunoreactivity. Kidney sections from rats 2 weeks after clipping or sham operation. A and B, Localization in the macula densa demonstrated in cross-sections of the distal tubule. Only one macula densa cell stains for COX-2 in a section from a nonclipped kidney of a 2K1C rat (A) whereas many cells per macula densa exhibit COX-2 staining in the clipped kidney (B). C, oblique section of a distal tubule leading to the vascular pole of the glomerulus, demonstrating COX-2 staining in distal tubular cells at or near the macula densa. Section from a clipped kidney. D, COX-2 staining in patches of distal tubular cells not associated with the macula densa. Section from a sham-operated animal. The results are representative for 6 (A through C) or 24 (D) rats, respectively.
Figure 3. Double stainings for immunoreactive COX-2 (brown) and renin (dark purple). Kidney sections from rats 2 weeks after clipping or sham operation. A through D, Serial 4-μm sections from the clipped kidney of a 2K1C animal. Marked COX-2 staining in the distal tubular cells most closely related to the renin-positive cells of the afferent arteriole. E, Section from a sham-operated animal. Immunoreactive COX-2 in macula densa cells close to the renin-positive afferent arteriole. F, Section from the clipped kidney of a 2K1C rat. The macula densa cells and many adjacent distal tubular cells stain for COX-2. Results are representative for four rats from each group.
Figure 4. Bar graphs show the percentage of glomeruli with adjacent macula densa positive for COX-2 and adjacent afferent arteriole positive for renin. Upper panel, 2 weeks after clipping; lower panel, 4 weeks after clipping. Values are mean±SEM from 6 animals each, *denotes significant (P<.05) differences between 2K1C and sham-operated animals. Values from right and left kidneys of sham-operated animals were not different from each other and were therefore pooled.
The percentage of the renin-positive glomeruli per kidney correlated positively with the percentage of COX-2-positive glomeruli (r=0.78, n=48, P<.05). After double staining for COX-2 and renin, we attempted to quantify the actual number of JGAs expressing renin, COX-2, or both. Only glomeruli with clearly identifiable JGA (vascular pole and adjacent distal tubule) were counted. The Table shows that the most prominent change was in the number of either double-negative or double-positive JGA. Further, we performed double stainings on serial sections to determine whether the distal tabular cells most close to the renin-producing cells were positive for COX-2. Fig 3A through 3D shows a representative example. The macula densa cells close to the afferent arteriole stained positive for COX-2.
Percentage of JGA with Immunoreactive Staining for COX-2 and/or Renin
Discussion
Our data show that COX-2 is expressed in the macula densa of the rat. Our main finding was the parallel regulation of renin and COX-2 expression in renovascular hypertension. COX-2 expression in the macula densa clearly increased in poststenotic kidneys, and there was a downregulation in the kidney exposed to high blood pressure. The number of JGA positive for COX-2 increased in the clipped kidney, and there appeared to be an increased number of COX-2-positive cells per macula densa, reminiscent of the recruitment of renin-producing cells in the afferent arteriole.23 Moreover, COX-2 expression correlated with renin expression. COX-2 and renin were coexpressed in the same JGA in the clipped kidneys of 2K1C rats. These data are consistent with a role for COX-2 in the regulation of renin in renovascular hypertension.
Stimulation of renin synthesis and secretion by prostaglandins is long since known.6,8⇓ Prostaglandins E2 and I2 act directly on renin-producing juxtaglomerular cells to stimulate renin mRNA and release.24 In vivo, suppression of endogenous prostanoid formation by nonisozyme-specific COX inhibitors, which do not distinguish between COX-1 and COX-2, impairs the stimulation of renin synthesis and secretion in salt depletion25 and renovascular hypertension.26 The origin of these prostaglandins is unknown, both with regard to the cell type and the COX isoform involved. COX-1 is present in vascular endothelial and smooth muscle cells27 and could thus contribute to local prostanoid formation in the vicinity of renin-producing cells. However, recent reports of COX-2 expression associated with the macula densa4,28⇓ suggested a role for this isoform. Indeed, COX-2 expression in the macula densa increased with salt restriction.4 Moreover, blockade of COX-2 by a specific inhibitor that does not affect COX-1 impaired the macula densa mediated increase of renal renin content in mice.3
Our results confirm the data of Harris et al4 on the localization of COX-2 in the rat macula densa. Unlike Harris et al,4 we often detected more than one COX-2-positive cell per macula densa. Further, we observed patches of distal tubular cells positive for COX-2, which could not be clearly associated with a macula densa. However, we did not observe the strong staining of medullary interstitial cells described by Harris et al.4 Different affinities of the antibodies used or strain differences between rats may contribute to these minor discrepancies. The macula densa localization of COX-2 has also been described in the rabbit,28 and the results obtained by Harding et al5 with a COX-2 specific inhibitor are at least consistent with macula densa associated COX-2 in mice. One study27 reported the absence of COX-2 from the macula densa in human tissue. One could speculate that besides species differences, different characteristics of the antibodies, or a different level of activation of the macula densa mechanism might account for these findings.
The parallel regulation of COX-2 in the macula densa and renin in the afferent arteriole also point to a possible role of the macula densa mechanism in regulating renin synthesis or secretion in renovascular hypertension. Interestingly, Bosse et al14 reported similar findings for macula densa-associated brain-type nitric oxide synthase: the expression of the enzyme was increased in the clipped kidney and decreased in the contralateral kidney. Schricker et al13 recently reported that renal artery stenosis and furosemide to stimulate the macula densa are not additive with regard to the stimulation of renin mRNA. Our results are consistent with a role for macula densa-associated COX-2 in renin-dependent hypertension. Our findings are somewhat limited because we cannot exclude the possibility that the altered expression of COX-2 may be a consequence rather than a cause of the altered renin expression or an epiphenomenon unrelated to the regulation of renin. However, if previous findings concerning the role of prostanoids and COX-2 for the regulation of renin are taken into account, it appears more likely that altered COX-2 expression in renovascular hypertension may in fact contribute to the regulation of renin. Further studies will be necessary to elucidate the role of the macula densa in renovascular hypertension and the role of nitric oxide and prostanoids as mediators of the macula densa mechanism.
Acknowledgments
This study was supported by a grant from the Deutsche Forschungsgemeinschaft (Klinische Forschergruppe “Molekulare Regulations-mechanismen in glomerulären Zellen der Niere”, STE-196/3-1) and by the Förderverein des Institutes für Nephrologie Nürnberg. We gratefully acknowledge the expert technical assistance of Elisabeth Weichmann and Rita Zitzmann. We thank Walter Fischli, PhD, for supplying us with the renin antiserum.
- Received September 16, 1997.
- Revision received October 15, 1997.
- Accepted October 29, 1997.
References
This Issue
Jump to
Article Tools
Share this Article
- Coordinate Expression of Cyclooxygenase-2 and Renin in the Rat Kidney in Renovascular Hypertension
