The Serum- and Glucocorticoid-Induced Kinase Is aPhysiological Mediator of Aldosterone Action*

ADITI BHARGAVA, MERYL J. FULLERTON, KATHY MYLES, TIMOTHY M. PURDY,JOHN W. FUNDER, DAVID PEARCE, AND TIMOTHY J. COLE

Baker Medical Research Institute (M.J.F., K.M., J.W.F., T.J.C.), Melbourne 8008, Australia; andDepartment of Medicine, University of California (A.B., T.M.P., D.P.), San Francisco, California 94143

ABSTRACTAldosterone plays a major role in regulating sodium and potassium

flux in epithelial tissues such as kidney and colon. Recent evidencesuggests that serum- and glucocorticoid-regulated kinase (SGK) isinduced by aldosterone and acts as a key mediator of aldosteroneaction in epithelial tissues. Induction of SGK messenger RNA(mRNA) has previously been shown within 30 min of addition ofsupraphysiological doses of aldosterone to Xenopus A6 cells andwithin 4 h in rat kidney in vivo. In this study we determined the timecourse of SGK induction, at doses of aldosterone in the physiologicalrange, in rat kidney and colon, using Northern and Western blotanalyses and in situ hybridization and determined concurrentchanges in urinary sodium and potassium excretion by Kagawa bio-assay. On Northern blot analysis, SGK mRNA levels were signifi-

cantly elevated in both kidney and colon 60 min after the injection ofaldosterone. SGK protein in late distal colon was significantly ele-vated 2 and 4 h after aldosterone treatment. In situ hybridizationshowed SGK mRNA to be induced in renal collecting ducts and distaltubular elements in both cortex and medulla by doses of aldosteroneof 0.1 mg/100 g BW or more within 30 min of steroid treatment.Significant changes in urinary composition were similarly seen withan aldosterone dose of 0.1 mg/100 g BW from 90 min after aldosteroneinjection. The early onset of SGK induction in kidney and colon andthe correlation with urinary changes in terms of both time course anddose response suggest that SGK plays an important role in mediatingthe effects of aldosterone on sodium homeostasis in vivo. (Endocri-nology 142: 1587–1594, 2001)

THE CLASSICAL EFFECTS of aldosterone in epithelia aremediated by the intracellular mineralocorticoid recep-

tor (MR), a member of the nuclear receptor superfamily thatacts as a ligand-dependent nuclear transcription factor toregulate gene expression. In contrast with other members ofthe nuclear receptor superfamily, it has proven difficult toidentify physiological early response genes for MR for sev-eral reasons. First, MR bind the physiological glucocorticoidscortisol and corticosterone with affinity comparable to thatof aldosterone (1, 2), and in epithelia these glucocorticoidshave been shown experimentally (3, 4) and clinically (4, 5) toact as mineralocorticoid agonists. Secondly, there is ampleevidence that activation of glucocorticoid receptors (GR) inaldosterone-responsive epithelia mimics the action of agonistactivation of MR (3, 4), suggesting that either receptor can,at least experimentally, regulate the genes affecting trans-epithelial sodium transport. Over the past decade it hasbecome clear that although mechanisms may exist to distin-guish MR and GR transcriptional activities in some tissues(6), the primary determinant of aldosterone selectivity interms of epithelial sodium transport is the enzyme 11b-hy-droxysteroid dehydrogenase type 2 (11bHSD2), which isexpressed at high levels in classical mineralocorticoid targettissues and metabolizes and thus excludes physiological glu-

cocorticoids (corticosterone and cortisol) from these cells(7, 8).

Several aldosterone-responsive genes have been reported:for example, the a1- and b1-subunits of Na1/K1-adenosinetriphosphatase (9); the epithelial sodium channel (ENaC)subunits a, b, and g, which are responsive to aldosterone inrat kidney (10) but differentially induced in distal colon (11);and a factor dubbed channel-inducing factor, which is aldo-sterone induced in colon but not in kidney (12). There is,however, considerable evidence that most, if not all, of theserepresent secondary rather than direct genomic effects ofaldosterone. Importantly, whether they are primary or sec-ondary response genes, they are induced relatively slowly,over hours to days, whereas changes in Na1 transport beginin less than 1 h. On the other hand, it has recently beenreported that aldosterone rapidly and directly induces tran-scription of the serum and glucocorticoid induced kinase(SGK) gene, and that the encoded protein SGK may play acentral role in mediating early aldosterone effects by stim-ulating ENaC-mediated sodium transport (13, 14). Dexa-methasone rapidly induces SGK in the collecting duct (A6)cell line, as does aldosterone in the rat kidney collecting duct(13), in cultured rabbit cortical collecting duct cells (14), andin distal colon (15). In all of experimental situations, however,the conditions were nonphysiological due to the use of cul-tured cell systems, high hormone concentrations, and/orsynthetic glucocorticoid agonists such as dexamethasone.Furthermore, although parameters in cultured cells that re-flect epithelial Na1 transport were measured, no correlationwith the physiologically relevant changes in urinary Na1 andK1 in vivo has been shown.

Hence, although these initial studies suggested that SGK

Received August 4, 2000.Address all correspondence and requests for reprints to: Dr. David

Pearce, Department of Medicine, Box 0532, 513 Parnassus Avenue,University of California, San Francisco, California 94143. E-mail:pearced@medicine.ucsf.edu.

* This work was supported in part by a block grant from the NationalHealth and Medical Research Council of Australia (to J.W.F.) and in partby NIH and American Heart Association grants (to D.P.).

0013-7227/01/$03.00/0 Vol. 142, No. 4Endocrinology Printed in U.S.A.Copyright © 2001 by The Endocrine Society

1587

is an aldosterone-induced mediator of epithelial Na1 trans-port in the kidney, they are primarily premised on in vitrostudies and supraphysiological doses of aldosterone. Thepresent studies were therefore designed to examine the SGKresponse to aldosterone in a more physiological context, todetermine whether the time course and sensitivity of the SGKresponse are consistent with those of an essential regulatorof ENaC activity. To this end we administered physiologicaldoses of aldosterone to rats in vivo, formally eliminating thepossibility of GR activation by the concurrent administrationof the GR antagonist RU486 and measured SGK inductionagainst urinary electrolyte responses.

Time-course and dose-response studies were performed inthis model system, and SGK messenger RNA (mRNA) levelswere determined in colon, kidney, and heart by Northernblot analysis. SGK protein levels in distal colon were esti-mated by Western blot analysis. In addition, kidneys fromthese time-course and dose-response studies were subjectedto in situ hybridization analysis to establish more preciselythe characteristics of the SGK response to aldosterone at thecellular level.

Materials and MethodsAnimals

Animals were treated in accordance with the principles and proce-dures defined in guidelines of the animal ethics committee of the BakerMedical Research Institute and the University of California-San Fran-cisco committee on animal research. Adult male Sprague Dawley rats,approximately 100 g for Exp 1 and 2 and 200–250 g for Exp 3, werebilaterally adrenalectomized (adx) on the day before use, and main-tained on 0.9% NaCl solution to drink overnight without chow forKagawa bioassay (16). At the end of the bioassay procedure, they werekilled by decapitation, and tissues were immediately removed for anal-ysis. Kidney, colon, and heart samples were either snap-frozen in liquidnitrogen for Northern blot analysis or frozen on a dry ice-ethanol bathinto OCT compound (Tissue-Tek, Miles Corp., Elkhart, IN) for in situhybridization.

Kagawa bioassay

Male Sprague Dawley rats (100–120 g) were used. On the day ofexperiment the bladder of each rat (n 5 5–8/group) was emptied at 60min by gentle suprapubic pressure and a whiff of ether, the urine wasdiscarded, and 3 ml 0.9% saline/100 g BW were injected ip. At time zerothe bladder was again emptied, and the urine generated was taken asthe pretreatment sample for determination of Na1 and K1.

In Exp 1 at time zero six groups (n 5 5–8/group) of rats were injectedsc with 10 mg RU 486 plus 0, 0.1, 0.3, 1.0, 3.0, or 10 mg aldosterone/100g BW. Urine from time zero until 60 min postinjection was discarded,and at that time animals were given a second ip injection of 0.9% NaClsolution (3 ml/100 g BW). The final urine collection for all animals wastaken at 120 min, for determination of Na1 and K1 by flame photometry(IL943, Allied Instrumentation Laboratories, Milan, Italy).

In Exp 2, one group of rats (n 5 5) was injected sc with 10 mg RU486at time zero and five groups (n 5 5–8/group) with 10 mg RU486 and 1mg aldosterone/100 g BW. A second ip injection of 3 ml 0.9% saline/100g BW was given 30 min before the final urine collection, which was at30, 60, 90, 120, or 180 min postaldosterone; the final urine collection forthe RU486 alone group was taken 30 min postinjection.

In Exp 3, rats were adx or sham operated. Adx rats were given saline,and all groups had ad libitum access to rat chow. Five days after adx, ratswere anesthetized and implanted sc with mini osmotic pumps (Alzetmodel 2001, Alza Corp., Palo Alto, CA) to deliver either propylene glycol(sham operated and adx 1 vehicle control groups, four rats per group)or 10 mg/day aldosterone. In addition to the pumps, subgroups of rats(n 5 4/group) also received a single sc injection of aldosterone (1 mg/100g BW) or saline (controls). The miniosmotic pumps were presoaked in

saline for 24 h and delivered either vehicle or aldosterone. Rats werekilled 2 or 4 h after treatment; tissues were removed and processedappropriately for Western blot analysis.

Isolation of RNA and Northern blot analysis

Total RNA was extracted from kidney, colon, and heart tissue of adultrats by homogenization in TRIzol reagent (Life Technologies, Inc., GrandIsland, NY). Homogenates were extracted with chloroform, and RNAwas precipitated from the aqueous phase with isopropanol. RNA pelletswere washed in 70% ethanol, dried, and finally redissolved in sterilewater. For Northern blot analysis, total RNA (15 mg) was separated informaldehyde containing 1.2% agarose gel and transferred to Gene-Screen Plus (NEN Life Science Products) by capillary Southern blotting(17). Filters were hybridized in 0.5 m Na2PO4 (pH 7.2), 7% SDS, and 2mm EDTA at 68 C overnight with an antisense 32P-labeled rat SGKriboprobe and washed as previously described (18). All filters wererehybridized with a riboprobe to a complementary DNA for mouseglyceraldehyde-3-phosphate dehydrogenase (GAPDH) to control forRNA loading. All washed filters were exposed for 1–3 days, and levelsof SGK mRNA were analyzed on a phosphorimager (Fuji Photo Film Co.,Ltd., Tokyo, Japan).

In situ hybridization

Frozen tissue sections (8 mm) were cut on a cryostat, thaw-mountedon SuperFrost slides, fixed, acetylated, and hybridized as previouslydescribed (13). A 671-bp fragment (nucleotides 314–985) of rat SGK wasused as a probe. This probe does not cross-hybridize to the other SGKisoforms (2 and 3) as determined by Blast search, and no other isoformswere detected on Northern blots. Briefly, [33P]UPT-labeled antisenseriboprobe was synthesized with T3 polymerase from 1 mg plasmidlinearized with BamHI. Hybridization was performed in a hybridizationsolution containing 50% formamide at 55 C for 16–18 h. Coverslips wereremoved in 2 3 SSC (standard saline citrate); sections were treated withribonuclease A for 30 min, washed in 1 3 SSC and finally washed in 0.1 3SSC at 65 C. They were then passed through alcohol series, air-dried, andexposed to Hyperfilm. Slides were dipped in Kodak emulsion (Eastman

FIG. 1. Response of SGK mRNA, as determined by Northern blotanalysis, to increasing doses of aldosterone administered by sc injec-tion. Rats were pretreated with 10 mg RU486 for 30 min before thefollowing doses of aldosterone were injected per 100 g BW: 0.1, 0.3, 1.0,3.0, and 10 mg. Kidney and colon were collected from each rat (n 55–8) 120 min after aldosterone injection. SGK mRNA, normalized forGAPDH, was determined in kidney (A) and colon (B). Values are themean 6 SEM. #, P , 0.01; *, P , 0.05 (compared with control byANOVA).

1588 ALDOSTERONE REGULATES SGK IN VIVO Endo • 2001Vol. 142 • No. 4

Kodak Co., Rochester, NY), exposed in the dark at 4 C for 10 days,developed, and counterstained with hematoxylin and eosin.

Western blot analysis

Distal colon was homogenized in a Polytron (Brinkmann Instru-ments, Inc., Westbury, NY) for 15 sec in PBS supplemented with 1 mmphenylmethylsulfonylfluoride, 0.5 mm dithiothreitol, and 1 3 proteaseinhibitors (Roche Molecular Biochemicals, Indianapolis, IN). The ho-mogenate was spun at 14,000 rpm for 30 min at 4 C, and the supernatantwas collected and stored in aliquots at 280 C until use. Protein con-centrations were estimated on each lysate with Bradford’s reagent (Bio-Rad Laboratories, Inc., Richmond, CA). Samples were separated on a 8%polyacrylamide gel and transferred electrophoretically to nitrocellulosemembranes (Micron Separations, Westboro, MA). Membranes were im-munoprobed with SGK antibody as previously described (13). Subse-quently, the membranes were stripped in solution containing 62.5 mmTris-Cl (pH 7.6), 2% SDS, and 100 mm b-mercaptoethanol at 50 C for 30min, blocked with 5% milk powder, and immunoprobed with actinantibody as described.

Statistical analysis

SGK/GAPDH mRNA levels and Na1/K1 levels from Kagawa bio-assays were analyzed by one-way ANOVA and Tukey’s post-hoc test,with statistical significance set at P , 0.05.

ResultsAldosterone effects on SGK mRNA dose-response studies

We first examined the induction of SGK mRNA levels (Exp1; see Materials and Methods for details) in rat kidney andcolon over a range of aldosterone doses (0.1–10 mg/100 g BW)that more reflect normal physiological levels. By Northernblot analysis of total kidney RNA, SGK mRNA levels did notchange significantly at the two lowest doses of aldosterone,but increased over the control value at aldosterone doses of1 mg or more (Fig. 1A), with levels doubling at doses of 3 and10 mg aldosterone. In colon there was a trend toward in-creased SGK/GAPDH ratios with progressive increase in thedose of aldosterone used, although no dose of aldosteroneshowed a statistically significant increase at 2 h; at the 10 mgdose of aldosterone, mean values for SGK mRNA/GAPDHmRNA were double the control value (Fig. 1B).

We next examined induction of SGK mRNA by in situhybridization in the kidney over this range of doses of al-dosterone (Fig. 2). In the cortex, hybridization was signifi-cantly enhanced at all doses of aldosterone used, with higher

FIG. 2. Expression of SGK mRNA in the rat kidney by in situ hybridization 2 h after induction with different doses of aldosterone. Four sectionsper animal (and five to eight animals per group) were analyzed. One representative area of an emulsion-dipped section is shown. Hybridizationto SGK mRNA is shown over the glomerulus (A) and distal tubule (B) and in the outer medulla (C) and papilla (D) at 0 (control) and 0.1 mgaldosterone/100 g BW. Hybridization to SGK mRNA is shown in both light- and darkfield at 320 magnification. All experiments were performedin the presence of 10 mg RU486.

ALDOSTERONE REGULATES SGK IN VIVO 1589

levels of hybridization over distal tubular areas of the cortexdetected at 0.1 mg aldosterone (Fig. 2B). Hybridization in theouter medulla appeared to respond slightly to the lowestdose of aldosterone, with uniformly scattered hybridizationonly slightly more intense at this dose, although levels wereconsiderably higher with 0.3 and 1 mg aldosterone (data notshown). Whereas marked hybridization was seen in the cor-tex at 0.1–0.3 mg aldosterone, medullary staining did notappear maximal until a dose of 1 mg aldosterone was used.At higher doses of aldosterone (3 and 10 mg), the hybrid-ization pattern did not differ significantly from that observedat 1 mg (data not shown), although the clustering of SGKsignal in both cortex (over distal tubules) and outer medullawas more marked at the two higher doses of aldosterone.Papillary hybridization in control sections (Fig. 2D) appeared

significantly more intense than in other areas of control kid-neys, and papillary signal intensity remained unchangedover the range of doses of aldosterone tested. Taken togetherthese results indicate that even at low doses of aldosterone(0.1 mg) there is significant induction of SGK mRNA in outermedulla and cortical collecting duct, with papillary expres-sion of SGK constitutively higher but unresponsive toaldosterone.

Aldosterone effects on urinary electrolytes:dose-response studies

To determine whether increases in SGK mRNA in kidneycorrelated with functional effects on urinary Na1 and K1

levels, we measured changes in urinary electrolytes (fromrats in Exp 1) by Kagawa bioassay. Significant changes inurinary electrolyte ratios ([K2/(K1 1 K2)]/[Na2/(Na1 1Na2)]) were seen even at the lowest dose (0.1 mg) of aldo-sterone used, with a progressive increase to a maximal re-sponse at aldosterone doses of 1 mg/100 g BW or more (Fig.3A). No change in urinary sodium concentrations (Na2/Na1;Fig. 3B) was seen until a dose of 0.3 mg aldosterone wasinjected, which produced a halving of sodium levels; athigher aldosterone doses urinary sodium concentrations fellto one quarter of the control value. In contrast, urinary po-tassium concentrations, as determined by the ratio K2/K1,rose to approximately double the control level from the low-est dose (0.1 mg), reaching a maximum level 4-fold the controlvalue at 10 mg (Fig. 3C). The plot of Na2/K2 was notable for

FIG. 3. Effects of dose on urinary electrolyte composition determinedby Kagawa bioassay (A). Effect of aldosterone on post/pre urinary Naratios (B), post/pre urinary K ratios (C), and post urinary Na/K ratios(D) in adx rats pretreated with 10 mg/100 g BW RU486 and injectedsc with 0, 0.1, 0.3, 1.0, 3.0, and 10 mg aldosterone. Rats were killed ingroups at 120 min postinjection. Values are the mean 6 SEM (n 55–8/group). #, P , 0.01; *, P , 0.05 (compared with control byANOVA).

FIG. 4. Time course of regulation of SGK mRNA by aldosterone inkidney (A) and colon (B), as determined by Northern blot analysis.Adrenalectomized rats (n 5 5–8/group) were pretreated for 30 minwith 10 mg/100 g BW RU486 and then injected sc with 1 mg aldoste-rone/100 g BW. Rats were killed in groups at intervals from 30–180min postinjection. One representative blot is shown. Total RNA wasisolated from kidney and colon and probed for SGK, and ratios ofSGK/GAPDH were plotted as the mean 6 SEM. *, P , 0.05 comparedwith time zero by ANOVA.

1590 ALDOSTERONE REGULATES SGK IN VIVO Endo • 2001Vol. 142 • No. 4

the clear efficacy of 0.1 mg aldosterone on this value (Fig. 3D),an effect that increased with increasing doses of aldosterone.

Aldosterone effects on SGK mRNA: time-course studies

SGK mRNA levels were analyzed by Northern blot and insitu hybridization. Northern blot analysis of whole kidneyand colon RNA (from Exp 2) showed that induction of SGKmRNA levels by aldosterone was maximal in both kidney

(Fig. 4A) and whole colon (Fig. 4B) within 60 min of injection,with levels approximately double those in control rats killedat time zero. SGK levels in heart were 10 times lower thanthose in kidney and did not change at the time points ex-amined (data not shown). The extent of induction of SGKmRNA by aldosterone appeared similar in kidney and wholecolon, with the maximum in kidney double the control value(17 6 3 at 90 min vs. 8 6 1 at time zero; P , 0.05) and that

FIG. 5. A, Time course of induction of SGK mRNA by aldosterone in the rat kidney by in situ hybridization over 120 min. The 33P-labeled ratSGK antisense riboprobe was hybridized to tissue sections of whole kidney of adx rats at 0, 30, 60, and 120 min after injection of aldosterone(1 mg/100 g BW). Four sections per animal (and five to eight animals per group) were analyzed. One representative section from each time pointis shown. B, Hybridization of SGK probe is shown in both light- and darkfield at 320 magnification over glomeruli (G) in the cortex (A), distaltubules (DT) of the cortex (B), outer medulla (C), and inner medulla (D) of the kidney. Results are shown for 0 (control) and 60 min after inductionby aldosterone. All experiments were performed in the presence of 10 mg RU486.

ALDOSTERONE REGULATES SGK IN VIVO 1591

in colon almost double (2.3 6 0.3 at 60 min vs. 1.2 6 0.2; P ,0.05). At other time points examined, there was a trendtoward an increase in SGK mRNA over the control value,but the levels did not attain statistical significance (Fig. 4, Aand B).

In parallel, we performed a more detailed analysis of SGKinduction in the kidney by in situ hybridization after injectionof a 1-mg dose of aldosterone (Fig. 5A). Hybridization to SGKmRNA is shown in both light- and darkfield at 0 and 60 minafter injection of aldosterone in both cortex and medulla (Fig.5B, A–D). In the cortex hybridization was seen in both theglomerulus and distal nephron, but was aldosterone respon-sive only in the distal nephron (Fig. 5B, B and C). Admin-istration of aldosterone did not change hybridization to glo-meruli at 30 min or any other time point examined. AlthoughSGK mRNA began to appear over few distal tubules in the

cortex by 30 min, no significant increase in hybridization toSGK mRNA was seen at this time point in the outer med-ullary collecting ducts or the inner medullary area. By 60 min,however, substantial levels of hybridization to SGK mRNAwere clearly seen over many collecting tubules in both cortexand outer medulla (Fig. 5B, B and C). At 90 and 120 min, SGKhybridization patterns did not differ significantly from thoseat 60 min, although clustering of signal in outer medullaincreased at 90 and 120 min (data not shown). High levels ofSGK expression were observed in inner medulla/papilla,with levels of expression remaining constant despite aldo-sterone treatment, consistent with a recent report that SGKmRNA remains unchanged in papilla but is responsive toaldosterone treatment in cortex and outer medulla, as de-termined by Northern blot analysis (19).

Aldosterone effects on urinary electrolytes:time-course studies

Significant changes in urinary electrolyte ratios were seenin samples excreted over the 60- to 90-min period postinjec-tion, with the change in ratio maintained for the subsequenturine collections (Fig. 6A). Changes in the urinary sodiumratio Na2/Na1, taken pre- and postaldosterone treatment(Fig. 6B), did not appear until the 90 min point, when ahalving of sodium excretion was seen. This decrease wassustained for each of the collection periods up to 180 min. Incontrast, potassium excretion, as determined by the K2/K1ratio, rose progressively from the second collection period(30–60 min) postaldosterone treatment, reaching maximumlevels at 120 min.

SGK levels in colon: Western blot analysis

Recently, major effects of aldosterone were observed indistal descending colon with increased b and g ENaC ex-pression, whereas expression of these subunits remainedunresponsive in more proximal descending colon (11). Inaddition, SGK message was induced in distal descendingcolon 2 h after aldosterone administration. The modest effectof aldosterone on SGK induction in whole colon describedabove probably reflects this graduation in response to aldo-sterone throughout the colon. We therefore examined thestatus of SGK protein in distal descending colon (Exp 3) inadx, sham-operated, and adx plus aldosterone-treated rats.

FIG. 6. The effects on urinary electrolyte composition as determinedby Kagawa bioassay at different time points (A). Adx rats (n 5 5–8/group) were pretreated for 30 min with 10 mg/100 g BW RU486, andthen injected sc with 1 mg aldosterone/100 g BW. Rats were killed ingroups at intervals from 30–180 min postinjection. Effects of time onpost/pre urinary Na ratios (B), post/pre urinary K ratios (C), and posturinary Na/K ratios (D) in adx rats. Values are the mean 6 SEM. #, P ,0.01; *, P , 0.05 (compared with control by ANOVA).

FIG. 7. SGK immunoreactive protein is induced by aldosterone treat-ment in late distal colon. Total protein was separated on a polyacryl-amide gel as described in Materials and Methods. The blot shown hereis a representative of three or four animals per group. Both phos-phorylated and unphosphorylated SGK bands are visible upon hor-mone treatment. Top panel, SGK antibody; bottom panel, same blotstripped and immunoprobed with actin.

1592 ALDOSTERONE REGULATES SGK IN VIVO Endo • 2001Vol. 142 • No. 4

Aldosterone treatment resulted in a significant increase inimmunoreactive SGK protein 2 and 4 h after hormone treat-ment (Fig. 7), consistent with the induction of SGK proteinseen 2 h after dexamethasone treatment in A6 cells (13).

Discussion

The findings in the present study confirm and extend thosefor SGK from this laboratory (13) and others (15, 19–21).These initial studies provided strong evidence in A6 cells,isolated cortical collecting tubule cells, and adx rats that SGKcould be induced by dexamethasone (in A6 cells) or by veryhigh doses of aldosterone. Whether SGK induction by aldo-sterone occurred at physiological concentrations of hormoneand with a time course consistent with the effects on urinaryelectrolyte excretion remained to be established. The presentstudy thus set out to examine more precisely the time courseand dose response of SGK induction by aldosterone and toshow that SGK induction by aldosterone in rat collecting ductis via MR. To this end, we performed time-course and dose-response studies, measuring mRNA levels by Northernblots, localizing the renal mRNA response by in situ hybrid-ization, and comparing the tissue responses so obtained withthe end-organ effects, those of changes in urinary Na1 andK1 excretion. Parallel changes in SGK mRNA and proteinlevels were seen in colon, an epithelial target tissue similarto kidney distal nephron; in heart (data not shown) nochanges in SGK mRNA in response to aldosterone weredetected.

In previous studies of adrenalectomized rats (13) a highdose (50 mg/100 g BW) of aldosterone was used for in situhybridization studies. At this dose, the effect of aldosteronecould possibly be mediated via GR, for which aldosteronehas low, but significant, affinity. Actions on urinary sodiumand potassium excretion via GR are not merely a formalpossibility; in both cultured cortical collecting tubule prep-arations (3) and adx rats (4) highly selective glucocorticoidshave been shown to be equipotent to aldosterone when theirmetabolism by 11bHSD2 is blocked. In addition, the originalcloning of Xenopus SGK from A6 cells exploited this nonse-lectivity, with dexamethasone used as a ligand to induce SGKexpression via GR. Very recently, Brennan and Fuller haveshown dexamethasone to have effects on SGK mRNA in-duction in rat kidney and colon at least equal to those ofaldosterone (15). For this reason we included RU486 withaldosterone in all studies to minimize the possibility ofsubstantial GR occupancy (22, 23) and activation by aldo-sterone, however unlikely that might be at the doses usedand given its relatively low affinity for GR. Further evidencefor aldosterone acting via MR comes from studies in whichadministration of RU26752, an MR antagonist, preventedaldosterone-mediated induction of SGK (19).

In terms of both mRNA levels and urinary electrolytes, theED50 for aldosterone appears to be about 0.3 mg/100 g BW,with the bioassay in this instance proving a more precisemeasurement system. This reflects in large part the fact thatrelatively high levels of SGK are expressed in renal glomer-uli, which do not respond to adrenalectomy or aldosteroneadministration; there is, therefore, a high background levelof SGK expression in the kidney, against which probably

substantial increases in a relative minority of cells must beset. This is very clearly shown by the results of the in situstudies at the lower end of the dose range used, at shortertime points. Although such data are semiquantitative at bestand not easily subjected to statistical analysis, they providegraphic evidence for a clear induction of SGK in corticalcollecting duct and papilla at the lowest dose of aldosteroneused (0.1 mg/100 g BW) and in all tissues at 0.3 mg/100 g BW.

The time-course studies described in this paper confirm,both in situ and by Northern blot assay, that SGK is rapidlyinduced, and that over the 3-h period of study chosen, SGKmRNA levels in kidney are returning to baseline at a timewhen the effect on urinary electrolytes (presumably viaENaC activated by the induced SGK) remains high. Again,the in situ time-course studies show clear effects in a rangeof renal target tissues at 30 min, before any effects on urinaryNa1 or K1 can be seen, as expected. The effects of aldosteroneon urinary K1 in this study appear to precede those on Na1.This may reflect an effect of aldosterone that is independentof SGK, given the clear demonstration of the inactivity ofSGK in terms of ROMK2 activation (13), although furtherexamination of electrolyte excretion (for example, urinary Naexcretion volume or urinary Na under conditions of collect-ing duct impermeability to water) will be required to explorethe mechanisms underlying this difference. A similar dis-junction in the time course of mineralocorticoid effects onurinary Na1 and K1 excretion has previously been describedby Morris et al. (24).

In the whole colon, the changes paralleled those in thekidney, consistent with the recent report that SGK mRNA isinduced by aldosterone in distal colon (15, 19). It has alsobeen shown that the distal descending colon is sensitive tonanomolar levels of aldosterone, and that expression of otheraldosteron- responsive genes (ENaCs) is differentially reg-ulated over the course of the distal colon (11). Our studiesused whole colon, which may thus have contributed to therelatively modest response observed. However, when latedistal colon was examined, we observed a robust inductionin SGK protein after aldosterone treatment. Recently, SGKprotein has been shown to be induced in the distal nephronby aldosterone (25) in mouse kidney. It is thus clear that theepithelial response of SGK in vivo to aldosterone can beextended to the colon as an additional epithelial tissue, sug-gesting that SGK may be an essential mediator of aldoste-rone-stimulated ion transport in all classical mineralocorti-coid target tissues.

Although MR have been described in a variety of othertissues (26), as has SGK (27), it is possible that SGK inductionmay be a uniquely epithelial response to increased aldoste-rone; one fragment of evidence in support of this are theunchanged cardiac levels of SGK in the present studies overthe fairly substantial (0.1–10 mg/100 g BW) range of aldo-sterone doses used. It is also possible that other isoforms ofSGK (28) are present in the heart, which may be responsiveto aldosterone treatment, but that their mRNAs are unde-tectable with the probe used in this study. Finally, it is clearthat these studies on MR-mediated effects of aldosterone onSGK do not exclude the possibility of additional modulatorsof enzyme activity in the kidney and colon.

Other agents are known to induce SGK expression.

ALDOSTERONE REGULATES SGK IN VIVO 1593

Whereas SGK was initially identified as a SGK from a ratmammary tumor cell line (27), in the brain, SGK is not re-sponsive to glucocorticoid treatment, but is induced aftercentral nervous system injury (29). Whereas in liver cells SGKis responsive to hyperosmotic shock (21), in A6 cells SGK isresponsive to osmotic shock (Rozansky, D. J., and D. Pearce,unpublished results) as well as to glucocorticoids (13). SGKwhen coexpressed with ENaC subunits in Xenopus oocytesresults in a marked increase in the amiloride-sensitive so-dium current (13), and similar observations have been madeby others (14, 19). SGK is phosphorylated at specific threo-nine residues (30), and this is mediated via the action of thephosphatidylinositol 3-kinase signaling pathway (31, 32). Al-though prevention of phosphorylation of SGK by inhibitorsof phosphatidylinositol 3-kinase results in inhibition ofhormone-induced activation of the sodium current (31, 32),the substrate(s) phosphorylated by SGK remains unidenti-fied, and the exact mechanism by which SGK augmentsENaC-mediated increases in sodium transport remains to bedetermined.

In summary, the present studies confirm that SGK is animmediate early gene regulated by aldosterone via MR, withresponses to aldosterone at low (0.1 mg/100 g BW) doses andwithin 30 min of administration. In terms of both mRNAlevels and the urinary electrolyte response, the ED50 for al-dosterone is on the order of 0.3 mg/100 g BW, as a single scdose. Kagawa bioassay changes, in terms of urinary electro-lyte ratios, follow SGK mRNA profiles in a temporal sense,with some evidence for a slightly earlier effect on urinary K1

than Na1; colonic levels of SGK mRNA appear lower thanthose in the kidney, and both mRNA and protein levelschange in parallel in response to aldosterone, in contrast withthose in the heart. Additional studies of SGK activation andother aldosterone-regulated genes and their products, arerequired for a more comprehensive understanding of aldo-sterone action in epithelial and nonepithelial target tissues.

Acknowledgments

We thank Nicola Solomon for research assistance in the Cole labo-ratory. Mary F. Dallman is gratefully acknowledged for her expert helpwith adrenalectomy and providing equipment for in situ studies. Dr. G.Firestone kindly provided the SGK antibody. Jian Wang’s technical helpin the Pearce laboratory is also appreciated.

References

1. Krozowski ZS, Funder JW 1983 Renal mineralocorticoid receptors and hip-pocampal corticosterone-binding species have identical intrinsic steroid spec-ificity. Proc Natl Acad Sci USA 80:6056–6060

2. Arriza JL, Weinberger C, Cerelli G, Glaser TM, Handelin BL, Housman DE,Evans RM 1987 Cloning of human mineralocorticoid receptor complementaryDNA: structural and functional kinship with the glucocorticoid receptor. Sci-ence 237:268–275

3. Naray-Fejes-Toth A, Fejes-Toth G 1990 Glucocorticoid receptors mediatemineralocorticoid-like effects in cultured collecting duct cells. Am J Physiol259:F672–F678

4. Funder JW, Pearce PT, Myles K, Roy LP 1990 Apparent mineralocorticoidexcess, pseudohypoaldosteronism, and urinary electrolyte excretion: toward aredefinition of mineralocorticoid action. FASEB J 4:3234–3238

5. Oberfield S, Levine L, Carey R, Greig F, Ulick S, New M 1983 Metabolic andblood pressure responses to hydrocortisone in the syndrome of apparentmineralocorticoid excess. J Clin Endocrinol Metab 56:332–339

6. Pearce D, Yamamoto KR 1993 Mineralocorticoid and glucocorticoid receptoractivities distinguished by nonreceptor factors at a composite response ele-ment. Science 259:1161–1165

7. Funder JW, Pearce PT, Smith R, Smith AI 1988 Mineralocorticoid action:target-tissue specificity is enzyme, not receptor, mediated. Science 242:583–585

8. Edwards CR, Stewart PM, Burt D, Brett L, McIntyre MA, Sutanto WS, deKloet ER, Monder C 1988 Localisation of 11b-hydroxysteroid dehydrogenase-tissue specific protector of the mineralocorticoid receptor. Lancet 2:986–989

9. Ikeda U, Hyman R, Smith T, Medford R 1991 Aldosterone-mediated regu-lation of Na1,K1-ATPase gene expression in adult and neonatal rat cardio-cytes. J Biol Chem 266:12058–12066

10. Masilamani S, Kim G, Mitchell C, Wade J, Knepper M 1999 Aldosterone-mediated regulation of ENaCa, b, and g subunit proteins in rat kidney. J ClinInvest 104:R19–R23

11. Epple HJ, Amasheh S, Mankertz J, Goltz M, Schulzke JD Fromm M 2000Early aldosterone effect in distal colon by transcriptional regulation of ENaCsubunits. Am J Physiol Gastrointest Liver Physiol 278:G718–G724

12. Brennan FE, Fuller PJ 1999 Acute regulation by corticosteroids of channel-inducing factor gene messenger ribonucleic acid in the distal colon. Endocri-nology 140:1213–1218

13. Chen S, Bhargava A, Mastroberardino L, Meijer O, Wang J, Buse P, FirestoneG, Verrey F, Pearce D 1999 Epithelial sodium channel regulated by aldoste-rone-induced protein SGK. Proc Natl Acad Sci USA 96:2514–2519

14. Naray-Fejes-Toth A, Canessa C, Cleaveland ES, Aldrich G Fejes-Toth G 1999SGK is an aldosterone-induced kinase in the renal collecting duct. J Biol Chem274:16973–16978

15. Brennan FE, Fuller PJ 2000 Rapid upregulation of serum and glucocorticoid-regulated kinase (sgk) gene expression by corticosteroids in vivo. Mol CellEndocrinol 166:129–136

16. Kagawa CM, Bouska DJ, Anderson ML 1964 Relationship of plasma alda-diene levels and antimineralocorticoid effects of spironolactone in the labo-ratory. Proc Soc Exp Biol Med 115:873–840

17. Sambrook J, Fritsch E, Maniatis T 1989 Molecular Cloning: A LaboratoryManual, Ed 2. Cold Spring Harbor Laboratory Press, Cold Spring Harbor

18. Cole T, Harris H, Hoong I, Solomon N, Smith R, Krozowski Z, Fullerton M1999 The glucocorticoid receptor is essential for maintaining basal and dexa-methasone-induced repression of the murine corticosterone-binding globulingene. Mol Cell Endocrinol 154:29–36

19. Shigaev A, Asher C, Latter H, Garty H, Reuveny E 2000 Regulation of sgk byaldosterone and its effects on the epithelial Na1 channel. Am J Physiol RenalPhysiol 278:F613–F619

20. Naray-Fejes-Toth A, Fejes-Toth G 2000 The SGK, an aldosterone-inducedgene in mineralocorticoid target cells, regulates the epithelial sodium channel.Kidney Int 57:1290–1294

21. Waldegger S, Barth P, Raber G, Lang F 1997 Cloning and characterization ofa putative human serine/threonine protein kinase transcriptionally modifiedduring anisotonic and isotonic alterations of cell volume. Proc Natl Acad SciUSA 94:4440–4445

22. Funder J, Myles K 1996 Exclusion of corticosterone from epithelial mineralo-corticoid receptors is insufficient for selectivity of aldosterone action: in vivobinding studies. Endocrinology 137:5264–5268

23. Wagner BL, Pollio G, Giangrande P, Webster JC, Breslin M, Mais DE, CookCE, Vedeckis WV, Cidlowski JA, McDonnell DP 1999 The novel progester-one receptor antagonists RTI 3021–012 and RTI 3021–022 exhibit complexglucocorticoid receptor antagonist activities: implications for the developmentof dissociated antiprogestins. Endocrinology 140:1449–1458

24. Morris D, Brem A 1987 Metabolic derivatives of aldosterone. Am J Physiol252:F365–F373

25. Loffing J, Zecevic M, Feraille E, Kaissling B, Asher C, Rossier BC, FirestoneGL, Pearce D, Verrey F, Aldosterone induces rapid apical translocation ofENaC in early portion of renal collecting system: role of SGK induction andactivity. Am J Physiol-Renal Physiol, in press

26. Funder JW 1997 Glucocorticoid and mineralocorticoid receptors: biology andclinical relevance. Annu Rev Med 48:231–240

27. Webster M, Goya L, Ge Y, Maiyar A, Firestone G 1993 Characterization ofSGK, a novel member of the serine/threonine protein kinase gene family whichis transcriptionally induced by glucocorticoids and serum. Mol Cell Biol13:2031–2040

28. Kobayashi T, Deak M, Morrice N, Cohen P 1999 Characterization of thestructure and regulation of two novel isoforms of serum and glucocorticoid-induced protein kinase. Biochem J 344:189–197

29. Imaizumi K, Tsuda M, Wanaka A, Tohyama M, Takagi T 1994 Differentialexpression of sgk mRNA, a member of the Ser/Thr protein kinase gene family,in rat brain after CNS injury. Brain Res Mol Brain Res 26:189–196

30. Kobayashi T, Cohen P 1999 Activation of serum- and glucocorticoid-regulatedprotein kinase by agonists that activate phosphatidylinositide 3-kinase is me-diated by 3-phosphoinositide-dependent protein kinase-1 (PDK1) and PDK2.Biochem J 339:319–28

31. Blazer-Yost BL, Paunescu TG, Helman SI, Lee KD, Vlahos CJ 1999 Phos-phoinositide 3-kinase is required for aldosterone-regulated sodium reabsorp-tion. Am J Physiol 277:C531–C536

32. Wang J, Barbry P, Maiyar AC, Rozansky DJ, Bhargava A, Leong M, FirestoneGL, Pearce D 2001 SGK integrates insulin and mineralocorticoid regulation ofepithelial sodium transport. Am J Physiol Renal 280:F303–F313

1594 ALDOSTERONE REGULATES SGK IN VIVO Endo • 2001Vol. 142 • No. 4