Endocrinology and Metabolism

Glucocorticoids stimulate human sgk1 gene expression by activation of a GRE in its 5′-flanking region

Omar A. Itani, Kang Z. Liu, Kristyn L. Cornish, Jason R. Campbell, Christie P. Thomas


In lung and collecting duct epithelia, glucocorticoid (GC)-stimulated Na+ transport is preceded by an increase in the protein kinase sgk1, which in turn regulates the activity of the epithelial Na+ channel (ENaC). We investigated the mechanism for GC-regulated human sgk1 expression in lung and renal epithelia. sgk1 mRNA was increased in these epithelia by GCs, and this was inhibited by actinomycin D and superinduced by cycloheximide, consistent with a transcriptional effect that did not require protein synthesis. To understand the basis for transcriptional regulation, the transcription initiation site was mapped and the 5′-flanking region cloned by PCR. A 3-kb fragment of the upstream region was coupled to luciferase and transfected into A549 cells. By deletion analysis, an imperfect GC response element (GRE) was identified that was necessary and sufficient for GC responsiveness. When tested with cell extracts, a specific protein recognized by an anti-GC receptor (GR) antibody bound the GRE in gel mobility shift assays. We conclude that GCs stimulate sgk1 expression in human epithelial cells via activation of a GRE in the 5′-flanking region of sgk1.

  • epithelial Na+ channel
  • glucocorticoid response element
  • corticosteroids
  • airway epithelia
  • gene transcription
  • gel mobility shift assays

na+ is absorbed from the lumen of the collecting duct, the distal colon, airway and alveolar epithelia, sweat glands, and salivary ducts principally via an amiloride-sensitive Na+ channel. The molecular correlate of this Na+ transport pathway is the epithelial sodium channel (ENaC), composed of a heteromultimer of α-, β-, and γ-subunits (11, 28). Two of the principal regulators of Na+ absorption at these sites are the corticosteroids, aldosterone and cortisol, which act within minutes to hours to increase Na+ transport. Much recent attention has been focused on understanding the mechanism of corticosteroid action to stimulate Na+ transport in each of these sites.

The effects of corticosteroids on Na+ transport can be substantially reduced by inhibitors of transcription and translation, clearly suggesting that new RNA synthesis is required for the increase in Na+ transport. Among the putative targets for corticosteroid action are the ENaC subunits themselves. There is now considerable evidence that ENaC expression and function can be regulated by glucocorticoids (GCs) and aldosterone. When given during the antenatal period to the developing fetus or when infused during adult life, exogenous GCs increase lung ENaC expression dramatically (29, 31). Several studies have also reported that either dexamethasone or aldosterone infusion leads to increased expression of α-ENaC mRNA in the renal cortex and β- and γ-ENaC mRNA in the distal colon (1, 10, 29). The rise in α-ENaC mRNA expression is due to the trans-activation of a GC response element (GRE) in the 5′-flanking region of the α-ENaC gene, but the basis for the corticosteroid-mediated increase in β- and γ-ENaC remains unknown (17, 19, 23, 25).

Although ENaC subunits are regulated by corticosteroids, in some tissues the increase in Na+ transport precedes the increase in ENaC, suggesting that there may be other steroid-regulated proteins that are required for the early increase in Na+ transport. Two groups simultaneously identified another gene product, the serum- and GC-regulated serine/threonine protein kinase sgk (now renamedsgk1), as an aldosterone-induced gene in the amphibian, rodent, and rabbit kidney (6, 21). Sgk1 was first described in rat mammary epithelia and rat-2 fibroblasts as an immediate early response gene that is rapidly induced in vivo by serum and GCs (36). One of the targets of sgk1 action is ENaC: co-expression of sgk1 with α,β,γ-ENaC mRNA in Xenopusoocytes significantly enhances Na+ current (6, 21,26, 32). This effect of sgk1 is achieved by increasing the number of ENaC channels assembled at the cell surface (8,33) and is the most direct evidence that sgk1 positively regulates the function of ENaC.

Despite the abundant evidence that GCs regulate sgk1 in lower mammals, an earlier report indicated that GC did not regulate human sgk1 (34). We (14) and others (22) have recently shown that, in human airway epithelia, GCs regulate sgk1. In this study we demonstrate that GCs increase sgk1 expression in human lung and renal cortical epithelial cells via an increase in transcription of thesgk1 gene. We also confirm that an imperfect hormone response element in the human sgk1 5′-flanking region binds the GC receptor (GR) and is required for the GC-mediated transcription of sgk1.



Dexamethasone, aldosterone, and cycloheximide were purchased from Sigma Biochemicals (St. Louis, MO). Actinomycin D was obtained from Roche Molecular Biochemicals (Indianapolis, IN), and the radionucleotides [α-32P]UTP, [γ-32P]ATP, and [α-32P]dCTP came from NEN Life Science Products (Boston, MA). RU-38486 was a generous gift from Roussel Uclaf (Romainville, France), and culture media were obtained from Life Technologies (Gaithersburg, MD) or from Clonetics Cell Systems (Cambrex, East Rutherford, NJ). DNA sequencing and synthesis were a service provided by the University of Iowa DNA core facility. The monoclonal anti-GR antibody, BuGR, was a gift from Thomas Schmidt, Department of Physiology, University of Iowa. Iowa City, IA.

Tissue culture, RNA extraction, and ribonuclease protection assay.

The H441 and A549 human lung epithelial cells were cultured as previously described (14). The human renal cortical epithelial (HRCE) cell line was obtained from Clonetics cell systems (Cambrex) and maintained in human renal epithelial growth medium supplemented with 10 ng/ml epidermal growth factor, 500 ng/ml hydrocortisone, 500 ng/ml epinephrine, 5 μg/ml insulin, 6.5 ng/ml triiodothyronine, 10 μg/ml transferrin, and 0.5% fetal bovine serum. To examine the effects of dexamethasone or aldosterone on gene expression, cell cultures were switched to serum-free hormone-free medium and then were treated with 100 nM steroids for various time periods or with various concentrations of steroid or vehicle for 2 h. Some experiments were done in the presence of 1 μM actinomycin D or 10 μM cycloheximide or vehicle. Total RNA was extracted from H441, A549, and HRCE cells, as previously described (19).

Steady-state levels of sgk1 mRNA were measured by ribonuclease protection assay (RPA) in H441, A549, and HRCE cells grown as monolayers in six-well plates. To obtain human sgk1 (hsgk1) cDNA, H441 RNA was reverse transcribed using oligo-(dT) and M-MLV reverse transcriptase, and first-strand cDNA was subjected to PCR amplification using primers 5′-ACGTCTTTCTGTCTCCCCG and 5′-GGCTCCACCAAAAGGCTAAC. A 181-bp ApaI-DraI fragment of hsgk1 was then cloned into pCDNA3 (Invitrogen), linearized, and used to synthesize an antisense radiolabeled cRNA probe from the T7 promoter. To measuresgk1 gene expression, 10 μg of RNA samples were hybridized with the sgk1 cRNA probe and an 18S rRNA probe as control for gel loading. The method for solution hybridization, nuclease digestion, and identification of nuclease-protected products by PAGE has been previously described (14, 25).

5′ Rapid amplification of cDNA ends.

Total RNA was extracted from dexamethasone-treated H441 cells and used to construct a modified cDNA library by use of the RLM-RACE kit (Ambion, Austin, TX), following the manufacturer's instructions but with some modifications. Briefly, 10 μg of RNA were first treated with calf intestinal phosphatase (CIP) in 1× CIP buffer at 37°C for 60 min, phenol extracted, and then incubated with tobacco acid pyrophosphatase (TAP) in 1× TAP buffer at 37°C for 60 min. A single-stranded 5′ rapid amplification of cDNA ends (RACE) adapter was added to decapped mRNA with T4 RNA ligase at 37°C for 60 min, and modified RNA was stored at −20°C. Two microliters of adaptor-ligated RNA were then reverse transcribed with oligo-(dT) and M-MLV reverse transcriptase (RT) in 1× RT buffer with 4 dNTPs and RNAse inhibitor at 42°C for 60 min to create first-strand cDNA. Gene-specific reverse primers X2: 5′-AGTCGTTCAGACCCATCC [+113 to +94 from the original initiation codon (see Fig. 4A)] (34) and X1: 5′-AGCAGCCTCAGTTTTCACC (+23 to +5 from the initiation codon) were used sequentially with adapter-specific primers in PCR reactions by use ofTaq polymerase. The amplification reactions occurred for 35 cycles each at 94°C for 30 s, with annealing at 53 or 58°C for 30 s and extension at 72°C for 3 min. Reaction products were cloned into pCR-XLTOPO (Invitrogen, Carlsbad, CA) and sequenced.

Primer extension.

Total RNA from dexamethasone-treated H441 cells or yeast RNA was used in primer extension reactions, with the nested sgk1 primer used in 5′ RACE reactions. Primer (12.5 pmol) was end-labeled with T4 polynucleotide kinase and [γ-32P]ATP and purified using a G-25 Sephadex column (Quick Spin, Roche Molecular Biochemicals, Indianapolis, IN). Labeled primer [100,000 counts/min (cpm)] was combined with 10 μg of RNA and denatured at 65°C for 5 min and then added to 15 units of Thermoscript RT (Invitrogen) in a reaction mixture that included 50 mM Tris-acetate, pH 8.4, 75 mM K-acetate, 8 mM Mg-acetate, 5 mM DTT, 1.25 mM dNTPs, and 2 μl RNAsin at 60°C for 60 min. Primer extended products were run alongside a labeled 50-bp ladder on a polyacrylamide gel, as previously described (2).

Cloning of the 5′-flanking region of human sgk1.

Sequence information from the 5′ end of the locus was found in GenBank (accession no. AL 355881). DNA sequence was analyzed for transcription factor binding motifs with Omiga (Oxford Molecular, Campbell, CA) and for CpG islands with GRAIL DNA sequence analysis software athttp://compbio.ornl.gov/Grail-1.3/. To clone the ∼3,000 bp of the proximal 5′-flanking region of sgk1 upstream of exon 1, primer F4: 5′-TGCGTGCTGGGGGTGTAATAAA and primer R4: 5′-CCATGCCCCTCATCCTGGAGTA were used to amplify a DNA fragment from human genomic DNA. This fragment, which includes 117 nt of the 5′ UTR of sgk1, was directionally subcloned into pGL3basic (Promega, Madison, WI) upstream of the firefly (Photinus pyralis) luciferase coding region. Deletion variants of this construct were created by using internal restriction enzyme sitesSpeI, SacI, and MscI.

To test the GRE with a heterologous promoter, a double-stranded oligonucleotide was constructed by annealing primer sgk1_GRE (S): 5′-CGGAACGGACAAAATGTTCTCAGAC and primer sgk1_GRE (AS): 5′-CTAGGTCTGAGAACATTTTGTCCGTTCCGAGCT. This oligonucleotide withAvrII and SacI ends was phosphorylated with T4 polynucleotide kinase and then introduced as two copies at anAvrII site (−142) upstream of the promoter of human α-ENaC-1 (α-hENaC-1) in a firefly luciferase construct (25).

To mutate the GRE in the human sgk1 (hsgk1) promoter, the technique of splicing by overlap extension was used to substitute 2 bp in this cis-element (13). Briefly, the sgk1 5′-flanking region was amplified as two separate fragments overlapping at the GRE with primer pairs F4 and mutGRE_AS2: 5′-GAAGTAATCTCTGAGATGATTTTGTCCGT TCCGC and mutGRE_S2: 5′-GCGGAACGGACAAAATCA TCTCAGAGATTACTTC and R4. The PCR products were polished with the Klenow fragment of DNA polymerase, gel purified, combined, and reamplified for 25 cycles with primers F4 and R4 to splice both fragments together. The sequence extending from −1899 to +117 was then cloned into pGL3basic for transfection studies.

Transient transfection and analysis of reporter activity.

A549 cells were grown in 24-well plates until subconfluent and were then transfected, using LipofectAMINE Plus (Life Technologies), with 1 μg of the sgk1 promoter-reporter construct, and as a control for transfection efficiency, 1 μg of pRL-SV40 (Promega), a plasmid vector in which the SV40 viral promoter drives sea pansy luciferase (Renilla reniformis). In some cases, the α-hENaC promoter-luciferase construct (−1388 + 55) that contains a functional GRE was used as a positive control (19). On the day after transfection, cells were treated with 100 nM dexamethasone or vehicle, and 24 h later cell lysates were prepared, and reporter gene activity was performed with the Dual Luciferase Assay Kit (Promega) as previously described.

Preparation of whole cell extracts and the gel mobility shift assay.

A549 cells were grown until nearly confluent, switched to serum-free media, and treated with 100 nM dexamethasone for 24 h. Cells were then scraped up into PBS using a rubber policeman and washed twice with PBS. The cell pellet was resuspended in an extraction buffer containing 25 mM Tris, pH 7.9, 400 mM NaCl, 0.1 mM EDTA, 0.3 mM DTT, 0.5 mM phenylmethylsulfonyl fluoride, and 1.9 μg/ml aprotinin and was subjected to three freeze-thaw cycles from −80°C to 4°C. The cell lysate was centrifuged at 15,000 g for 15 min at 4°C, and the supernatant was collected. Protein concentration was calculated by the Bradford method and then stored in aliquots at −80°C.

Double-stranded oligonucleotides that correspond to thehsgk1 GRE (discussed above) or the α-hENaC GRE (25) were labeled by a fill-in reaction with [α-32P]dCTP by use of the Klenow fragment of DNA polymerase and then purified by G-25 Sephadex chromatography. Labeled probe (50,000 cpm) was incubated with or without 1–2.5 μg whole cell extracts at 4°C for 20 min in a reaction mixture that contained 4 mM HEPES, pH 7.9, 10 mM KCl, 1.25 mM MgCl2, 40 μM DTT, 16.7 ng/μl poly(dI-dC), 4 mM spermidine concentration, and 10% glycerol. A 50-fold excess of nonradioactive oligonucleotides was used for competition experiments. For supershift assays, A549 extracts were preincubated with 2 μl of the anti-GR Ab BuGR at 4°C for 20 min and were then added to labeled probe. Samples were analyzed on a 3.5% nondenaturing polyacrylamide gel (19:1 acrylamide-bis) in 0.3× Tris-borate-EDTA buffer run at 150 V, and the gel was dried and subjected to autoradiography.


We have previously shown that H441 cells, an airway epithelial cell line, and A549 cells, an alveolar epithelial cell line, have regulated expression of α-, β-, and γ-ENaC subunits and sgk1 (14, 25). These findings were in contrast to earlier studies (34) reporting that GC did not stimulate the human sgk1 transcript in HepG2 cells. To determine the molecular basis for the increase in sgk1 expression in airway and alveolar cells, we performed a more detailed study of the characteristics of the corticosteroid response, focusing on early time points. Dexamethasone stimulated the expression of sgk1 in H441 and A549 cells in a time-dependent manner, with the first response seen at 40–60 min (Fig. 1, A and B). Whereas aldosterone had no effect in H441 cells, a small effect was seen in A549 cells at a single time point. In H441 and A549 cells, dexamethasone increased sgk1 expression in a dose-dependent manner, with stimulation beginning at 3.3 nM (Fig. 1, C andD). The maximal response in H441 cells and A549 cells was 2- to 5-fold and 4- to 10-fold over basal conditions. We then tested the effect of actinomycin D, a general transcription inhibitor, and cycloheximide, a protein synthesis inhibitor on basal and GC-stimulatedsgk1 expression in A549 cells. Actinomycin D blocked both basal and GC-stimulated sgk1 expression, consistent with a transcriptional effect of GC on sgk1 gene expression (Fig.2 A). Cycloheximide increased basal sgk1 expression and further enhanced the GC effect onsgk1 expression (Fig. 2 A), an effect that we have also seen with α-, β-, and γ-ENaC subunits (14, 25) and is a well described phenomenon for early response genes. These results indicated to us that the effect of GC on sgk1 gene transcription was direct and did not involve synthesis of an intermediary protein. To determine whether the effect of dexamethasone and aldosterone was via the classic cytosolic GR, we used the GR blocker RU-38486 with dexamethasone and aldosterone. In the presence of RU-38486, the effect of dexamethasone and aldosterone onsgk1 expression was abolished (Fig. 2 B). Collectively, these data suggest that the lack of GC effect onsgk1 expression, previously reported in HepG2 cells, may reflect the relative paucity of GR in that cell line or the absence of cell-specific co-activators (3).

Fig. 1.

Effect of dexamethasone and aldosterone on sgk1 expression in airway epithelial cells. Human sgk1 (hsgk1) mRNA and 18S rRNA were measured by ribonuclease protection assay (RPA). A andB: dexamethasone (Dex) 100 nM increases sgk1 mRNA in a time-dependent manner in H441 (A) and A549 (B) cells, whereas aldosterone (Aldo) 100 nM has a weak effect seen only in A549 cells. C: dexamethasone increases sgk1 mRNA in a concentration-dependent manner in H441. Y, yeast RNA hybridized with probe and digested with RNAse; last lane, undigested labeled probe. Results are representative of ≥3 experiments.D: dexamethasone dose response in A549 cells. Ratio of sgk1 to 18S rRNA levels quantitated by densitometry and expressed as the degree of (fold) increase from basal levels. Values are data (means ± SE) pooled from 3–4 experiments.

Fig. 2.

Effect of actinomycin D, cycloheximide, and RU-38486 on dexamethasone-stimulated sgk1 expression in A549 cells. Sgk1 mRNA and 18S rRNA were measured by RPA. A: actinomycin D (act) or cycloheximide (chx) was added simultaneously with vehicle (ctrl) or Dex to A549 cells for 2 h. Actinomycin D abolishes Dex-stimulated sgk1 expression. Cycloheximide stimulatessgk1 independently of Dex and enhances Dex-stimulated stimulated sgk1 expression. B: A549 cells were stimulated for 2 h with Dex (100 nM) or Aldo (100 nM) in the presence or absence of the glucocorticoid receptor (GR) blocker RU-38486 (RU, 10 μM). Dex and Aldo induction of sgk1 mRNA occurs via binding to GR.

Because corticosteroids increase Na+ transport in the collecting duct, we wondered whether GCs could stimulatesgk1 expression in human collecting duct epithelia. There are no suitable human collecting duct cell lines with regulated Na+ transport available, and so we looked at expression ofsgk1 in HRCE cells derived from primary cultures of dispersed human kidney cortex epithelia. As in human airway epithelia, GC increased sgk1 expression in these cells in a time-dependent manner (Fig.3 A). Because aldosterone is considered to be a more physiological regulator of Na+transport in the distal nephron, we also looked at the effect of aldosterone in these epithelia. The results demonstrated that aldosterone increased sgk1 expression in HRCE cells, which was evident at the first time point selected (Fig. 3 B).

Fig. 3.

Effect of Dex (A) and Aldo (B) on sgk1 expression in human renal cortical epithelial (HRCE) cells. Human sgk1 (hsgk1) mRNA and 18S rRNA were measured by RPA in all panels.A and B: Dex and Aldo increase sgk1 mRNA in a time-dependent manner in HRCE cells.

To address the mechanisms of transcriptional regulation ofsgk1, we first identified the transcription start site by two complimentary techniques: 5′ RACE and primer extension analysis. Because the 5′ end of the sgk1 transcript was GC rich, we hypothesized that this feature may prematurely terminate RT and give a truncated product by primer extension or 5′ RACE. To identify the authentic 5′ end of the transcript by 5′ RACE, we constructed a cDNA library, which was designed to include the 5′ cap site of mRNAs. RNA from dexamethasone-treated H441 cells was treated with CIP to remove 5′ phosphate from incomplete transcripts that did not include the 5′ cap. RNA was then treated with TAP to remove the 5′ cap, and an adaptor sequence was ligated to decapped RNA. Adaptor-ligated full-length mRNA was reverse transcribed with oligo-(dT) and M-MLV RT and then subjected to two rounds of PCR by use of gene-specific primers. A single band was identified by PCR that on sequence analysis extended the known 5′ end of the sgk1 transcript by 13 bases (Fig.4 A).

Fig. 4.

Mapping of 5′ end of sgk1 mRNA. A: 5′ portion of sgk1 cDNA sequence. A reverse arrow marks primer used for 5′ rapid amplification of cDNA ends (RACE); translation start site is underlined. New sequence identified by 5′ RACE is shown in bold.B: schematic of transcription start site mapping by 5′ RACE and primer extension. Gene-specific reverse primers used in both assays are shown as solid lines, and the extension reaction is shown as dotted arrows. The original exon is shown as an open bar and the new sequence as an open bar with hatched lines. Immediately upstream of the transcription start site, a TATA box is present within genomic sequence. C: primer extension analysis. Primer was 5′ end-labeled and hybridized to H441 RNA (lane 1) or yeast RNA (lane 2). The product of primer extension is shown. A band at ∼80 bp is seen with H441 RNA. The molecular mass markers are a radiolabeled 50-bp DNA ladder (Life Technologies).

We then performed primer extension analysis in H441 mRNA with the nested primer used for 5′ RACE and again identified a single extended product that, by size, matched that seen by 5′ RACE (Fig.4 B). To overcome the possible secondary structure in the 5′ end of this RNA, primer extension was carried out at 60°C with a thermostable form of RT (Thermoscript RT, Invitrogen, Carlsbad, CA). These results thus confirmed the 5′ end of sgk1 mRNA, at least in H441 cells. Compared with genomic DNA sequence, the extended sequence appeared to be within the same exon; furthermore, 23 bp upstream of the 5′ end of this gene, a consensus TATA box, TATAA, was identified.

Having mapped the 5′ end of the sgk1 mRNA, we cloned ∼3 kb of the 5′-flanking region by PCR by use of sequence information available in GenBank. We directionally inserted this upstream of the firefly luciferase reporter gene in pGL3basic (Promega, Madison, WI) and then transiently transfected this into A549 cells along with a plasmid, pRLSV40, to control for transfection efficiency. As a positive control, the GC-responsive α-ENaC promoter-luciferase construct, −1388 +55, was used in some wells (19). Twenty-four hours after transfection, cells were treated with dexamethasone or its vehicle, and lysates were prepared on the following day to measure luciferase activity. Our results demonstrated a sixfold increase in luciferase activity of the sgk1-luciferase construct when exposed to dexamethasone, confirming that 3 kb of genomic sequence 5′ and flanking the sgk1 gene are sufficient to confer GC responsiveness to the luciferase gene (Fig. 5). These results confirm that a GC-responsive enhancer transcriptionally regulates the sgk1 gene.

Fig. 5.

The 5′-flanking region of sgk1 contains a glucocorticoid (GC)-responsive cis-element. A: schematic of the 5′ flanking region of sgk1. Primers F4 and R4 used to amplify ∼3 kb of the 5′-flanking region are shown as arrowheads. The 5′-flanking region is GC rich and contains 3 CpG islands. B:sgk1 promoter coupled to luciferase was transfected into A549 cells and treated with dexamethasone (dex) 100 nM or vehicle (ctrl) for 24 h. The human α-epithelial sodium channel (α-hENaC) promoter is used as a positive control. The activity of the empty plasmid, pGL3basic, was extremely low, although a small increase was noted with dexamethasone (dex). Values are means ± SE (n = 4). *P < 0.001 vs. control (ctrl). These data are representative of 4 separate experiments.

To map the GC-responsive enhancer in the 5′-flanking sequence, we made a series of deletion constructs with convenient restriction sites (Fig.6 A). When tested in A549 cells, deletions up to −1899 maintained the GC effect, whereas a deletion to −735 was enough to abolish the effect, pointing to a GC-responsive element(s) between −1899 and −735 (Fig.6 B). A careful examination of this region demonstrated an imperfect hormone response element, CGGACAAAATGTTCT, present at −1159 to the transcription start site. To test this as a candidate element, we transferred this 15-bp sequence as two tandem copies to the core promoter of α-hENaC-1, −141 +55 α-ENaC/luciferase (19). This construct does not contain the GRE of α-ENaC and hence does not respond to GCs. Consistent with its role as a GC-responsive enhancer, the transferred sgk1 cis-element supported GC stimulation of a heterologous promoter (Fig. 6 C). To demonstrate that the sgk1 cis-element was necessary for the GC effect, a 2-bp substitution was introduced into the 3′ hexameric half-site (CGGACAAAATCATCT) within the −1899 +117 hsgk1sequence. This construct was no longer stimulated by GC when transfected into A549 cells, confirming that the imperfect GRE was necessary and sufficient to confer GC responsiveness to thesgk1 gene (Fig. 6 D).

Fig. 6.

Mapping the GC-responsive cis-element. A: schematic indicates position of restriction sites in 5′-flanking region of sgk1 used to make deletion constructs for transfection assays. An imperfect GC response element (GRE) is present between theSacI and MscI sites. B: full-lengthsgk1 and three deletion variants coupled to luciferase were transfected into A549 cells and treated with dexamethasone (dex) 100 nM or vehicle (ctrl) for 24 h. Results indicate that acis-element between −1899 and −937 is sufficient to confer this effect. Values are means ± SE; n = 3.# P < 0.05; * P < 0.01).C: the GRE in sgk1 was dimerized and ligated to an α-hENaC promoter construct, transfected into A549 cells, and treated with dexamethasone (dex) 100 nM or vehicle as above. The α-hENaC promoter construct (ENaC) and the empty plasmid pGL3basic (pGL3bas) were used as controls. Results demonstrate that the 15-bp GRE sequence is sufficient to confer GC responsiveness when coupled to a heterologous promoter. D: the GRE in sgk1promoter construct was mutated (−1899+117/mutGRE), transfected into A549 cells, and compared with the unmodified sgk1 promoter construct (−1899+117) and with the empty plasmid pGL3bas. Dexamethasone does not stimulate luciferase expression from the mutated construct, confirming that the GRE is necessary for the GC response.

We then asked, by performing gel mobility shift assays, whether GR could bind this GRE. Whole cell nuclear extracts from dexamethasone-treated A549 cells were used with sgk1 GRE or the α-ENaC GRE. Compared with probe alone, increasing the amount of cell extract led to the shift of an increasing fraction of thesgk1 GRE to a higher position, indicating the presence of a DNA-binding protein (Fig. 7 A). This binding was specifically competed by a 50-fold excess of cold sgk1 or α-ENaC GRE but not by a nonspecific (NS) oligonucleotide. When cell extracts were preincubated with an anti-GR antibody, a further shift or supershift of the GRE to a higher level was noted, confirming that GR was part of the complex that binds to the sgk1 GRE. Compared with the sgk1 GRE, there appeared to be enhanced binding of GR to the α-ENaC GRE, as evidenced by a greater intensity or mass of the shifted probe with 1 μg of cell extract and a substantially greater mass of supershifted probe with the anti-GR antibody (Fig. 7 B).

Fig. 7.

Gel mobility shift analysis demonstrates that the GC receptor (GR) binds to the sgk1 GRE. The sgk1 GRE (A) and the α-hENaC GRE (B) were end-labeled and used with A549 nuclear extracts. A: compared with free probe, a broad radiolabeled band is noted (closed arrowhead), representing protein(s) that bind this DNA to retard its mobility. Increased binding is seen with increasing whole cell extract (1–2.5 μg), and this is competitively displaced by a 50-fold excess of nonradioactive sgk1 GRE or α-ENaC GRE, but not by a nonspecific (NS) DNA. When protein DNA complexes are incubated with an anti-GR antibody, a specific band appears at a higher position, consistent with a supershifted GR protein-DNA complex (open arrowhead).B: when an oligonucleotide corresponding to the GRE in α-ENaC is used as the probe with cell extracts (1 μg), a band is seen, similar to that in A, that represents probe with retarded mobility. This is displaced effectively by nonradioactivesgk1 GRE or α-ENaC GRE. When protein DNA complexes are incubated with an anti-GR antibody, a specific and intense band appears at a higher position consistent with a supershifted GR protein-DNA complex.


Corticosteroids are important physiological regulators of transepithelial Na+ transport in several sites, such as the collecting duct of the kidney and throughout the airway epithelia. One of the consequences of corticosteroid action in collecting duct epithelia is the increase in mRNA abundance for sgk1 (6, 19,21). This increase in sgk1 expression precedes the increased sodium transport in these epithelia, and in heterologous expression systems, co-expression of sgk1 with α,β,γ-ENaC enhances Na+ transport (8, 33). In vivo, after aldosterone infusion, cytosolic ENaC subunits are distributed to the apical membrane of the cortical collecting duct, an effect that correlates with increasing sgk1 expression (18). A dominant negative form of sgk1 inhibits ENaC function when heterologously expressed with α-, β-, and γ-ENaC inXenopus oocytes, suggesting that an endogenous sgk1 is required for functional expression of ENaC (33). There is thus good circumstantial evidence that the increase in sgk1 abundance is a prerequisite, at least for the early increase in corticosteroid-mediated Na+ transport.

Because there is increasing evidence that sgk1 may be responsible, at least in part, for the corticosteroid-stimulated increase in Na+ transport in amphibian and rodent kidney collecting duct, we asked whether GC could stimulate sgk1 expression in human epithelia cells. This was not a trivial question, because the human sgk1 transcript was reported to be unresponsive to GCs (34). We examined sgk1 expression in H441 and A549 cells, human lung cell lines where α,β,γ-ENaC subunits and sgk1 are expressed and where Na+ transport occurs via amiloride-sensitive Na+ channels whose biophysical properties are those of an α,β,γ-ENaC heteromultimer (14,16). We show that GCs increased sgk1 expression in a time- and dose-dependent manner and that this effect was via the GR and could be blocked by actinomycin D. These results suggested that the GC effect on sgk1 expression in these cells requires gene transcription.

We also determined that GCs and aldosterone stimulate sgk1expression in HRCE, cells derived from primary culture of renal cortical epithelial cells. The response to aldosterone was less than that seen with GCs. This may reflect the likelihood that these epithelial cells are a mixture of proximal convoluted tubule, distal tubule, and collecting duct cells, in which only the collecting duct cells are expected to be aldosterone responsive. A second possibility that we did not directly test is that these cells in culture no longer express usual levels of the mineralocorticoid receptor as has been observed for many other cell lines derived from the collecting duct (7, 19, 30).

The genomic organization of the sgk1 gene has previously been reported (35). In that study, the 5′ end of thesgk1 gene was mapped by 5′ RACE, and the upstream genomic DNA was cloned and partially sequenced. The authors noted that the 5′-flanking region did not contain a GRE, corroborating their earlier study demonstrating that the human sgk1 gene was not GC regulated in HepG2 cells, a human hepatocellular carcinoma cell line (34). These findings were in contrast to evidence that a GC stimulated rat sgk1 gene expression via an authentic GRE in its 5′-flanking region (37). When we obtained data demonstrating that sgk1 was GC regulated in two human lung epithelia (Fig. 1, A-D), and that this effect was likely to be transcriptional (Fig. 2 A), we decided to reexamine this issue. We wondered whether the earlier study reporting the absence of an authentic GRE in the 5′-flanking region of the human sgk1gene might have overlooked this cis-element. We reasoned that, if the rat sgk1 gene had a GRE in its 5′-flanking region that was the basis of its GC regulation, and if the humansgk1 gene was also GC regulated, then there was a very high likelihood that the human sgk1 gene was also controlled by a GRE. The previous study may have failed to recognize the GRE in the 5′ flanking region of the sgk1 gene for a variety of reasons:1) the transcription start site had not been properly mapped and that the putative 5′-flanking sequences were, in fact, sequences within an upstream intron; 2) the GRE that regulated thesgk1 gene did not bear any resemblance to the ratsgk1 GRE; or 3) the GRE in the human was not in the proximal 5′-flanking region but, rather, elsewhere in the gene.

Because the characterization of the promoter and othercis-regulatory elements of any gene is critically dependent on accurate identification of its transcription start site, we performed our own experiments to map the start site of sgk1 by use of 5′ RACE and primer extension. Our studies identified a single site, 13 bases upstream of that previously identified but still within the same exon (Fig. 4 A). We then cloned the 5′-flanking region upstream of a luciferase reporter gene, and after its transfection into A549 cells, we were able to identify GC-stimulated transcriptional activity. Although the level of induction was modest compared with the induction seen with the α-ENaC promoter (6-fold vs. 44-fold), it correlated with the modest increase in endogenous sgk1 gene seen in A549 cells (Fig. 1 B). By deletional analysis and by testing with a heterologous promoter, we identified an imperfect GRE in the 5′-flanking region that is required for this GC effect in A549 cells. The classic GRE (AGAACAnnnTGTTCT) is a bipartite sequence composed of two hexameric nucleotides that are palindromic and separated by three nucleotides. Interestingly, the humansgk1 GRE is different from the previously reported ratsgk1 GRE by two bases, one of which is within a hexameric half-site. In contrast to the rat sgk1 GRE (AGGACAGAATGTTCT), the human sgk1 GRE differs from the classic GRE also by two nucleotides, suggesting that its affinity for or trans-activation by GR may be different, with resulting effects on the kinetics or amplitude of the GC response in these two species. We were, however, unable to determine a difference in binding affinity for A549 extracts between the human and rat sgk1 GRE in gel mobility shift assays (data not shown).

The sgk1 gene is conserved from yeast to human and is expressed in many tissues in response to a variety of stimuli, including corticosteroids, serum, and inflammatory mediators (5). We have now demonstrated that GC regulation of the human sgk1 gene is mediated via a GRE in its 5′-flanking region. It is interesting that there is a sequence difference within the GRE of the human sgk1 gene compared with the rat. It is difficult to speculate on the direction of change, because both life forms branched off from a common ancestral mammal many millions of years ago. However, an analysis of the mouse genome reveals an imperfect GRE ∼1 kb 5′ and flanking the mouse sgk1 gene, which is identical to the rat GRE. The level of induction by GCs, at least in cultured cell lines, appears to be considerably less in the human compared with rat or even amphibian epithelial cells, suggesting that the more poorly conserved human GRE may have evolved later (6, 20, 37).

In the collecting duct and in airway epithelial cells, stimulation ofsgk1 expression by corticosteroid hormones coincides with an upregulation of Na+ transport. Emerging studies are beginning to provide some insights into how sgk1 may regulate Na+ transport. Activation of sgk1 requires the phosphorylation of residues in the COOH-terminal region of the peptide and is phosphoinositide 3-kinase dependent (12, 15, 24). In turn, sgk1 phosphorylates a number of downstream proteins, including the transcription factor FKHRL1, the cytosolic kinase B-Raf, and Nedd4–2, a WW domain protein that targets membrane ENaC for removal and ubiquitination (4, 9, 27, 38). Phosphorylation of Nedd4–2 at one of three consensus phosphorylation sites reduces its affinity for ENaC subunits, thus providing at least one mechanism for the enhanced expression and function of the channel at the cell surface after GC stimulation.


We thank Tom Schmidt for the gift of BuGR.


  • The sgk1 cDNA sequence reported here has been deposited in GenBank with accession number AF460178. Portions of the work were presented in abstract form at the American Thoracic Society meeting in 2002.

  • This work was supported in part by March of Dimes Birth Defects Foundation Research Grant 6-FY99–444 and by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-54348. C. P. Thomas is an Established Investigator of the American Heart Association.

  • Address for reprint requests and other correspondence: C. P. Thomas, Dept. of Internal Medicine, E300 GH, Univ. of Iowa, 200 Hawkins Drive, Iowa City, IA 52242 (E-mail:christie-thomas{at}uiowa.edu).

  • 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.

  • July 9, 2002;10.1152/ajpendo.00021.2002


View Abstract