Abstract
Abstract. Adrenalectomized (ADX) and sham-operated rats received either dexamethasone (DEX) or vehicle. Renal tissue was used for morphologic analysis, assessment of cyclooxygenase-2 (COX-2) protein expression and mRNA accumulation, and quantitation of COX-2 activity. In untreated or shamoperated rats, COX-2 protein was observed in a subset of tubular epithelial cells (<2%), which were located mainly in the cortex. All COX-2-positive cells also expressed Tamm-Horsfall glycoprotein, a highly selective marker for thick ascending limb (TAL) cells. After ADX, >30% of TAL cells expressed COX-2 in a manner consistent with recruitment of COX-2-positive TAL cells toward the medulla. Treatment of ADX rats with DEX reduced the number of COX-2-positive cells to that observed in sham-operated or intact rats. COX-2 mRNA accumulation was increased by ADX and partially attenuated by treatment with DEX. Western blot analysis of cortical microsomes revealed a substantial increase in COX-2 expression in ADX rats, compared with ADX/DEX-treated, sham-operated, or intact rats. The increase in COX-2 protein expression was associated with a twofold increase in prostaglandin E2 formation by cortical microsomes obtained from ADX rats, compared with sham-operated rats. It is concluded that ADX induces expression of enzymatically active COX-2, such that expression occurs in the cortical TAL and proceeds in a defined pattern toward the outer medullary TAL. It is suggested that ADX induces expression of TAL cells that, in the basal state, do not express COX-2 protein.
There are two isoforms of prostaglandin (PG) H synthase, i.e., PGH synthase 1 and PGH synthase 2, which are commonly referred to as cyclooxygenase-1 (COX-1) and COX-2, respectively. The amino acid sequences for COX-1 and COX-2 are similar (approximately 75% homology), and the residues that are important for the catalytic activities of these enzymes are highly conserved (1,2). However, COX-1 and COX-2 differ significantly at amino acids before residue 30, and COX-2 has an 18-amino acid insert close to the carboxyl-terminus that is not present in COX-1 (3). These two enzymes catalyze the same reactions; i.e., arachidonic acid (AA) is converted to PGG2 via the COX reaction, followed by a peroxidase reaction in which the 15-hydroperoxyl group of PGG2 is reduced to the 15-hydroxyl group of PGH2. Specific isomerases metabolize PGH2, in a cell-specific manner, to products including PGE2 and PGI2, which contribute importantly to renal function (4).
COX-1 is constitutively expressed in many cell types, whereas COX-2 gene transcription is induced by mitogens, growth factors, cytokines, and tumor promoters (1,5,6,7). The mRNA encoding COX isoenzymes (constitutive COX-1 and inducible COX-2) has been demonstrated in rat kidneys (8). Moreover, immunohistochemical localization studies have demonstrated that COX-2 is present in the peri-macula densa region and a subset of thick ascending limb (TAL) epithelial cells located in the cortex and outer medulla (9,10).
The regulation of renal COX-2 expression and function in vivo is not completely understood. However, this isoform is expressed during NaCl depletion (9,11,12) and is responsible for renal developmental changes, particularly morphogenesis of the nephron (13,14,15). Chronic dietary salt depletion produced a threefold increase in COX-2 immunoreactivity in cortical TAL cells and cells of the macula densa, which is an extension of the TAL and the renal tubular component of the juxtaglomerular apparatus that includes the glomerular afferent artery and contiguous mesangium (9). Although COX-2 expression is regulated by glucocorticoids in extrarenal cells (16), the effects of glucocorticoids (which have well described effects on renal function) on renal COX-2 expression have not been studied. This study was designed to assess the effects of endogenous glucocorticoids on renal COX-2 expression.
Materials and Methods
Animals
This study was performed using adult male Sprague-Dawley rats (220 to 250 g, n = 48), which were maintained with free access to water and were fed normal rat chow.
Adrenalectomy
Animals (n = 24) were adrenalectomized (ADX) under general anesthesia (17) with a combination of ketamine (25 mg/kg, intraperitoneally) and xylazine (2.5 mg/kg, intraperitoneally) and were maintained with normal rat chow and 0.9% NaCl in the drinking water for 5 d after surgery. They were then randomly assigned to two groups; one group received dexamethasone (DEX) (1 mg/kg per d, subcutaneously) for 2 d (ADX plus DEX group, n = 12) and the other group received saline solution (ADX group, n = 12). DEX (Oradexon, 5 mg/ml; NV Organon, Oss, Holland) was dissolved in sterile saline solution, to yield a 1 mg/ml solution, before administration. Other experimental groups (n = 6 each) consisted of sham-operated rats that were treated with DEX or vehicle, as described above. The animals were maintained in the Pontificia Universidad Catolica de Chile animal care facilities, and the experimental procedures were in accordance with institutional and international standards for the humane care and use of laboratory animals (Animal Welfare Assurance Publication A5427-01, Office for Protection from Research Risks, Division of Animal Welfare, National Institutes of Health). At the appropriate times, the kidneys were removed under anesthesia (ketamine/xylazine); one kidney was used for morphologic analysis and the other kidney for reverse transcription (RT)-PCR, Western blot analysis, and COX activity studies.
Tissue Processing and Immunohistochemical Analyses
The localization of COX-2 protein was evaluated by immunohistochemical analysis using specific antibodies (see below). The cellular origin of COX-2 was assessed by two complementary methods, i.e., immunostaining of consecutive sections with antibodies against COX-2 and Tamm-Horsfall glycoprotein (a specific marker for TAL cells) (18) and sequential double-immunolabeling for COX-2 and Tamm-Horsfall glycoprotein in the same tissue sections. Macrophages, which are a possible source of COX-2, were identified with a monoclonal antibody (ED1) to a cytoplasmic antigen present in monocytes and macrophages (19). The renal tissue samples from the different groups were coded and studied independently, in a blinded manner, by two expert observers who were not aware of the codes. Renal slices (3-mm thick), including the cortex, medulla, and papilla, were fixed by immersion in Bouin's solution for 24 h at room temperature. The tissue was then dehydrated, embedded in Paraplast plus (Monoject Scientific, St. Louis, MO), serially sectioned at 5-μm thickness with a rotatory microtome, mounted on glass slides, and stored until immunostaining.
Single-Immunolabeling in Consecutive Sections
Immunostaining was performed using an indirect immunoperoxidase technique, as described previously, to localize COX-2 and other proteins in normal rat kidneys (10). Briefly, for sequential double-immunolabeling for COX-2 and Tamm-Horsfall glycoprotein, consecutive tissue sections were dewaxed, rehydrated, rinsed in 0.05 M Tris-phosphate-saline (TPS) buffer (pH 7.6) and incubated with the primary antiserum overnight at 22°C, followed by 30-min incubations at 22°C with the secondary antibody (1:20) and then the peroxidase-antiperoxidase (PAP) complex (1:150). Peroxidase activity was detected by incubation of the sections with 0.1% (wt/vol) 3,3′-diaminobenzidine and 0.03% (vol/vol) hydrogen peroxide. When a monoclonal antibody (ED1) was used, incubation with a rabbit antimouse IgG antibody was included as a link between the primary and secondary antibodies. The antisera and PAP complex were diluted in TPS buffer containing 0.25% (vol/vol) Triton X-100 and 0.7% (wt/vol) λ-carrageenan. The sections were rinsed with TPS buffer between incubations, counterstained with hematoxylin, dehydrated and cleared with xylene, and coverslipped.
Double-Immunolabeling in the Same Tissue Sections
This technique was performed with a modification of a previously described procedure (20). Sections were incubated overnight at 22°C with the first primary antibody (anti-murine COX-2), followed sequentially by unlabeled secondary antibody, PAP complex, and diaminobenzidine/hydrogen peroxide for color development, yielding a brown color. Incubation in 3% hydrogen peroxide/methanol for 20 min eliminated any remaining peroxidase activity. Subsequently, the second primary antibody (anti-Tamm-Horsfall glycoprotein) was applied overnight at 22°C, followed by unlabeled secondary antibody, PAP complex, and Vector SG substrate (Vector Laboratories, Burlingame, CA) as the second reagent for color development, yielding a blue reaction product. The sections were dehydrated and cleared with xylene and then coverslipped, without hematoxylin counterstaining. Controls for the immunostaining procedures included preabsorption of the primary antibody with the COX-2 carboxyl-terminal peptide, omission of the primary antibody, and replacement of the primary antibody with normal or preimmune rabbit serum. The tissue sections were observed and photographed using a Nikon Optiphot microscope with a Nikon Microflex UFX IIA photographic system (Nikon Corp., Tokyo, Japan).
Antisera and Chemicals
Specific antibodies (dilution, 1:200 to 1:800) raised against peptides (15 to 29 amino acids) derived from carboxyl-terminal amino acid sequences unique to murine COX-2 were used. Two of the antisera were of rabbit origin. One antiserum was raised in our laboratory, as described (10). The second antiserum, 160106, was purchased from Cayman Chemicals (Ann Arbor, MI) (13,21). A third antiserum (SC-1747, of goat origin) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The specificity of these antisera was previously established by Western blotting and immunohistochemical analyses (10,13,21). The monoclonal antibody against macrophages (ED-1) (19), which was used at a 1:400 dilution, was purchased from Serotec (Oxford, England). Secondary antibodies and the corresponding PAP complexes were purchased from INC Pharmaceuticals-Cappel (Aurora, OH). Triton X-100, 3,3′-diaminobenzidine, λ-carrageenan, Tris-HCl, hydrogen peroxide, phosphate salts, and other chemicals were purchased from Sigma Chemical Co. (St. Louis, MO).
Quantitative Morphologic Analyses
TAL cells were identified by the presence of Tamm-Horsfall glycoprotein and the anatomic arrangement of rays beginning in the cortex (juxtaglomerular apparatus) and ending in the medulla (22). The percentage of TAL cells containing COX-2 was quantitated in consecutive sections stained alternately for Tamm-Horsfall glycoprotein and COX-2. In axially sectioned TAL tubules, COX-2-positive TAL cells were counted and expressed as percentages of the total numbers of TAL cells counted in the consecutive sections, according to the following formula: % COX-2 = number of COX-2-positive TAL cells/total number of TAL cells × 100.
Western Blot Analyses
Renal microsomes were prepared as described previously (23). Renal tissues were lysed, and protein concentrations were determined using a detergent-compatible Bio-Rad protein assay kit (Bio-Rad, Richmond, CA). Sixty micrograms of renal microsomes were mixed with an equal volume of 2× sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis sample buffer (100 mM Tris-HCl, pH 6.8, 200 mM dithiothreitol, 4% SDS, 0.2% bromphenol blue, 20% glycerol) and were boiled for 3 min. The proteins in renal microsomes were separated on 10% SDS-polyacrylamide gels and transferred to either nitrocellulose or polyvinylidene difluoride membranes. Nonspecific sites on the membranes were blocked by incubation in blocking solution (5% nonfat dry milk in Tris-buffered saline-Tween) for 30 min at room temperature. After blocking, membranes were probed with the rabbit polyclonal antibody against murine COX-2 (which was also used for immunohistochemical analyses) for 1 h at room temperature, washed with Tris-buffered saline-Tween, and incubated with horseradish peroxidase (HRP)-conjugated antiserum. Proteins were detected using enhanced chemiluminescence or enhanced chemifluorescence/PhosphorImaging techniques.
Isolation of Total RNA and RT-PCR Analyses
Total RNA was isolated using Trizol reagent (Life Technologies) and precipitation with isopropyl alcohol. A 3-μg aliquot of total RNA was used for cDNA synthesis, using the Superscript preamplification system (Life Technologies), in a 20-μl reaction mixture containing Superscript II reverse transcriptase (200 U/μl) and random hexamers (50 ng/μl). The reaction mixture was incubated at room temperature for 10 min to allow extension of the primers by reverse transcriptase, and was then incubated at 42°C for 50 min, at 70°C for 15 min, and at 4°C for 5 min. An aliquot of the cDNA was then amplified using Taq DNA polymerase (2.5 U) in the presence of sense and antisense primers (1 μM) for murine COX-2 or glyceraldehyde-3-phosphate dehydrogenase (GAPDH). The following primer sets were used: COX-2 sense, TACAAGCAGTGGCAAAGGC; COX-2 antisense, CAGTATTGAGGAGAACAGATGGG; GAPDH sense, ACCACAGTCCATGCCATCAC; GAPDH antisense, TCCACCACCCTGTTGCTGTA. The PCR primer sets for COX-2 and GAPDH amplified fragments with transcript sizes of 304 and 452 bp, respectively. Amplification (35 cycles) was performed with denaturation for 1 min at 94°C, annealing for 1 min at 53°C, and polymerization for 2 min at 72°C, followed by autoextension for 8 min at 72°C. Preliminary experiments were performed to ensure that PCR was performed within the linear range of product formation.
Assessment of COX Activity
COX activity was assessed in renal cortical microsomes (50 μg) that had been incubated for 10 min with AA (7 μM) in a Tris buffer (pH 8.0) containing tryptophan (5 mM), hematin (1 μM), glutathione (1 mM), and epinephrine (10 mM). Prostanoids were extracted from the aqueous phase of the reaction buffer three times with 2 vol of ethyl acetate, and PGE2 concentrations in the extracted samples were determined by enzyme-linked immunosorbent assay (Oxford, MI). Briefly, samples and HRP-conjugated PGE2 were added for 1 h to wells of a 96-well plate, which were coated with anti-PGE2 antibody. Substrate for HRP was added to each well for 30 min, and the reaction was stopped by addition of 1N HCl. Quantitation was performed by measuring absorbance at 450 nm; protein concentrations were determined using the Bradford method.
Statistical Analyses
The differences among percentages of COX-2-positive TAL cells were assessed with the Kruskal-Wallis test. The t test was used to detect differences in COX activities; P < 0.05 was considered significant.
Results
Characteristics of COX-2 Expression in the TAL
In all experimental groups, the cells displaying COX-2 were heavily stained, compared with unstained neighboring tubular epithelial cells (Figures 1,2,3). These tubules corresponded to TAL, as identified by the colocalization of COX-2 and Tamm-Horsfall glycoprotein using single-immunolabeling in consecutive sections and double-immunolabeling in the same tissue sections (Figures 2 and 3, a and b). Subcellular COX-2 immunostaining was concentrated in the perinuclear cytoplasmic region and was more diffuse over the cytoplasm (Figure 3). Similar results were obtained using three different antisera against the carboxyl-terminal amino acid sequences unique to murine COX-2. Moreover, no staining was observed when primary antisera were preabsorbed with COX-2 carboxyl-terminal peptides or when the antisera were omitted or replaced by normal or preimmune serum. An overall examination of the immunostained renal tissue revealed striking differences among the experimental groups in the number and distribution of cells containing COX-2. Quantitative morphologic analyses demonstrated that few TAL cells were stained for COX-2 in sham-operated rats (1.6 ± 0.3%), whereas the percentage of TAL cells containing COX-2 increased to 30.6 ± 4.7% after ADX (Figures 1 and 2, a and c). Administration of DEX (Figure 2e) to ADX rats (Figure 2c) reduced the number of COX-2-positive cells to levels similar to those observed in sham-operated or untreated rats (1.4 ± 0.23%). Dose-response experiments demonstrated that DEX reduced the number of COX-2-positive cells to a similar extent when used at doses of ≥50 μg (100, 250, or 500 μg/kg per d) (Vio CP, unpublished observations).
Recruitment of cyclooxygenase-2 (COX-2) protein along the thick ascending limb (TAL) segment. A composite of low-power micrographs of COX-2-immunostained renal sections from sham-operated (a) and adrenalectomized (ADX) (b) rats is shown. The plates, depicting the whole cortex from the kidney capsule (top) to the arcuate artery (A) (bottom), provide an overall view of the effects of ADX. In the sham-operated kidney (a), COX-2 is restricted to a few TAL cells located in the outer cortex. After ADX, however, COX-2-positive cells are recruited extensively toward the outer medullary TAL segment. *, all of the TAL segments with COX-2-positive cells in a and some of the TAL segments with COX-2-positive cells in b. Scale, 100 μm; sections were counterstained with hematoxylin. G, glomerulus.
Effects of ADX and dexamethasone (DEX) treatment on COX-2 protein expression in the TAL. Immunohistochemical localization of COX-2 (a, c, and e) and Tamm-Horsfall glycoprotein (b, d, and f) is shown. (a and b) Sham-operated rats. (c and d) ADX rats. (e and f) ADX rats treated with DEX. Each pair of photographs represents consecutive sections stained for either COX-2 or Tamm-Horsfall glycoprotein and counterstained with hematoxylin (the asterisks denote the same TAL segments). Note the small number of TAL cells containing COX-2 in sham-operated rats (a), the increase in the number of COX-2-positive TAL cells after ADX (c), and the attenuation of COX-2 expression after DEX treatment (e). All of the photographs are at the same magnification; scale, 50 μm. G, glomerulus.
Distribution of COX-2 protein along the TAL segment. (a to c) Double-labeling immunostaining of TAL segments for COX-2 (dark brown) and Tamm-Horsfall glycoprotein (blue) in samples from a sham-operated rat (a) and an ADX rat (b and c). Note that, in both groups, COX-2 is located in tubular epithelial cells containing Tamm-Horsfall glycoprotein. However, COX-2 expression is restricted to a few cells in the more cortical TAL segment in sham-operated rats, whereas expression along the TAL segment after ADX (c) is directed toward the medulla, recruiting cells that in the basal state do not express the enzyme. In sham-operated or ADX/DEX-treated rats, expression is restricted to more cortical portions of the TAL (Figures 1 and 2). (d) Higher-magnification view, showing the cellular distribution of COX-2 along the TAL segment. Note that most of the TAL cells shown contain COX-2; however, several COX-2-negative cells are also present. This feature, as well as the subcellular distribution of COX-2 in the perinuclear area and over the cytoplasm, is shown at higher magnification in the inset. The arrowhead indicates a COX-2-negative TAL cell between adjacent COX-2-positive TAL cells. The small arrow points to a COX-2-negative macula densa, and the back of the arrow points to a COX-2-positive TAL cell. Scale, 25 μm in a, b, and d; 50 μm in c; the scale of d corresponds to 10 μm in insert. Sections c and d were counterstained with hematoxylin. G, glomerulus.
ADX Induction of TAL COX-2 Protein Expression
In sham-operated or untreated rats, the expression of COX-2 protein was observed in small scattered groups of epithelial cells belonging to TAL tubules located in the outer medulla and cortex (Figure 2, a and b). The number of COX-2-positive cells increased greatly in ADX animals. Moreover, recruitment of COX-2-positive TAL cells toward the more medullary portion of the TAL was observed after ADX (Figures 1b and 3c). Thus, whereas COX-2 was observed in TAL cells close to the corresponding glomeruli of sham-operated or untreated rats, the increased numbers of COX-2-positive TAL cells in ADX rats yielded nearly continuous staining of TAL epithelial cells, intermingled with a few COX-2-negative TAL cells (Figures 2 and 3b). A more detailed examination of COX-2-positive TAL cells along axially sectioned tubules was performed to assess whether the increase in the percentage of COX-2-positive TAL cells was attributable to random expression or a defined recruitment pattern along the TAL segment. This analysis demonstrated that, whereas COX-2-positive TAL cells in shamoperated rats were confined to a few cells in the outer cortex (Figure 1a), the increased expression of COX-2 observed after ADX followed a well defined pattern of recruitment of COX-2-positive TAL cells toward the outer medulla (Figures 2 and 3, b and c).
Lack of Increase in COX-2 Protein Expression in Renal Macrophages after ADX
Because ADX was previously shown to increase COX-2 expression in murine peritoneal macrophages (24), renal macrophages were identified with a specific monoclonal antibody (ED-1), to assess the possible contribution of these cells to the COX-2 expression pattern. As shown in Figures 1,2,3, COX-2 protein expression was restricted to tubular epithelial cells. No staining was observed in the few macrophages located within the tubulointerstitium. Moreover, neither ADX nor DEX treatment altered the number of macrophages observed in each field. These data support the observation that the increase in COX-2 expression after ADX is restricted to tubular epithelial cells of the cortical and outer medullary TAL.
ADX-Induced Increases in COX-2 mRNA Accumulation and Protein Expression and Effects of DEX
Immunohistochemical analysis of COX-2 protein expression was confirmed by assessment of COX-2 mRNA accumulation and protein expression. Analysis of total RNA from renal cortex revealed small amounts of COX-2 mRNA in shamoperated rats, which increased after ADX; no concomitant increase in GAPDH mRNA accumulation was observed (Figure 4). This increase was attenuated when ADX animals were treated with DEX (Figure 4). Similarly, a substantial increase in COX-2 protein expression was observed in cortical microsomes obtained from ADX rats, and treatment with DEX attenuated this increase (ADX group, 0.34 ± 0.12 arbitrary units; ADX plus DEX group, 0.02 ± 0.01 arbitrary units; n = 4, P < 0.01) (Figure 5).
Reverse transcription (RT)-PCR analysis of COX-2 mRNA accumulation. (A) Total RNA was extracted from the renal cortex of sham-operated and ADX rats, which were treated with either DEX (+) or vehicle (-). Accumulation of mRNA was assessed by RT-PCR using primer sets specific for COX-2 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Animals designated as sham control animals underwent surgery, but their adrenal glands were not removed. (B) PhosphorImaging analysis of COX-2 mRNA accumulation, normalized to the level of GAPDH mRNA (n = 3). *, P < 0.01.
Western blot analysis of COX-2 protein expression. Renal cortical microsomes were prepared from ADX rats treated with either vehicle (-) or DEX (+), and COX-2 expression was determined by Western blot analysis. COX-2 expression of four individual rats in each group is shown.
Analysis of COX Activity
Renal cortical microsomes were prepared from kidneys removed from each of the treatment groups, for assessment of the effects of ADX on COX activity. PGE2 formation was more than twofold higher in microsomes prepared from ADX rats, compared with sham-operated or untreated control rats (Figure 6). Treatment with DEX abolished the increase in PGE2 formation induced by ADX (Figure 6). These data indicate that COX-2 protein expressed in response to ADX is enzymatically active, and they suggest that endogenous glucocorticoids may affect COX-2 expression via a mechanism that regulates protein expression.
Analysis of COX activity. Renal cortical microsomes were isolated from rats that had been subjected to various treatments, and microsomes were incubated for 10 min at 37°C with arachidonic acid (7 μM). Prostanoids were extracted and analyzed using an enzymelinked immunosorbent assay for prostaglandin E2 (PGE2). The data are mean ± SD (n = 3). *, P < 0.05. AdrX, ADX.
Discussion
We demonstrated that ADX increased COX-2 mRNA accumulation and protein expression in the kidney, an effect that was prevented by the administration of DEX. Immunohistochemical analyses revealed that COX-2 protein expression was restricted to a subset of TAL cells. Morphometric analyses revealed that the percentage of TAL cells containing COX-2 increased from 1 to 2% in sham-operated rats to approximately 30% in ADX rats. Moreover, COX-2 expression was observed in a well defined pattern, suggesting recruitment of TAL cells from the cortex to the outer medulla. The increased expression of COX-2 protein observed in TAL cells was confirmed by Western blot analysis of COX-2 protein expression and by RT-PCR analysis of COX-2 mRNA expression. The increase in COX-2 mass was associated with increased conversion of exogenous AA to PGE2 by renal cortical microsomes obtained from ADX rats, indicating that the newly expressed protein was enzymatically active. Our study identifies the TAL as the tubular segment in which COX-2 is predominantly expressed after ADX. COX-2 is also expressed in the macula densa and in papillary interstitial cells in response to salt depletion (9). Moreover, expression of COX-2 in the renal cortex and the medulla is differentially regulated in response to changes in salt intake. Low dietary salt intake increases COX-2 expression in the cortex, whereas medullary COX-2 is expressed in response to high salt intake (11,12).
The phenomenon of induction, which is described herein for COX-2 in the TAL, is similar to the well described developmental recruitment of renin-containing myoepithelial cells in afferent arterioles (25). In fetal and neonatal life, renin is expressed in large sections of the preglomerular vasculature, including afferent arterioles and interlobular arteries. Expression decreases with maturation and, in adult animals, is restricted to a small number of myoepithelial cells in the afferent arterioles close to the glomeruli (25). In adult animals, interruption of the inhibitory feedback mechanism mediated by angiotensin II, in response to inhibition of angiotensin-converting enzyme, resulted in recruitment of myoepithelial cells expressing renin (26). Therefore, blockade of this inhibitory mechanism elicited derepression of renin expression, resulting in recruitment of cells that did not express renin in the basal state. We postulate that, in adult rats, COX-2 in the TAL is under tonic inhibition by glucocorticoids. This inhibitory mechanism is removed after ADX, resulting in induction and recruitment of COX-2-expressing TAL cells. A similar mechanism may have contributed to the regulation of COX-2 expression in murine peritoneal macrophages, inasmuch as ADX produced two- to threefold elevations of COX-2 mRNA and protein levels in these cells, an effect that was suppressed by replacement with DEX (16). Similarly, DEX inhibited platelet-induced COX-2 expression in mesangial cells (27). Our results are also in agreement with the recent study by Zhang et al. (28), who reported that upregulation of renal COX-2 expression after ADX was reversed by replacement therapy with either corticosterone or deoxycorticosterone. These data suggest that factors other than the removal of glucocorticoids could contribute to the regulation of COX-2 in the kidney. In contrast to the study of Zhang et al. (28) and this study, no signs of COX-2 induction were observed in the kidneys of ADX mice (16). This discrepancy can be explained by the use of renal medullary microsomes (16,24), compared with the use of cortical microsomes in the study by Zhang et al. (28) and in this study. Moreover, a study investigating the distribution of COX-2 in dissected nephron segments demonstrated COX-2 expression in the macula densa-containing segment, cortical TAL, and inner medullary collecting ducts, in an approximate ratio of 100:10:1 (12), confirming that the principal source of COX-2 in tubular cells is the TAL segment. The abundant COX-2 expression in the inner medulla is likely attributable to extratubular cells such as interstitial cells (12).
Recently, elegant studies by Zhang et al. (14) linked COX-2 to the differentiation and maturation of nephrons. COX-2 mRNA and protein expression were observed by the 16th embryonic day, remained low until birth, peaked in the first 2 wk after birth, and subsided to the low levels characteristic of adults by approximately the sixth postnatal week. These observations were confirmed by Vio et al. (15); the small numbers of COX-2-positive cells observed during early postnatal life increased from day 5 to day 15 and then decreased to adult levels. The percentages of TAL cells displaying COX-2 were 2.6, 7.8, 20.2, and 1.8% at 5, 10, 15, and 60 d of age, respectively (15). These studies support the existence of a reciprocal relationship between glucocorticoid levels and renal COX-2 expression, inasmuch as increased expression of renal COX-2 was evident during the first 2 to 3 wk after birth, when plasma levels of glucocorticoids were very low, suggesting derepression of COX-2 (29). Interestingly, a subset of cortical TAL cells in adult rats express low levels of COX-2 mRNA and protein (9,10). COX-2 expression in these cells appears to be insensitive to DEX administration, possibly because the cells are devoid of glucocorticoid receptors, which are expressed at higher densities in medullary TAL than in cortical TAL (30). Therefore, the glucocorticoid receptor distribution gradient is inversely proportional to the expression of COX-2; COX-2 expression is relatively abundant in the cortical TAL and decreases toward the outer medullary TAL, whereas glucocorticoid receptor expression is more abundant in the medullary TAL and decreases toward the cortical TAL. This suggests that, in adult rats, COX-2 expression is under the tonic inhibitory control of glucocorticoids, an effect that is unmasked by ADX. Glucocorticoid repression of COX-2 may be a protective mechanism against fibrosis, because induction of COX-2 under abnormal conditions (such as in the remnant kidney) may mediate tissue fibrosis and cell proliferation in glomeruli and the tubulointerstitium (31).
The observation of restricted distribution of COX-2 within TAL segments is not easily explained in terms of cell heterogeneity, because the TAL is derived from only one cell type, with very well defined features. The obvious exception is the macula densa, which is situated within and is functionally part of the TAL. Although subtle structural differences between medullary and cortical TAL have been observed, they develop gradually and may be unrelated to the differential expression of COX-2 (32,33,34). Serious consideration should be given to the possibility that the uneven (or scattered) distribution of COX-2 in the TAL could correspond to the previously described smooth- and rough-surface subtypes of TAL cells (33,34). It is interesting to note that the distribution of ROMK channels in the kidney also exhibits cellular heterogeneity along the TAL, as demonstrated using immunohistochemical methods (35). The eventual ultrastructural characterization of this subset of cells may have important functional implications, indicating either a different phenotype of TAL cells or a different functional state of TAL cells in response to the induction and recruitment phenomena.
The absence of COX-2 expression in the macula densa in this study may be related to differences in our experimental conditions, compared with those used by other investigators (9). For example, COX-2 expression in the macula densa of rats with normal sodium intake may be below the level of detection of our immunohistochemical methods. However, this was not the case after ADX, when the increase in COX-2 expression was observed without a concomitant increase in macula densa COX-2 levels. Perhaps a low-sodium diet is a selective stimulus for macula densa COX-2 expression, as suggested by Zhang et al. in their study of rat nephron development; those authors stated that “intense COX-2 immunoreactivity fills the cytoplasmic domain of cells that are concentrated in the vicinity of the nascent macula densa, but COX-2 immunoreactivity is not observed in the macula densa cells themselves” (14).
The differential expression of COX isoforms in the kidney may have important functional consequences. For example, recent studies demonstrated that COX isoforms are differentially regulated, in a zone-specific manner, by dietary salt (12), suggesting that COX-2 may have different functions in the cortex (vascular resistance) and medulla (salt and water homeostasis). Regarding the functional significance of COX-2 in TAL cells, this isoform participates in the regulation of tubuloglomerular feedback (36), the modulation of afferent arteriolar responses to increases in pressure (37), and renin release (38,39). Recent work from our laboratory demonstrated that TNF-mediated decreases in rubidium uptake (an in vitro correlate of natriuresis) by primary cultures of medullary TAL cells were COX-2 dependent (8). Moreover, ADX has been associated with a decrease in sodium reabsorption by the loop of Henle (40). Therefore, the preferential expression of COX-2 in the cortex, with extension to the corticomedullary junction, after ADX may reflect a functional adaptation to conserve salt in the absence of aldosterone. These findings suggest that the concept of COX-2 expression in the kidney, and other tissues only under pathologic conditions should be expanded to include enzyme expression also in normal tissue with regulation in a manner that permits participation in physiologic compensatory mechanisms (41).
In summary, ADX induces expression of enzymatically active COX-2 in the kidney, suggesting that renal COX-2 expression is subject to tonic inhibition by endogenous glucocorticoids. The induction of COX-2 in the cortical TAL after ADX proceeds in a defined pattern from the cortex toward the outer medullary TAL, indicating recruitment of TAL cells that, in the basal state, either do not express COX-2 or express it at levels below the threshold of detection.
Acknowledgments
This work was supported by grants from the Fondo Nacional de Desarrollo Cientifico y Tecnologico (Chile) to Drs. Vio and Céspedes (Grant 1980951), from the National Institutes of Health to Drs. McGiff (Grants HL25394 and HL34300) and Ferreri (Grant HL56423), and from the American Heart Association to Dr. Ferreri (Grant 9740001N) and by a Fogarty International Research Collaboration Award to Drs. Ferreri and Vio (Award TW01115). Dr. Ferreri is an Established Investigator of the American Heart Association.
- © 2001 American Society of Nephrology
References
Jump to section
More in this TOC Section
Cited By...
- Bradykinin Stimulates Renal Na+ and K+ Excretion by Inhibiting the K+ Channel (Kir4.1) in the Distal Convoluted Tubule
- Bradykinin Regulates Cyclooxygenase-2 in Rat Renal Thick Ascending Limb Cells
- Renal Cyclooxygenase-2: Evidence for Recruitment of Thick Ascending Limb of Henle Cells in Microdissected Nephron Segments