| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Department of Physiology and Pharmacology, University of Southern Denmark, DK-5000 Odense, Denmark; 2Division of Nephrology, University of Utah School of Medicine, and Salt Lake Veterans Affairs Medical Center, Salt Lake City, Utah 84132; and 3Department of Anatomy, Charité, Humboldt University, 10115 Berlin, Germany
Submitted 7 March 2003 ; accepted in final form 11 September 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-Hydroxysteroid dehydrogenase 2 was induced shortly after birth and was widely distributed in the whole collecting duct system in the suckling period and then returned to an adult pattern. Supplementation with corticosterone (20 mg·kg-1·day-1) or GR-specific dexamethasone (1 mg·kg-1·day-1) during low endogenous corticosterone suppressed renal COX-2 mRNA and protein and led to a restricted distribution of COX-2 immunolabeling. The ability of glucocorticoids to affect COX-2 was reflected in colocalization of GR-
and COX-2 immunoreactivity and mRNAs in thick ascending limb of Henle's loop. The MR antagonist potassium canrenoate (20 mg·kg-1·day-1) enhanced COX-2 expression from days 5 to 10, but low MR-specific concentrations of DOCA (1 mg·kg-1·day-1) had no effect on COX-2. Renomedullary interstitial cells expressed GR- and COX-2. Dexamethasone suppressed COX-2 in these cells. Thus low plasma concentrations of corticosterone allowed for cortical and medullary COX-2 induction during postnatal kidney development. Increased circulating glucocorticoid in the postnatal period may damage late renal development through inhibition of COX-2.
mineralocorticoid; aldosterone; 11-hydroxysteroid dehydrogenase 2; nephrogenesis; prostaglandin
-hydroxysteroid dehydrogenase type 2 (11HSD-2) metabolizes glucocorticoids to inactive metabolites and thereby reduces glucocorticoid sensitivity in a cell-specific way (9). 11HSD-2 is expressed at high levels in placenta and in mineralocorticoid target cells, also during human fetal development (5). 11HSD-2 protects the mineralocorticoid receptor (MR) from illicit activation by glucocorticoids (9), which bind to MR with the same affinity as to GR. Developmental changes in 11HSD-2 in the kidney could significantly modify steroid sensitivity in COX-2-expressing cells. In the present study, we explored these possibilities by determination of plasma aldosterone and corticosterone concentrations and by mRNA and localization studies of steroid receptors, 11HSD-2, and COX-2 during kidney development in rats. Moreover, we examined the effect of GR and MR agonists and antagonists on COX-2 expression during kidney development. |
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Radioimmunoassay for plasma corticosterone and aldosterone concentrations. Plasma aldosterone was measured using a commercial kit (COAT-A-COUNT, Diagnostic Products, Los Angeles, CA). The detection limit was 13.0 pg/ml, and the intra-assay coefficient of variation was <4%. Plasma total corticosterone was measured with the radioimmunoassay kit from Amersham's Biotrak series using 125I-corticosterone as a tracer.
RT-PCR and cDNA cloning. PCR, cloning, and sequencing of cDNAs for rat 11HSD-2, rat MR, and GR have been previously described (25). Primers and PCR for -actin and COX-2 were as previously published (2, 15). To amlify aquaporin-1 (AQP1), the following primers were used: AQP1 sense 5'-CCA GCG AAA TCA AGA AGA AGG CT-3'; antisense 5'-CTA TTT GGG CTT CAT CTC CAC C-3', covering the translated region, 806 bp (GenBank acc. no. L07268
[GenBank]
).
Solution hybridization and ribonuclease protection assays. Frozen tissue samples (150–200 mg) were homogenized (Polytron PT300, Kinematica, Switzerland), and total RNA was isolated with the RNeasy midi kit (Qiagen, Alberteslund, Denmark). RNA was eluted in pure water, and the yield was quantified by measuring OD at 260 nM (GeneQuant II, AmershamPharmacia, Hørsholm, Denmark). Messenger RNA levels were estimated by solution hybridization followed by A/T1 ribonuclease protection assay using plasmids and protocols as described (2, 15). mRNA-cRNA hybrids were separated by denaturing PAGE. Autoradiography (Biomax, Kodak) was performed at -80°C for 6 h to 3 days. Subsequently, protected probes were excised and radioactivity was quantitated in a -counter.
Immunohistochemical and immunofluorescence analysis of kidney sections. For immunolabeling, rat pup kidneys were fixed by systemic perfusion through the left cardiac ventricle with 4% paraformaldehyde solution in PBS (pH 7.35) for 5 min. Processing of tissue for immunochemical analysis was essentially as described in detail previously (8, 30, 31).
Primary antibodies used were goat polyclonal anti-rat COX-2 directed against the COOH terminus of rat COX-2 (1:500, Santa Cruz Biotechnology, sc-1747); rabbit polyclonal anti-rat GR (1:100, Santa Cruz, SC-1002) directed against the COOH terminus of the hormone binding GR- isoform; and rabbit polyclonal anti-rat 11HSD-2 (1:1,000, Chemicon, AB 1296) polyclonal rabbit anti-rat Tamm-Horsfall glycoprotein (THP) antibody (1:100, a gift from Dr. J. Hoyer, Philadelphia, PA). For immunoperoxidase labeling, the sections were incubated for 1 h with horseradish peroxidase (HRP)-conjugated rabbit secondary antibody directed against the relevant species (DAKO). Signals were visualized by incubation for 2–30 min with 0.01% diaminobenzidine and 0.02% H2O2. Negative controls included incubation without primary antibodies and preabsorption where COX-2 and GR- antibodies were incubated with the peptides used for immunization (10 µg/ml for 1.5 h). Double immunofluorescence labeling for GR- and COX-2 was performed on paraffin-embedded rat kidney sections from P10. Sections were incubated with anti-COX-2 (1:500), anti-GR (1:100), anti-THP (1:100), and anti-11HSD-2 (1:1,000) antibodies for 16 h at 4°C in 5% dry milk in TTBS. After being washed several times, the sections were incubated with donkey anti-goat IgG coupled to Alexa 594 (1:250; Molecular Probes) for 1 h and, after several washes, with goat anti-rabbit IgG (1:100) coupled to Alexa 488 (Molecular Probes). In other series, we used secondary antibodies from Dianova for THP visualization coupled to Cy2 (1:100) and for 11-visualization coupled to Cy3 (1:250). Next, the sections were washed three times and then mounted with fluorescence mounting medium (DAKO). Sections were inspected with an Olympus BX51 microscope (Olympus Denmark) equipped with an HBO fluorescence lamp. Photos were captured with a digital camera (Olympus DP50) with DP-Soft (Olympus) software and processed with Adobe Photoshop/Corell draw.
Cell culture experiments. Rat renal medullary interstitial cells (RMIC) were kindly provided by Dr. T. Maack (19). The cells were routinely propagated in RPMI medium supplemented with 10% fetal bovine serum, 100 U/ml penicillin, and 100 µg/ml streptomycin. Confluent RMIC were treated for 16 h with vehicle or dexamathasone at various concentrations. Cells were lysed in lysis buffer containing 1% Triton X-100, 100 µM PMSF, 30 mM Tris·HCl, pH 7.5, and further homogenized by sonication for 10 s.
Western immunoblotting. Tissue samples (
100 mg) were homogenized in buffer as previously described in detail (25, 31). Protein concentration was determined spectrophotometrically (Bio-Rad protein assay reagent) using serial dilution of BSA as a standard. The samples were separated by SDS-PAGE (7–10% gel) at 150–200 V for 30–40 min. Proteins were electroblotted (Bio-Rad) onto PVDF immobilon membranes (Milipore), and the membranes were reacted with primary and secondary antibodies as described (25, 31). Proteins from whole interstitial cell lysates were separated by SDS-PAGE and transferred onto nitrocellulose membrane. The blots were blocked overnight with 5% nonfat dry milk in Tris-buffered saline, followed by incubation for 1 h with rabbit anti-murine polyclonal antiserum to COX-2 (Cayman) at a dilution of 1:500. After being washed with Tris-buffered saline, blots were incubated with a goat anti-rabbit HRP-conjugated secondary antibody and visualized with ECL kits (Amersham). The band densities were evaluated by QuantityOne software from Bio-Rad and background subtracted. Data are presented as means ± SE densitometric units per square millimeter.
Microdissection of cortical thick ascending limb segments and RT-PCR analysis. Microdissection was performed basically as described previously (2). Rat pups were decapitated, and kidneys were cut in coronal slices that were digested with collagenase (Roche, Hvidovre, Denmark) for 25 min. The cortical thick ascending limb (cTAL) segments were identified by their localization in cortex, often closely associated with glomeruli, by their straight unbranched course and their relatively thin appearance. The isolated segments were transferred in 3–5 µl medium to 500 µl fresh DMEM in a 24-well cell culture plate on ice. Segments from one rat kidney were pooled, and the total length of loops was assessed by using a calibrated micrometer scale built into the ocular. We obtained between 8- and 17-mm cTAL in each dissection session. The segments were transferred to 400 µl guanidinium thiocyanate (4 M) solution, and 12 µg yeast tRNA (Roche) were added as carrier. Samples were stored at -80°C until extraction of RNA. RNA was isolated by a modified phenol-chloroform extraction protocol (2). Part of the RNA was reverse transcribed as described (2), and PCR was performed on cDNA corresponding to 1 mm of cTAL length.
Statistics. All values are given as means ± SE. When several sets of data were compared at the same time (e.g., data from different developmental stages), a one-way ANOVA was used to indicate statistical significant differences within the set of data. If the ANOVA was significant at the 5% level, differences between groups were established using 95% confidence intervals. An unpaired Student's t-test was used to compare a control with an intervention group. A value of P
0.05 was considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
200 nmol/l and aldosterone was 414.3 ± 160.1 pg/ml, 1.4 nmol/l). Aldosterone peaked on the day of birth and then decreased on P1 and remained stable during suckling. In unstressed adult rats (P80), plasma aldosterone concentration was 82.4 ± 21.8 pg/ml (n = 5), which was not different from levels measured in the suckling period (P1-P23). Thus both plasma aldosterone and corticosterone displayed a birth peak and then rapidly decreased to stable levels around P3. Whereas plasma aldosterone remained at this level through development and maturity, corticosterone increased 10 times to adult levels in late postnatal week 3 and early postnatal week 4 as renal COX-2 decreased.
|
Developmental changes in renal GR, MR, and 11HSD-2 mRNA expression in the perinatal period. The developmental pattern of MR and GR and of 11HSD-2 mRNA expression in postnatal rat kidney was characterized with a focus on the stages before renal COX-2 induction (E19, P0, P3), during COX-2 induction (P7), and after the developmental decrease of renal COX-2 (P28, P56). MR and GR mRNAs were expressed at all stages investigated, ranging from E19 to P56. GR mRNA abundance increased significantly after birth, as previously reported (Ref. 17; Fig. 2). MR was stably expressed in the suckling period and decreased significantly after weaning. Of note, in the period of COX-2 changes (P3, P7, P28), there were no statistical significant differences in either GR or MR expression levels in the kidney. -Actin expression was determined as an RNA quality control; however, -actin expression was also developmentally regulated, with higher levels in embryonal and early postnatal kidneys than at later stages. Thus -actin levels were decreasing as GR and MR were stable or increasing, which implies that RNA quality differences were not the cause of the observed changes in mRNA levels.
|
Next, we investigated renal 11HSD-2 expression. 11HSD-2 mRNA increased between birth and P3 and then stabilized (Fig. 3A). Notably, between P3 and P28, when renal COX-2 expression changes dramatically, there were no significant changes in 11HSD-2 mRNA in the kidney. Similar to MR, 11HSD-2 mRNA decreased in weanling rat kidney between days 28 and 56. Changes in whole kidney 11HSD-2 protein abundance were less pronounced compared with mRNA (Fig. 3B), and stable levels were encountered from P7 (Fig. 3B). With respect to the potential significance for regulation of COX-2, it was of particular interest to determine the localization of 11HSD-2 during kidney development. Immunoreactivity for 11HSD-2 was associated with distal nephron segments and the whole, immature, collecting duct system in early postnatal stages (P2 and P7; Fig. 3C), as previously reported (30). Labeling reached from collecting ducts in the superficial most recently differentiated cortex to the tip of the immature papilla. Between postnatal weeks 3 and 4, there was a significant redistribution of 11HSD-2 immunoreactivity that corresponded to the lower level of mRNA expression. Thus, at P29, immunolabeling had disappeared in the inner medullary collecting ducts and diminished in intensity along the axis from cortex to outer medulla (Fig. 3C, P29, bottom). Staining for 11HSD-2 was absent in the vasculature, proximal tubules, and loops of Henle and was not seen in the absence of primary antibody (not shown).
|
Effect of GR pathway on renal COX-2 expression in postnatal period. Administration of GR-selective and 11HSD-2-resistant dexamethasone (1 mg/kg) through the period of low plasma corticosterone level and high renal COX-2 (P5-P10) essentially stopped growth, whereas corticosterone (20 mg/kg) had less marked effects on body weight (Table 1). Total kidney mass was reduced significantly only by dexamethasone, but, when normalized to body weight, the kidney-to-body weight ratios were increased by dexamethasone treatment compared with control rats (Table 1). This suggests less-pronounced catabolic effects or even stimulatory effects of glucocorticoids on kidney growth. Mifepristone had no effect on any of the parameters when given from P5 to P10 (Table 1). Dexamethasone reduced renal COX-2 mRNA to one-fourth the level in vehicle-injected rats (Fig. 4A). Corticosterone had a less-marked effect and reduced COX-2 mRNA level by one-half compared with vehicle-injected rats (Fig. 4A). -Actin expression was not changed by glucocorticoid treatment, and COX-2 values were normalized with respect to -actin. The GR antagonist mifepristone did not significantly alter kidney COX-2 expression when administered from P5 to P10, which is the period with COX-2 induction and low endogenous corticosterone (control 547 ± 51 cpm/20 µg RNA, n = 6, mifepristone 517 ± 31 cpm/20 µg RNA, n = 5, not shown). COX-2 protein level was significantly lowered in response to both dexamethasone and corticosterone administration, as evidenced by densitometric evaluation of Western immunoblots (Fig. 4B).
|
|
Microscopic inspection of kidney sections did not reveal obvious changes in kidney morphology (e.g., cysts or necrotic areas) after 5 days of steroid treatment. Immunohistochemical labeling of paraffin-embedded kidney sections with a specific COX-2 antibody showed that COX-2 immunoreactivity was associated with cTAL of Henle's loop in control kidneys at P10, as shown previously (Refs. 31, 36, 40; Fig. 5). Sections from dexamethasone- and corticosterone-treated kidneys displayed a very faint immunoreactivity for COX-2 that was restricted to a few cells in the loops of Henle (Fig. 5). When the COX-2 antibody was omitted, or preabsorbed with the peptide used for immunization, no labeling was observed (Fig. 5). Thus both the level of mRNA and protein and the distribution of COX-2 in cTAL are markedly reduced by glucocorticoid supplementation in the early postnatal period.
|
Effect of the aldosterone-MR pathway on renal COX-2 in the postnatal period. Next, we examined whether the aldosterone-MR pathway had any effects on growth and COX-2 expression during postnatal development. The MR antagonist potassium canrenoate (20 mg·kg-1·day-1 soldactone) was administered before COX-2 induction from P1 to P5 and during COX-2 induction from P5 to P10. Potassium canrenoate did not change body growth or kidney growth (Table 1) but led to a significant increase in renal COX-2 expression from P5 to P10 (Fig. 6A). Immunohistochemical labeling of kidney sections from rat pups given potassium canrenoate from P5 to P10 did not display any obvious changes in labeling distribution or intensity compared with vehicle-treated pup kidneys (Fig. 6B). When given during early postnatal stages, from P1 to P5, potassium canrenoate had no effect on renal COX-2 mRNA abundance (control 427 ± 22, n = 12; potassium canrenoate 439 ± 5, n = 8, cpm normalized for -actin). The effect of the MR agonist DOCA on renal COX-2 was examined from P5 to P10, when endogenous plasma aldosterone is stable and renal COX-2 is elevated. Treatment with low doses of DOCA (1 mg/kg) preferentially activates MR but did not alter kidney or body growth (Table 1). Renal COX-2 expression was not significantly altered by DOCA from P5 to P10, although there was a tendency toward lower COX-2 expression (control 547 ± 51, n = 6; DOCA 447 ± 51 cpm/20 µg RNA, n = 5).
|
Localization of GR in rat kidney cortex in the postnatal period. On the basis of the ability of GR activation to suppress renocortical COX-2, we hypothesized that GR was localized in the loop of Henle during postnatal development. Immunohistochemical analysis of kidney sections with a specific antibody directed against the COOH terminus of hormone binding subunit GR- showed immunoreactive labeling at all stages examined (P2, P7, P10, and P28) in accordance with the mRNA data. GR- immunoreactivity was associated preferentially with nuclei in some immunopositive structures (e.g., glomeruli, loops of Henle) and with cytoplasm in others (proximal tubules). GR immunoreactivity was absent in undifferentiated mesenchyme cells in P2 kidneys and was first observed in immature stage III-IV glomeruli, where nuclei were strongly labeled (Fig. 7; P2). In the mature deep cortex, intraglomerular cell nuclei were also immunopositive, whereas proximal tubules displayed cytoplasmic labeling. Immature loops of Henle with no lumen and the bend still in superficial parts of cortex were not labeled for GR. In more mature stage IV glomeruli, where the loops have lumen, labeling for GR was associated with nuclei along the tubules, including both the macula densa segment and outer medullary loops of Henle (Fig. 7). At postnatal day 2, the collecting ducts were negative for GR, whereas at later stages, single, dispersed cells along the cortical ducts were positive, most likely intercalated cells, since this labeling was absent in the inner medulla (Ref. 25; Figs. 7 and 8). Staining for GR was not observed in the absence of primary or secondary antibody or when the antibody had been blocked by the peptide used for immunization (Fig. 7). Thus considering the expression pattern of 11HSD-2, gluco- and mineralocorticoid signaling pathways seem to be expressed at mutually exclusive sites through postnatal kidney development.
|
|
Effect of glucocorticoids on COX-2 in the medulla. It has previously been shown that COX-2 is expressed in RMIC (11, 31, 40). In the above experiments, we analyzed whole kidneys because at P10, medullary COX-2 levels are at least 10 times lower than cortical levels (31). However, to ensure that the effect of glucocorticoids on COX-2 in cortex was not obscured by opposite changes in the medulla, we examined whether glucocorticoids affected COX-2 expression in renal medullary cells. As shown in the kidney section in Fig. 8A, GR- immunoreactivity was exclusively present in nuclei of medullary interstitial cells and not in inner medullary collecting duct cells. The interstitial cell expression of GR became more pronounced with development. We examined the effect of glucocorticoids on COX-2 expression in cultured RMIC from adult rats. RMIC were treated for 16 h with vehicle or GR-specific dexamethasone at 0.5–2 µmol/l, and COX-2 protein expression was determined by immunoblotting. RMIC expressed abundant COX-2 protein that was suppressed by dexamethasone (Fig. 8B). Immunolabeling of the inner medulla for COX-2 at P10 showed that immunoreactivity was associated with interstitial cells in particular with the perinuclear area and staining was less marked and seen in fewer cells in response to dexamethasone from P5 to P10 (Fig. 8C).
Colocalization of GR, 11HSD-2, and COX-2 in the loop of Henle. In a second set of experiments, we addressed in further detail whether GR, MR, and 11HSD-2 were expressed along with COX-2 in loops of Henle by RT-PCR analysis and double-immunofluorescence labeling. cTALs were microdissected from developmental stages with maximal COX-2 induction (P7–9, P11, and P18) (28, 31, 36, 40), and cDNA corresponding to 1-mm tubule length was used for PCR-amplification for 32 cycles. Only long segments reaching into the medulla were dissected. PCR for GR yielded amplification products at all stages examined. GR was amplified only in the presence of reverse transcriptase, confirming that the amplification products originated from mRNA (Fig. 9, top). MR and 11HSD-2 were also detected in the microdissected samples (Fig. 9, bottom). COX-2 was readily amplified as a positive control from all samples. To ensure correct identification of TAL vs. descending limbs of loops of Henle, which can be mistaken during dissection, we amplified AQP1, which is a marker for the descending limb. AQP1 was not found in the dissected cTAL segments at any stage but was readily amplified from control cDNA from whole kidney (Fig. 9, bottom). We did not observe amplification products in the absence of added cDNA in any of the PCRs.
|
Subsequently, we performed double-immunofluorescence labeling experiments to specifically address whether cortical COX-2 immunoreactivity is colocalized with GR and 11HSD-2-immunoreactive proteins in the early postnatal period. Kidney sections were labeled on P10, when COX-2 was maximally elevated and widely distributed. In the cortex, COX-2 immunofluorescence (Fig. 10, top and middle micrographs, red) was associated with loops of Henle only, whereas labeling for GR (middle, green) was associated with glomeruli, proximal convoluted tubules, and loops of Henle. All cells positive for COX-2 also expressed GR (double labeling, orange fluorescence), showing that both mRNA transcripts and proteins are present in cortical loop of Henle cells during postnatal development. In contrast to the colocalization of COX-2 and GR, 11HSD-2 immunofluorescence (Fig. 10, top, green) was localized in the collecting ducts and there was no overlap with COX-2 (red fluorescence). This observation was confirmed by the TAL marker THP (Fig. 10, bottom, green fluorescence), which was not observed together with 11HSD-2 (red fluorescence).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
mRNA and immunoreactive protein in TAL of Henle's loop (and in renomedullary interstitial cells). GR- was expressed in developing loops of Henle at all stages of postnatal renal development, which is a prerequisite for the sensitivity of COX-2 to endogenous changes in circulating GR agonist. After a birth peak, plasma aldosterone rapidly stabilized at a level found in control adult rats and did not correlate with developmental changes in COX-2. Selective action of aldosterone depends on coexpression of 11HSD-2. PCR amplification showed expression of 11HSD-2 in microdissected loops of Henle in the first and second postnatal week, whereas immunohistochemical staining with an anti-11HSD-2 antibody yielded weak or no labeling at all associated with this segment during postnatal development (Figs. 9 and 10), while collecting ducts were strongly labeled. The contrasting findings indicate that 11HSD-2 is much less abundant in the loop of Henle compared with the collecting duct and suggest that an abrupt change in the ability to inactivate glucocorticoids in the loop of Henle cannot explain the induction of COX-2 during development. The sensitivity of COX-2 for corticosterone in the loop of Henle implies that 11HSD-2 activity is insufficient for protection against GR activation, although high doses of exogenous corticosterone may have exceeded enzymatic capacity. However, the nuclear localization of GR immunoreactivity in loop of Henle cells during normal postnatal development also supports insufficient 11HSD-2 activity. Nuclear translocation of activated GR is inhibited by the presence of active 11HSD-2 (16, 27). On the other hand, blockade of MR by potassium canrenoate during P5–P10 increased cortical COX-2 expression, which is compatible with the notion that activation of MR, by either gluco- or mineralocorticoid, suppresses COX-2 expression in the loop of Henle, as observed previously in young and adult rats (39, 41). This effect could be mediated directly, because by PCR we found MR expression in the loop in the first 2 postnatal weeks. Previous reports showed very low expression of MR, or none at all, in the loop of Henle in fetal and adult kidney (5, 32). Supplementation with DOCA in low doses that are MR specific failed to change COX-2 levels, which argues against a dirct, MR-mediated effect on COX-2. The stimulatory effect of potassium canrenoate on COX-2 could also be indirectly mediated through changes in NaCl and water homeostasis. Potassium canrenoate would be expected to block MR in the collecting duct system and increase sodium excretion in the pups. Renal COX-2 expression in the postnatal period is highly sensitive to changes in NaCl balance, and an increase in NaCl excretion would increase COX-2 (41). Because of the low-sodium content in rat milk and high endogenous activation of MR, supplementation with DOCA without additional NaCl would not be expected to increase sodium reabsorption significantly. Taken together, the data support that adrenal effects on renocortical COX-2 expression during postnatal development are mediated by glucocorticosteroid and that regulation takes place at the level of circulating steroid rather than by changes in renal receptor expression or 11HSD-2 expression. Although it cannot be excluded that corticosterone in the rat pups also interacts with MR in loop of Henle target cells, data suggest that the suppression of COX-2 is mediated predominantly through interaction with GR.
COX-2 is expressed in medullary interstitial cells (11, 31, 40). Corticosterone has been shown to suppress COX-2 expression in cultured interstitial cells (38). Our data confirm that this effect is indeed GR mediated because it was mimicked by GR-specific dexamethasone in vitro and in vivo and reflected in coexpression of GR- and COX-2 immunoreactivity in these cells in vivo. Overall, our data on GR and 11HSD-2 localization indicate separate and mutually exclusive sites of glucocorticoid and mineralocorticoid action along the nephron, collecting ducts, and interstitial cells in the postnatal period. We found that GR- was localized primarily in 11HSD-2-negative structures in the developing kidney (e.g., glomeruli, proximal tubules, interstitial cells) similar to the reported distribution in adult rat kidney (25, 32) and in adult and fetal human kidney (5, 39). On the basis of the changes in 11HSD-2 during development, aldosterone selectivity is expected to be enhanced during suckling along 11HSD-2-positive segments, connecting tubule, and the whole collecting duct system. Aldosterone sensitivity is also likely to be enhanced, because MR expression was increased during suckling and decreased after weaning. The combination of normal circulating plasma aldosterone and low circulating corticosterone in the suckling period further adds to the selectivity of aldosterone over corticosterone in renal target cells and underlines the established, crucial role of intact aldosterone signaling for sodium conservation and survival in the first critical days and weeks of life (3). The high 11HSD-2 activity in kidney shortly after birth might provide an explanation for the marked tissue-specific difference in glucocorticoid induction of aldosterone targets, e.g., ENaC, during early postnatal development, where the kidneys are relatively insensitive compared with lung (24).
Administration of glucocorticoid in the fetal or early postnatal period has profound effects on renal funtion at sites compatible with the present immunolocalization of GR-. Thus, in fetal sheep, glucocorticoid increases glomerular filtration rate, renal blood flow, urinary flow rate, and tubular acidification (13), and expression of the proximal tubular Na-H exchanger-3 is augmented (10). In humans, glucocorticoid increases creatinine clearance and fractional sodium reabsorption when given to mothers before preterm delivery (1). Thus functional data, and the widespread expression of GR in the fetus (5), indicate that glucocorticoids regulate diverse organ funtions, such as lung maturation, at birth, where glucocorticoid levels peak in many species. On the other hand, during intrauterine life the fetus is protected against glucocorticoids by high expression of 11HSD-2 in both placenta and multiple fetal organs (5). The present set of data suggests that, apart from the birth peak, low levels of circulating corticosterone prevail in the first weeks of life in the rat, thus mimicking intrauterine conditions in other species. In this "window" of development, which is extrauterine in the rat, circulating glucocorticoid spontaneously decreases and allows expression of glucocorticoid-suppressed pathways such as renal COX-2. Inhibition of COX-2-mediated signaling during late nephrogenesis could be one pathway by which excessive glucocorticoid exposure leads to aberrant development of the kidney. Glucocorticoid treatment in early postnatal life or an elevation of endogenous glucocorticoid leads to fewer glomeruli, sodium retention, and hypertension in adult rats (4, 22, 29). Similarly, COX-2 deficiency or COX-2 inhibition leads to cortical damage during development (7, 21, 23, 26). The window where renal COX-2 is elevated could define an interval where glucocorticoids are deleterious for nephrogenesis. This interval is likely to vary considerably between species but should be taken into account when glucocorticoids are administered in pregnancy or early postnatal life.
|
ACKNOWLEDGMENTS |
|---|
GRANTS
This work was supported by grants from the Danish Medical Research Council (22010122, 22010159), the Novo Nordisk Foundation, The Danish Heart Foundation (01123022896), the Danish Medical Association Research Fund, and Kønig-Petersens Fund for Kidney Diseases.
|
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 |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1G-subunit of a voltage-dependent Ca2+ channel is localized in rat distal nephron and collecting duct. Am J Physiol Renal Physiol 279: F997-F1005, 2000.-hydroxysteroid dehydrogenase and corticosteroid hormone receptors in early human fetal life. J Clin Endocrinol Metab 83: 4490-4497, 1998. hydroxysteroid dehydrogenase type 2 in rat colon but not in kidney by low dietary NaCl intake. Am J Physiol Renal Physiol 285: F348-F358, 2003.-hydroxysteroid dehydrogenase type 2. J Biol Chem 276: 28484-28492, 2001.This article has been cited by other articles:
|
|
 |
|
  K. A. Brennan, S. Kaufman, S. W. Reynolds, B. T. McCook, G. Kan, I. Christiaens, M. E. Symonds, and D. M. Olson Differential effects of maternal nutrient restriction through pregnancy on kidney development and later blood pressure control in the resulting offspring Am J Physiol Regulatory Integrative Comp Physiol, July 1, 2008; 295(1): R197 - R205. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Liu, W. Ye, G. Guan, Z. Dong, Z. Jia, and T. Yang Developmental regulation of calcineurin isoforms in the rodent kidney: association with COX-2 Am J Physiol Renal Physiol, December 1, 2007; 293(6): F1898 - F1904. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Baserga, M. A. Hale, Z. M. Wang, X. Yu, C. W. Callaway, R. A. McKnight, and R. H. Lane Uteroplacental insufficiency alters nephrogenesis and downregulates cyclooxygenase-2 expression in a model of IUGR with adult-onset hypertension Am J Physiol Regulatory Integrative Comp Physiol, May 1, 2007; 292(5): R1943 - R1955. [Abstract] [Full Text] [PDF]
|
|
X174/Hae III DNA fragments.