J Am Soc Nephrol 13: 1190–1198, 2002 Sgk1 Expression in Kidney and Its Regulation by : Spatio-Temporal Heterogeneity and Quantitative Analysis

JIANGHUI HOU, HELEN J. L. SPEIRS, JONATHAN R. SECKL, and ROGER W. BROWN Molecular Medicine Centre, Western General Hospital, Edinburgh, United Kingdom.

Abstract. The serine-threonine kinase sgk1 was recently iden- ␮g/kg per d [II], 150 ␮g/kg per d [III], and 750 ␮g/kg per d tified as a gene rapidly induced by mineralocorticoids, result- [IV]) generated elevation of circulating aldosterone. Across ing in increased sodium transport in vitro. To carefully localize these treatments (I through IV), the circulating level correlated and quantify the renal sgk1 expression response to aldosterone, with the progressive induction of sgk1 expression, with highly in situ hybridization was performed on kidneys of mice having stimulated tubules first appearing in cortex (I) and continuing aldosterone excess over a range of doses and durations. In downward (II) until there was a strong stimulation throughout control and adrenalectomized animals, the glomeruli and inner outer medulla (III and IV). Interestingly, chronic but not acute medullary collecting ducts were the major sites of sgk1 ex- aldosterone excess caused a slight increase of sgk1 expression pression, which was maintained independent of aldosterone. in glomerulus (30 to 50%; P Ͻ 0.01) and a dramatic down- Sgk1 upregulation induced by aldosterone excess exhibited regulation in the initial portion of inner medulla, which could spatio-temporal heterogeneity. Both acute (3-h) and chronic result from diminished interstitial osmolarity. Relative quanti- (6-d) aldosterone excess stimulated sgk1 expression in the fication (versus control) of sgk1 upregulation in individual distal nephron, i.e., from the distal convoluted tubules through tubules revealed: (1) a 1.8-fold increase of sgk1 mRNA at 3 h to the outer medullary collecting ducts. Treatments for 6 d with (150 ␮g/kg injection) and (2) a dose-dependence of chronic low sodium diet (0.03% [I]) and aldosterone infusions (50 upregulation reaching a ceiling of eightfold elevation.

Regulation of sodium reabsorption in kidney is crucial for the pathway of aldosterone action, leading to the finding of an maintenance of whole body fluid . The Naϩ trans- early-response gene, sgk1 (serum/ kinase), the port induced by aldosterone, either in vivo or in vitro,is transcription of which is rapidly (Ͻ3 h) stimulated by aldoste- ϩ sensitive to the K -sparing diuretic amiloride, which is a rone excess (7). Intriguingly, sgk1 was shown to significantly ϩ potent blocker of the epithelial Na channel (ENaC) (1–3). elevate ENaC-mediated sodium transport via increasing the The major site of aldosterone action is in the cortical collecting surface expression of ENaC subunits in Xenopus oocytes (8), duct, including its initial portion. Receptor sites and metabolic which is in keeping with evidence suggesting sgk1 may facil- effects of mineralocorticoids have been documented through itate a similar increased apical translocation of ENaC in distal the second half of distal convoluted tubules (DCT) and in the nephron in vivo (9–10). Other stimuli, such as , glu- connecting tubules and the medullary collecting ducts (4–5). cocorticoids, and cell volume change, were documented to Two phases are distinguished in its regulation of sodium trans- regulate sgk1 (11–14). These in vitro findings identify sgk1 as port: an early phase starting after a lag period of 20 to 60 min, a key gene in hormonal regulation of sodium transport in cells during which the preexisting channels and Na/K-ATPase of distal nephron. It thus becomes of great interest to determine Ͼ pumps are activated, and a late phase ( 2.5 h), which involves the distribution of sgk1 in intact kidney and study its hormonal increasing contribution from newly synthesized channel and regulation in vivo. pump proteins (6). Previous animal studies have been focusing on the short- In the A6 Xenopus cell line, which is a good model of term (0.5- to 4-h) effect of aldosterone on sgk1 expression mammalian distal nephron, efforts have been made to elucidate (7,15–16). First, we aim to elucidate whether regulation of sgk1 expression is consistent with playing a role in transduc- tion of aldosterone-driven sodium reabsorption in distal Received August 3, 2001. December 22, 2001. nephron in both short-term (acute) and chronic hyperaldoste- Correspondence to: Dr. Roger W. Brown, Molecular Medicine Centre, West- ern General Hospital, Edinburgh EH 2XU, United Kingdom. Phone: 44-131- ronism. Second, we look for evidence of other potential regu- 651-1024/1037; Fax: 44-131-651-1085; E-mail: [email protected] lators of and roles for sgk1 in kidney. In particular, sgk1 1046-6673/1305-1190 expression may exhibit an aldosterone dose dependence in Journal of the American Society of Nephrology such a way that a plateau or a peak occurs at a dose above Copyright © 2002 by the American Society of Nephrology which there is no further induction or even a fall off in the DOI: 10.1097/01.ASN.0000013702.73570.3B degree of sgk1 induction. This would be of particular interest, J Am Soc Nephrol 13: 1190–1198, 2002 Sgk1 in Kidney 1191 as it may identify a response moderating aldosterone-sgk1– ␮g/kg per d) for 6 or 21 d. To investigate the acute effects of driven sodium retention and imply that such a mechanism aldosterone, 150 ␮g/kg injections were given to mice 3 h before contributes to escape from aldosterone-induced Na retention. sacrifice. Additionally, one group of mice was kept on 0.03% low- Finally, these detailed studies on a wide range of aldosterone- sodium diet for 6 d. Table 1 gives details of these groups. related treatments will allow a clearer view of sgk1 expression and regulation along the nephron and will help to guide future Plasma Aldosterone Assay studies. To address these issues, we performed in situ hybrid- Blood samples were removed by cardiac puncture rapidly after ization using radiolabeled riboprobes, autoradiographic films, initiation of terminal anesthesia and centrifuged at 4°C for 10 min. Ϫ and silver grain counting, which are well-established means of Plasma was separated and stored at 20°C. Aldosterone was assayed using a 125I-RIA kit (DPC, Los Angeles, CA). relative quantitation of expression level (17). This approach allows comparison of sgk1 mRNA expression levels in renal regions and nephron segments. In Situ Hybridization On the basis of mouse sgk1 sequence (accession number: AF139638), primers were designed to amplify a fragment encompass- Materials and Methods ing a region (1091 to 1336 bp) that had no significant homology to Materials other sequences. Flanking T3 and T7 phage polymerase consensus Steroids, propylene glycol, DMSO, and molecular biology–grade sites were introduced by nested PCR as described previously (18). chemicals were purchased from Sigma Chemical Company (Poole, Single-stranded [35S]UTP-labeled RNA probes were generated using Dorset, UK), 1-kb ladder DNA size markers and synthesized oligo- the required RNA phage polymerases. Six mice were examined in nucleotides were purchased from Life Technologies/Life Technolo- each group. For each mouse, one sagittal cryostat section (10 ␮m) was gies (Paisley, Renfrewshire, UK), ALZET 1007D osmotic mini- cut from intact kidney, which included cortex and outer and inner pumps and mice were purchased from Charles River UK (Margate, medulla. Sections were thaw-mounted onto 3-amino propyltriethox- 35 Kent, UK), and S[UTP] was purchased from Amersham Interna- ysilane–coated slides and stored at Ϫ80°C. 4% paraformaldehyde tional (Little Chalfont, Buck, UK). was used for fixation, followed by acetylation and prehybridization at 50°C for 2 h. Hybridization with 4 ϫ 106 c.p.m [35S]UTP-labeled Animal Treatments RNA probe per slide was performed at 50°C for 14 to 16 h and then The animal work was approved by the local ethical committee for followed by RNaseA treatment and washes with a maximum strin- the care of animals and was in accordance with the NIH Guidelines gency of 0.1xSSC at 60°C. This follows well-established methodol- for the Care and Use of Laboratory Animals. Eight-week-old male ogy that reflects findings that show that these conditions create the C57BL/6J mice (23.5 to 25 g) were segregated into ten groups (each high subsaturation/saturation riboprobe concentrations that provide n ϭ 6) and caged in pairs within their own group (Table 1). All mice specific radiolabeled probe hybridization that reflects the amount of had free access to normal diet (0.3% sodium) and water (with saline target RNA present. When quantitatively detected (film and dipped additionally available to adrenalectomized mice). The bilateral adre- emulsion autoradiography) this then allows relative quantitation for nalectomy operations were performed under gaseous anesthesia. An- sgk1 expression between slides processed together. After washes, imals were killed by terminal anesthesia. The ALZET 1007D osmotic slides were dehydrated and placed against ␤-Max Hyperfilms. Three mini-pumps (preloaded with treatments or vehicle) were implanted to five days of exposure gave a satisfactory range of autoradiographs subcutaneously to mice, generating groups having aldosterone infu- in the linear range of the films as judged by coincident exposure of the sions with various doses (50 ␮g/kg per d, 150 ␮g/kg per d, and 750 film to radioactive microscale standards for this purpose (calibrated

Table 1. Experimental groups

Group Name Group Description Infusion Methods Drug Delivery Rate Lasting Time

Adx Adrenalectomy Vehiclea mini-pump 6d Con Control Vehicle mini-pump 6 d Aldo50 Aldosterone Aldosterone/vehicle mini-pump 50 ␮g/kg per d 6 d Aldo150 Aldosterone Aldosterone/vehicle mini-pump 150 ␮g/kg per d 6 d Aldo750 Aldosterone Aldosterone/vehicle mini-pump 750 ␮g/kg per d 6 d Aldo21d Aldosterone Aldosterone/vehicle mini-pump 150 ␮g/kg per d 21 d ConInj Control Saline injection 3 h AldoInj Aldosterone Aldosterone injection 150 ␮g/kg 3 h Spiron Aldosterone and Aldosterone injection and 150 ␮g/kg 3 h of aldosterone pretreatment with pretreatment with spironolactone/ aldosterone and 20 and6dof spironolactone vehicle mini-pump mg/kg per d spironolactone spironolactone LowSalt 0.03% sodium Vehicle mini-pump — 6d diet

a Vehicle: 0.9% saline, 2% (vol/vol) ethanol. 1192 Journal of the American Society of Nephrology J Am Soc Nephrol 13: 1190–1198, 2002 for moderate ␤-emitters [14C, 35S] in thin tissue slices: #RPA504, tical analyses. For each mouse, the mean density of each such renal RPA511; Amersham International, Little Chalfont, UK). All slides component assessed (glomeruli and cortical and medullary tubules or were then dipped in photographic emulsion and exposed in a light- areas) was then calculated. tight box for 2 to 3 wk before being developed. This optimal time was determined by the optimal exposure for the film autoradiographs. In preliminary work, we emulsion-dipped slides with attached mi- Statistical Analyses croscales (such as above) and found a wide linear range (correlation, The significance of differences between groups was tested by r ϭ 0.98), relating grain-count density versus radioactivity. For all the ANOVA. When the all-effects F value was significant (P Ͻ 0.05), dipped slides in these sgk1 studies, grain-count densities did not post hoc analysis of differences between individual groups was made exceed this linear range. Selected sections were then counterstained with the Neuman-Keuls test. Values were expressed as mean Ϯ SE with the periodic acid-Schiff (PAS) method. unless otherwise stated.

Quantitative Image Analyses Results Renal Expression Profile of sgk1 Autoradiographic films were scanned on a high-resolution flatbed scanner. Developed emulsion-dipped slides were analyzed on a Zeiss Examination of control and adrenalectomized animals re- Axioskop Microscope (Carl Zeiss, Thornwood, NY) with the KS300 vealed that sgk1 mRNA had a discrete distribution in kidney, (v3) silver-grain counting software. Steps to minimize background with a punctate appearance in cortex, little expression in outer tissue hybridization included the employment of a postfixing acety- medulla, and high abundance in inner medulla (Figure 1A). lation procedure to block the positively charged ammonium groups on This expression profile was not dependent on circulating cor- tissue and a posthybridization RNase A treatment to cleave unhybrid- ticosteroids, as it was maintained in adrenalectomized animals ized single-stranded RNA molecules. Thus, the tissue background for (aldosterone Ͻ400 ϮϽ100 pmol/L; corticosterone Ͻ55 nmol/ antisense probe was considered as the nonspecific binding to tissue L). The cortical expression of sgk1 was primarily confined to RNA by probes with equal GC percentage, and so equal to the glomeruli (Figure 1B). Its expression in glomeruli was uniform hybridization level on sense control sections. When silver grains were and centrally distributed (Figure 1E), ruling out the possibility counted in high magnification fields, the tissue background was that the glomerular visceral or parietal epithelium was the normalized for each slide as follows. The background grain density of tissue and glass on the sense control slide were measured separately. expression site. This is consistent with the data from human The ratio between such paired background measurements was calcu- kidney (19), suggesting sgk1 expression in the glomerular lated for the slide. The tissue background for an antisense slide was mesangium. In inner medulla, nearly all cells strongly ex- thus estimated by multiplying its glass background by this ratio. pressed sgk1 as the papilla was approached (Figure 1, C and Quantification was done by counting the number of silver grains over D). This characteristic indicates that the inner medullary col- areas under high-power magnification (ϫ400). Grain-counting in- lecting ducts (IMCD) are the expression location, because the volved determining the grain density over glomeruli and tubules that number of thin limbs of Henle’s loop diminishes toward highly expressed sgk1. Identification of glomeruli was straightforward papilla. for all groups. Identification of tubules in both cortex and outer medulla involved comparing highly expressing tubules in animals with aldosterone excess and those in controls, where tubular expres- Chronic Effects of Aldosterone and Dosage Response sion merged with “background” levels. Although there was some A range of 6-d treatments with sodium depletion (I: 0.03% variation across non-glomerular cortex (mean Ϯ SE grain density, sodium diet) and aldosterone infusions (II through IV: 50 2 0.0869 Ϯ 0.0106 counts/␮m ), this did not clearly correspond to ␮g/kg per d, 150 ␮g/kg per d, and 750 ␮g/kg per d) resulted in tubule outlines and was present over a wider cortical area in control sustained elevation of circulating aldosterone levels: I, 1234 Ϯ and treatment groups than the highly expressing tubules readily iden- 124 pmol/L; II, 4660 Ϯ 120 pmol/L; III, 15485 Ϯ 471 pmol/L; tifiable in treatment groups. The best valid solution to estimate control and IV, 53569 Ϯ 3198 pmol/L. In view of the fact that severe tubular grain counts was to measure this grain density over a larger tubular area, (full-field [ϫ400] view of medulla or non-glomerular sodium deficiency has been reported to increase aldosterone cortex) rather than individual tubules. This control expression level secretion 20- to 25-fold in sheep or human (20–21), the two was used for comparison with animals with aldosterone excess. We high-dose infusions (III and IV) mimicked these physiologi- followed a protocol for random selection of regions, glomeruli, and cally extreme conditions of aldosterone excess (25-fold and tubules, which involved aligning the field of view over renal cortex at 75-fold elevation of plasma level versus 699 Ϯ 56 pmol/L of such low power that the viewer was unable to see local renal structure controls). Correlated with the elevation of circulating aldoste- or grain density. The view was then zoomed to ϫ100 to ϫ200, and rone across these animal groups, there was a change in the glomeruli or tubules in the central field of view were those designated pattern of sgk1 expression, with highly stimulated tubules first ϫ to be subjected to grain counting at 400. The process was repeated, appearing in cortex (I, Figure 2, A and a) and continuing counting ten glomeruli or tubules (with grain-density elevated allow- downward (II, Figure 2, B, b, and c) until there was a strong ing them to be demarcated against background) out of eight sampled stimulation throughout outer medulla (III and IV, Figure 2, C fields in each region (for cortex and medullary regions separately). Sampling cortical or medullary regions was carried out in the same and d). Silver-grain counting of individual tubules showed a fashion for each kidney section. When it came to control sections, five dose-dependent upregulation of sgk1 in both cortex and outer full-field (ϫ400) views were counted (minus glomeruli in cortex, five medulla (Figure 3). The mild excess of aldosterone in sodium- rather than eight because of their larger area). Rarely, part of the depleted animals (I) induced a 4.7-fold increase of sgk1 ex- section with poor morphology was excluded from subsequent statis- pression, which was confined to cortical tubules. Such cortical J Am Soc Nephrol 13: 1190–1198, 2002 Sgk1 Gene Expression in Kidney 1193

Figure 1. Sgk1 expression profile in kidney. Autoradiograph (A: expression ϭ black) and photomicrographs (B through D: dark-field, expression ϭ white; E: bright-field, expression ϭ visible silver- grains) showing sgk1 expression in detail in the kidney of a normal (control) mouse. The boxes in the autoradiograph in panel A show the sites of capture of the photomicrographs on the corresponding kidney section slide (B: ϫ100, cortex; C and D: ϫ50, medulla). A single glomerulus was visualized in panel E at ϫ400 magnification. Note the abundant expression in glomeruli and inner medulla. Glom, glomer- ulus; Cx, cortex; oM, outer medulla; iM, inner medulla.

induction reached the ceiling (eightfold) in animals infused with aldosterone (II and IV). In outer medulla, sgk1 stimu- lation (6.4-fold) was first observed in animals with moderate aldosterone excess (II). The ceiling of eightfold increase was reached at groups with severe aldosterone excess (III and IV). Moreover, in animals treated with 150 ␮g/kg Figure 2. Chronic effects of aldosterone on sgk1 expression. Autora- ϭ aldosterone for 21 d, sgk1 stimulation persisted and its diographs (A through C: expression black) and photomicrographs (a through e: dark-field, expression ϭ white) showing sgk1 expression pattern and extent remained similar to that at 6 d (Figure 3). in kidneys of animals treated with aldosterone excess (Magnifications: Finally, water intake of animals was substantially increased ϫ100 in a and b; ϫ50 in c through e). The sites of capture of the by the high-dose aldosterone infusions (twofold in the 150 photomicrographs on the kidney sections are indicated on the corre- ␮g/kg per d groups and threefold in the 750 ␮g/kg per d sponding autoradiographs. Note the highly stimulated tubules travers- group versus control). ing from cortex to outer medulla. T, stimulated tubules. 1194 Journal of the American Society of Nephrology J Am Soc Nephrol 13: 1190–1198, 2002

Figure 3. Quantification of sgk1 expression in response to chronic aldosterone excess. Mice were treated for 6 d with sodium depletion, low-dose (50 ␮g/kg) and high-dose (150 to 750 ␮g/kg) aldosterone infusions or for 21 d with 150 ␮g/kg infusion. Sodium depletion induced a 4.7-fold increase of sgk1 mRNA level in cortical tubules (P Ͻ 0.01 versus control). The 50 ␮g/kg infusion stimulated sgk1 by 8.2-fold in cortex and 6.4-fold in outer medulla. Sgk1 stimulation by high-dose infusions for 6 or 21 d reached a ceiling of upregulation (eightfold elevation) in cortex and outer medulla. The mean grain density in cortical tubules of control animals was designated as one unit.

Acute Effects of Aldosterone In mice injected with 150 ␮g/kg aldosterone 3 h before sacrifice, highly stimulated tubules were visualized in cortex and outer medulla (Figure 4). The rapid sgk1 induction caused by injection was completely abolished in animals pretreated with spironolactone (MR antagonist) (Figure 4), suggesting mediation via mineralocorticoid receptors (MR). Silver-grain counting to examine the acute effects of aldosterone revealed a 1.8-fold upregulation of sgk1 in cortical tubules and 1.4-fold in outer medullary tubules, lower than that elicited by the chronic treatments, e.g., 6 d and 21 d (Figure 5). Nevertheless we cannot rule out the possibility that a peak of sgk1 stimulation may have occurred before 3 h, considering the short half-life of circulating aldosterone (20 min) (22) and sgk1 mRNA (30 min) Figure 4. Acute effects of aldosterone on sgk1 expression. Autoradiographs (23). Furthermore, other studies showed that sgk1 stimulation (expression ϭ black) showing comparison of sgk1 expression among control peaked in rat kidney within 1 to 2 h after aldosterone injection (A), animals injected with 150 ␮g/kg aldosterone 3 h before sacrifice (B), and (15–16). those pretreated with spironolactone (MR antagonist) (C).

Aldosterone Induction of sgk1 Expression Is in Distal Nephron a minority of cortical tubules. The periodic acid-Schiff coun- Examination of emulsion-dipped slides revealed both acute terstaining method stains the brush border of proximal tubules and chronic aldosterone treatments induced sgk1 expression in with pink color. Sgk1 stimulation was visualized in tubules J Am Soc Nephrol 13: 1190–1198, 2002 Sgk1 Gene Expression in Kidney 1195

Discussion The expression of sgk1 mRNA and its regulation by aldo- sterone were studied in detail. In kidney of control mice, we found an abundant expression of sgk1 in glomeruli and IMCD, which was substantially maintained independent of corticoste- roids and thus persisted in adrenalectomized animals. In mice, both acute and chronic aldosterone infusions induced sgk1 expression in distal tubules through to OMCD. These regions have the potential to mediate amiloride-sensitive sodium trans- port, as they are the major sites of ENaC action. Although sgk1 was shown to increase the surface expression of ENaC subunits in Xenopus oocytes (8), a recent report suggested that ENaC translocation to the apical membrane was limited to the prox- imal portion of the distal nephron, i.e., from DCT to connecting tubules, in response to rapid aldosterone infusion (2 to 4 h) in rats (9). Both our findings (mRNA) and reports of other groups Figure 5. Quantification of sgk1 expression in response to acute (protein) showed that acute sgk1 accumulation was localized to aldosterone excess. Aldosterone-induced upregulation of sgk1 mRNA the entire aldosterone-sensitive distal nephron (ASDN: i.e., by 1.8-fold in cortex and 1.4-fold in outer medulla (P Ͻ 0.01 versus from DCT to OMCD), an axially heterogeneous mechanism control). The mean grain density in cortical tubules of control animals can be proposed to control the activation of sgk1 (e.g.,by was designated as one unit. phosphorylation) (9), and thereby restrict ENaC translocation to the more upstream portion of the distal nephron. Beyond previous findings, our work shows an axial heterogeneity of lacking this border (Figure 6, A, B, D, and E), indicative of prolonged upregulation of sgk1 mRNA, which correlates with more distal tubules, which in cortex were distributed widely the plasma level of aldosterone across groups on low-sodium across cortical labyrinths as well as in cortical medullary rays; diet to low-dose and high-dose aldosterone infusions. It ap- in fact, from DCT to cortical collecting ducts (CCD). More- peared that mild aldosterone excess (0.03% sodium diet) in- over, tubules at the vascular pole of glomerulus exhibited sgk1 duced sgk1 in the cortical portion of ASDN, whereas moderate induction (Figure 6, C and F), suggesting possible involvement (50 ␮g/kg infusion) and more severe (150 ␮g/kg and 750 of the early portion of DCT. However to characterize the DCT ␮g/kg infusions) excess affected the entire ASDN. Two recent distribution definitively would require additional studies. Sgk1 studies in rat (10) and mouse (24), have reported that treat- stimulation could be seen in long single tubules extending in ments with chronic aldosterone excess (10) or dietary sodium continuity from cortex through outer medulla to enter inner restriction (24) are associated with ENaC translocation in distal medulla without a major change in diameter (Figure 2, c nephron. Dietary sodium restriction in mice caused ENaC to through e). This pattern was fully consistent with the cortical shift from cytoplasm to the apical membrane in cells confined and outer medullary collecting ducts (OMCD) being sites of to the cortical segments of ASDN (24). This could be inter- upregulation of sgk1 expression. preted as potentially due to the accumulation of sgk1, which linked aldosterone effects to ENaC insertion. This study is the first to assess sgk1 regulation using in situ Distinct Regulation by Aldosterone in Glomeruli hybridization in mouse as well as the first to carefully quantify and IMCD the degree of its induction in different parts of the nephron It was somewhat unanticipated that sgk1 mRNA abundance (distal tubules, glomeruli, and inner medulla) using silver-grain was raised (30 to 50%) by chronic (50 ␮g/kg per d) but not counting. Previous work has examined aldosterone regulation acute aldosterone excess in glomeruli (Figure 7), a site not of sgk1 mRNA levels in renal cell lines (7,25), the whole regarded as a classical MR-responsive target. A similar up- kidney (15–16), or dissected renal components (9,26) by north- regulation was also detected in animals with higher doses of ern blotting or RT-PCR and has demonstrated the accumula- aldosterone infusion but not in those with sodium depletion. tion of sgk1 protein in response to hormone excess by immu- Interestingly, the abundant sgk1 expression in inner medulla nohistochemistry (7,9). As the current work examines the was downregulated in its initial portion by chronic and severe effects of a wider range of treatments (Table 1), this allows a aldosterone excess (150 or 750 ␮g/kg per d for 6 or 21 d). The clearer view of the way aldosterone affects sgk1 expression degree of this effect showed some individual variation within across the nephron. In the present study, up to an eightfold treatment groups and was more striking at 21 d. As illustrated increase of sgk1 mRNA was found in individual renal tubules in Figure 2, C and e, the outer-inner medulla junction was easy in response to chronic aldosterone excess and 1.8-fold to acute to identify on the autoradiograph, where the OMCD with high excess. The upregulation of sgk1 mRNA exhibited dose de- sgk1 expression were adjoining the inner medulla with much pendence over a range of circulating aldosterone (2- to 75-fold repressed expression. elevation). The reaching of its ceiling (eightfold elevation) in 1196 Journal of the American Society of Nephrology J Am Soc Nephrol 13: 1190–1198, 2002

Figure 6. Histology of renal cortical sections from aldosterone-treated groups. The images (magnification, ϫ100) of panels A through C were taken by phase-contrast microscopy, and D through F are their respective counterparts under dark-field view. The cortical labyrinth (CL) is illustrated on panel A. Panel B shows a region where a cortical medullary ray (CMR) protrudes into cortical labyrinth. Periodic acid-Schiff (PAS) stains the brush border of proximal tubules (PT). In dark-field view (D through F), the counterstaining caused the image to have a dark green (negative) or dark purple (positive) background, on top of which silver grains are visible. In cortex (both CL and CMR), the tubules in which upregulation of sgk1 by aldosterone is seen are PAS-stained negative, indicating they are tubules of the distal nephron. Glomerulus and surrounding structures are illustrated in panels C and F. The distal tubule approximating to the vascular pole of glomerulus exhibited sgk1 induction, suggesting the possible involvement of the early portion of distal convoluted tubules. Glom, glomerulus; VP, vascular pole; UP, urine pole; ϩ, high expression. animals with severe aldosterone excess suggested the satura- unclear. It is possible that aldosterone at high doses will bind tion of available MR sites and of sgk1-mediated sodium reten- to to some degree and subsequently tion via ENaC. induce sgk1 expression (27–28). However, circulating physio- The mediator that upregulated sgk1 in glomeruli remains logic glucocorticoid levels (18) supply a better glucocorticoid J Am Soc Nephrol 13: 1190–1198, 2002 Sgk1 Gene Expression in Kidney 1197

fluid intake with chronic and severe aldosterone excess, it may account for the corresponding variation in the degree to which inner medullary sgk1 expression is downregulated under these circumstances. The action of sgk1 in inner medulla is unclear, but it could be involved in controlling cell volume in the face of high interstitial osmolarity toward the papilla. Additionally, it could be involved in regulating sodium reabsorption in IMCD. In the latter case, a reduction in sgk1 expression with prolonged aldosterone infusion might contribute to facilitating natriuresis in “escape” from the sodium retention induced by aldosterone.

Acknowledgments We thank the University of Edinburgh Medical School and the UK Overseas Research Scheme (jointly supported studentship to JH); and the Scottish Hospital Endowment Research Trust (grant Rg77/00), Figure 7. Quantification of sgk1 expression in glomerulus. Silver- British Heart Foundation (grant FG/2001075), and Urquhart Charita- grain counting revealed a slight increase (30 to 50%) of sgk1 expres- ble Trust for their support. We thank June Noble, Lynne E. Ramage, sion induced by chronic (6 d) but not acute (3 h) aldosterone excess and Lawrence Brett for their expert technical assistance. We thank Dr. (* P Ͻ 0.01 versus control). Chris Kenyon for his help and valuable advice on the aldosterone and corticosterone assays. receptor ligand at even higher levels than the raised plasma References aldosterone in our aldosterone excess groups, especially the 50 1. Horisberger JD: Amiloride-sensitive sodium channels. Curr ␮g/kg per d group, which exhibits this glomerular sgk1 up- Opin Cell Biol 10: 443–449, 1998 regulation. Moreover adrenalectomy does not abolish glomer- 2. Rossier BC: The epithelial channel and the control of blood ular sgk1 expression, indicating that other pathways are in- pressure. 1996 J Am Soc Nephrol 8: 980–992, 1997 volved. As elevated mesangial sgk1 in diabetic nephrons has 3. Garty H: Regulation of sodium permeability by aldosterone. been thought to follow altered GFR (19), it may be that Semin Nephrol 12: 24, 1992 4. Doucet A, Katz AI: Mineralocorticoid receptors along nephron. glomerular hemodynamic changes during chronic aldosterone Tritiated-aldosterone binding in rabbit tubules. Am J Physiol 241: excess initiate the mesangial sgk1 upregulation. F605, 1981 The axial osmolality gradient in the renal medulla is made 5. Farman N: Steroid receptors: Distribution along the nephron. up of gradients of several individual solutes, including NaCl Semin Nephrol 12: 12–17, 1992 and urea. In antidiuresis, progressive increases in tissue NaCl 6. Verrey F, Pearce D, Pfeiffer R, Spindler B, Mastroberardino L, and urea concentration and osmolality are observed along the Summa V, Zecevic M: Pleiotropic action of aldosterone in epi- corticomedullary axis from the cortex to the papilla. In con- thelia mediated by transcription and post-transcription mecha- trast, the gradients are markedly attenuated during water di- nisms. Kidney Int 57: 1277–1282, 2000 uresis (29). We propose an osmotic theory to explain the high 7. Chen SY, Bhargava A, Mastroberardino L, Meijer OC, Wang J, abundance of sgk1 expression in IMCD, although we cannot Buse P, Firestone GL, Verrey F, Pearce D: Epithelial sodium rule out the possibility of sgk1 expression in interstitial cells. channel regulated by aldosterone-induced protein . Proc Natl Acad Sci USA 96: 2514–2519, 1999 Evidence from other research supports this proposal and it 8. Rosa de la DV, Zhang P, Naray-Fejes-Toth A, Fejes-Toth G, seems that the extracellular hypertonicity or cell shrinkage can Canessa CM: The serum and glucocorticoid kinase sgk increases stimulate sgk1 expression in vitro (14). Thus, a certain osmotic the abundance of epithelial sodium channels in the plasma mem- threshold sufficient to induce sgk1 may be reached at the brane of Xenopus oocytes. J Biol Chem. 274: 37834–37839, outer-inner medullary junction. The unexpected downregula- 1999 tion of sgk1 in inner medulla induced by severe aldosterone 9. Loffing J, Zecevic M, Ferailie E, Kaissling B, Asher C, Rossier excess indicates the engagement of regulatory mechanisms BC, Firestone GL, Pearce D, Verrery F: Aldosterone induces other than MR mediation. It is possible that the osmotic gra- rapid apical translocation of ENaC in the early portion of the dient was partially washed out by the increased water intake renal collecting system: Possible role of SGK. Am J Physiol. 280: (twofold in the 150 ␮g/kg per d groups and threefold in the 750 F675–682, 2001 ␮g/kg per d group versus control), which was elicited by 10. Masilamani S, Kim GH, Mitchell C, Wade JB, Knepper MA: Aldosterone-mediated regulation of ENaC ␣, ␤ and ␥ subunit long-term and severe aldosterone excess. Hence, in such a proteins in rat kidney. J Clin Invest 104: R19, 1999 partial “wash-out,” the osmotic gradient including the thresh- 11. Perrotti N, He RA, Phillips SA, Haft CR, Taylor SI: Activation old at which sgk1 was stimulated would move further distally of serum and glucocorticoid-induced protein kinase (sgk) by toward papilla; this may explain the reduction of sgk1 expres- cyclic AMP and insulin. J Biol Chem. 276: 9406–9412, 2001 sion in the initial part of inner medulla. Although further 12. Wang J, Barbry P, Maiyar AC, Rozansky DJ, Bhargava A, Leong investigation is required, as there is an individual variation in M, Firestone GL, Pearce D: SGK integrates insulin and miner- 1198 Journal of the American Society of Nephrology J Am Soc Nephrol 13: 1190–1198, 2002

alocorticoid regulation of epithelial sodium transport. Am J 21. Mills JN: Aldosterone secretion in man. Br Med Bull 18: 170, Physiol Renal Physiol 280: F303, 2001 1962 13. Webster MK, Goya L, Ge Y, Maiyar AC, Firestone GL: Char- 22. Tait JF, Tait SAS, Little B, Laumas KR: The disappearance of acterization of sgk, a novel member of the serine/threonine 7-H3-D-aldosterone in the plasma of normal subjects. J Clin protein kinase gene family which is transcriptionally induced by Invest 40: 72, 1961 and serum. Mol Cell Biol. 13: 2031–2041, 1993 23. Webster MK, Goya L, Firestone GL: Immediate-early transcrip- 14. Waldegger S, Barth P, Raber G, Lang F: Cloning and character- tional regulation and rapid mRNA turnover of a putative serine/ ization of a putative human serine/threonine protein kinase tran- threonine protein kinase. J. Biol. Chem. 293: 11482–11485, 1993 scriptionally modified during anisotonic and isotonic alterations 24. Loffing J, Pietri L, Aregger F, Bloch-Faure M, Ziegler U, Mene- of cell volume. Proc Natl Acad Sci USA 94: 4440–4445, 1997 ton P, Rossier BC, Kaissling B: Differential subcellular localiza- 15. Brennan FE, Fuller PJ: Rapid upregulation of serum and glu- tion of ENaC subunits in mouse kidney in response to high and cocorticoid-regulated kinase gene expression by corticosteroids low sodium diets. Am J Physiol 279: F252–F258, 2000 in vivo. Mol Cell Endocrinol 166: 129, 2000 25. Naray-Fejes-Toth A, Canessa C, Cleaveland ES, Aldrich G, 16. Bhargava A, Fullerton MJ, Myles K, Purdy TM, Funder JW, Fejes-Toth G: Sgk is an aldosterone-induced kinase in the renal Pearce D, Cole TJ: The serum and glucocorticoid induced kinase collecting duct. J Biol Chem 274: 16973–16978, 1999 is a physiological mediator of aldosterone action. Endocrinology 26. Shigaev A, Asher C, Latter H, Garty H, Reuveny E: Regulation 142: 1587, 2001 of sgk by aldosterone and its effects on the epithelial sodium 17. Wilkinson DG: In Situ Hybridization.2nd edition. Oxford, Ox- channel. Am J Physiol 278: F613–F619, 2000 ford University Press, 1998 27. Yan K, Kudo A, Hirano H, Watanabe T, Tasaka T, Kataoka S, 18. MacDonald P, MacKenzie, Ramage LE, Seckl JR, Brown RW: Nakajima N, Nishibori Y, Shibata T, Kohsaka T, Higashihara E, Corticosteroid regulation of amiloride-sensitive sodium channel Tahaka H, Watanabe H, Nagasawa T, Awa S: Subcellular local- subunit mRNA expression in mouse kidney. J Endocrinol. 165: ization of glucocorticoid receptor protein in the human kidney 25–37, 2000 glomerulus. Kidney Int 56: 65, 1999 19. Lang F, Klingel K, Wagner CA, Stegen C, Broer S: Deranged 28. Escoubet B, Coureau C, Blot-chabaud M, Bonvalet JP, Farman transcriptional regulation of cell volume sensitive kinase hSGK N: Corticosteroid receptor mRNA expression is unaffected by in diabetic nephropathy. Proc Natl Acad Sci USA 97: 8157– corticosteroids in rat kidney, heart and colon. Am J Physiol 270: 8162, 2000 C1343–C1353, 1996 20. Laragh JH, Ulick S, Januszewicz V, Deming QB, Kelly WG, 29. Hai MA, Thomas S: The time-course of changes in renal com- Lieberman S: Aldosterone secretion and primary and malignant position during lysine vasopressin infusion in the rat. Pflugers . J Clin Invest 39: 1091, 1960 Arch 310: 297, 1969