Tonicity-dependent induction of Sgk1 expression has a potential role in dehydration-induced natriuresis in rodents (original) (raw)
Like the Sgk1 gene transcript and the Sgk1 protein (40), Sgk1 promoter activity proved to be very responsive to increases in extracellular tonicity in IMCD cells. As shown in Figure 1A, IMCD cells transfected with the Sgk promoter–driven luciferase reporter demonstrated a 5- to 6-fold increase in promoter activity following exposure to an approximately 150-mOsm/kg increase (i.e., above the tonicity of DMEM medium, which is 296 mOsm/kg H2O) in extracellular tonicity. The background vector (pGL3-Luc) showed no sensitivity to extracellular tonicity.
Osmotic induction of rat Sgk1 promoter in rat IMCD cells. (A) IMCD cells were cotransfected with 1 μg of –1430 rat _Sgk1_-Luc (wild-type) and 0.2 μg actin–β-galactosidase. Twenty-four hours later, they were treated with 75 mM NaCl or 150 mM sucrose (Suc) for 24 hours. Cells were harvested, and luciferase levels were measured and normalized to β-galactosidase activity. C, control. (B) p38 MAPK inhibitor SB203580 (SB) and dominant negative MKK6 (dnMKK6) inhibit osmotic induction of the Sgk1 gene promoter in IMCD cells. The –1430 rat Sgk1_-Luc reporter and actin–β-galactosidase were transfected into IMCD cells in the presence or absence of 0.2 μg of a dominant negative mutant of MKK6. Twenty-four hours later, cells were incubated with 75 mM NaCl for 24 hours in the presence of different concentrations (M) of SB203580 (1 hour preincubation) as indicated. The pooled data are from 3–4 independent experiments. **P < 0.01 versus control; †_P < 0.05 versus NaCl alone.
The osmotic activation of this promoter was dependent on the p38 MAPK signaling system (Figure 1B). Treatment of the IMCD cells with the p38 MAPK inhibitor SB203580 led to a dose-dependent suppression of the osmotic induction of _Sgk1_-Luc reporter activity. Similarly, cotransfection with a dominant negative mutant of MKK6, an activating kinase immediately upstream from p38 MAPK (41), resulted in a nearly 50% decrease in the tonicity-dependent induction.
Earlier studies (34) indicated that osmotic induction of the Sgk1 promoter in NMuMg cells is dependent on a GC-rich region, shown in vitro to associate with the transcription factor Sp1, in close proximity to the transcription start site of the Sgk1 gene. Since Sp1-binding sites have not been linked to induction of osmotically sensitive genes in renal cells, we examined the promoter sequence for the presence of alternative sites that might contribute to the osmotic induction. We identified a consensus tonicity-responsive enhancer (TonE) (44) approximately 312 bp upstream from the transcription start site. The rat Sgk1 TonE (TGGAAAATCACC) (Figure 2A) is completely homologous with the murine and human aldose reductase osmoregulatory elements (13, 45). We introduced a series of point mutations into the putative TonE in the Sgk1 promoter–driven luciferase reporter and transfected these into IMCD cells. As shown in Figure 2B, one of the mutations (M3) led to complete reversal of the osmotic induction, while 2 of the others, M1 and M2, led to intermediate levels of inhibition; a fourth, M4, with base modifications outside the core TonE, was without effect.
Mutation of TonE blocks osmotic induction of the Sgk1 gene promoter in IMCD cells. (A) Location and site-directed mutagenesis of the TonE site in the rat Sgk1 promoter. Mutagenized bases are indicated by lower-case letters. (B) Wild-type –1,430 Sgk1_-Luc (1 μg ) or a series of TonE mutants (M1–M4) were cotransfected with 0.2 μg actin–β-galactosidase into IMCD cells. The cells were cultured with or without 75 mM NaCl for 24 hours prior to collection for luciferase and β-galactosidase measurement. **P < 0.01, *P < 0.05 versus respective controls; †_P < 0.01 versus NaCl alone (n = 3).
Given the previous association of the putative Sp1-binding element (34) with the osmotic induction of this promoter’s activity, we created a mutation in this site, alone and in combination with the TonE mutation (Figure 3A), and introduced the reporters into IMCD cells. As shown in Figure 3B, virtually the entire osmotic induction appears to flow through TonE in these cells. Mutation of the Sp1-binding element in the GC-rich region did not affect the osmotic induction when tested alone or in the presence of the TonE mutation. Thus, while this Sp1-binding element appears to play a role in controlling the Sgk1 promoter response to tonicity in NMuMg cells (34), in IMCD cells the TonE site dominates in orchestrating this response.
Proximal Sp1-binding elements do not participate in the osmotic induction of Sgk1 promoter activity in rat IMCD cells. (A) Locations and targeted mutagenesis of the TonE- and Sp1-binding sites in the rat Sgk1 promoter. TonE- and Sp1-binding sequences are identified in bold, and mutated bases are indicated by lower-case letters. (B) Actin–β-galactosidase (0.2 μg) was cotransfected into IMCD cells with 1 μg wild-type –1430 _Sgk1_-Luc or the M3 TonE mutant (TonE MUT), Sp1 mutant (Sp1 MUT), or a double mutant (TonE MUT/Sp1 MUT) and cultured for 24 hours. At that point, cells were treated with 75 mM NaCl or 150 mM sucrose for 24 hours. **P < 0.01 versus control (n = 3).
In the case of the aldose reductase (13) and betaine transporter (20) genes, a specific nuclear transcription factor termed NFAT5 has been shown to associate with TonE and stimulate transcriptional activity of the contiguous promoter. To explore the involvement of NFAT5 as a mediator of the osmotic induction of the Sgk1 promoter, we carried out EMSA of nuclear extracts from IMCD cells cultured under isotonic versus hypertonic (NaCl) conditions using radiolabeled Sgk1 TonE sequence as a probe. As shown in Figure 4A, exposure to the hypertonic environment resulted in an increase in protein association with the oligonucleotide harboring the TonE, here designated as the NFAT5 complex. Incubation with anti-NFAT5 antibody, but not with antibody directed against the p50 subunit of NF-κB, resulted in a supershift of the associated protein. The NFAT5 complex was competed by unlabeled oligonucleotide harboring the wild-type sequence and by the sequence harboring the M4 mutation, but not by oligonucleotides harboring mutations M1, M2, or M3 (Figure 4B). Thus, the relative affinity of NFAT5 for these mutated osmotic response element (ORE) sequences mirrors the ability of these sequences to signal the osmotic induction of the Sgk1 promoter (see Figure 2).
NFAT5 binding to TonE in the Sgk1 promoter in vitro. (A) Nuclear extracts were preincubated on ice for 2 hours with 1 μg of polyclonal antibody directed against NFAT5 or the p50 subunit of NF-κB. 32P-end-labeled, double-stranded wild-type oligonucleotide containing the TonE sequence from the Sgk1 promoter (32P-Sgk1) was added to the reaction and incubated at room temperature for 30 minutes. (B) Nuclear extracts were incubated with 32P-labeled TonE-containing Sgk1 oligonucleotide in the absence or presence of increasing concentrations of unlabeled wild-type or mutant TonE-containing oligonucleotides (10- or 100-fold molar excess). The reaction complexes were resolved on 5% nondenaturing polyacrylamide gels. (C and D) Cytoplasmic extract was collected simultaneously with the nuclear extract, and 150 mM NaCl and 10% glycerol were added. Cytoplasmic and nuclear extracts were subjected to Western blot analysis with anti-NFAT5 antibody. NFAT5 signals were normalized to GAPDH and vitamin D receptor (VDR), respectively. (E) IMCD cells were lysed with regular lysis buffer, and total protein was used for immunoblotting. The NFAT5 signal was normalized to VDR. The experiments were repeated 2–4 times. Representative autoradiographs are shown.
Western blot analysis for NFAT5 protein showed that there was a shift of protein from the cytoplasm into the nuclear compartment after 4 hours of exposure to the hypertonic environment (75 mM NaCl), and this nuclear sequestration persisted for 24 hours (Figure 4, C and D). In addition, there was a net increase in levels of total NFAT5 protein at 24 hours that was not present at the 4-hour time point (Figure 4E).
To confirm that the osmotically dependent NFAT5 association with the _Sgk_1 promoter takes place within the context of the intact cell, we carried out ChIP analysis of the endogenous rat Sgk1 promoter using an antibody directed against NFAT5. As shown in Figure 5A, exposure of IMCD cells to hypertonic medium (addition of 75 mM NaCl or 150 mM sucrose to culture medium) led to increased association of both NFAT5 and RNA polymerase II with the native Sgk1 gene relative to the isotonic control. Neither NFAT5 nor RNA polymerase II interacted to any significant degree with sequence lacking the TonE positioned 3.2 kb upstream from the Sgk1 gene. To prove that this interaction between NFAT5 and the Sgk1 gene promoter requires the TonE-binding sequence described above, we introduced reporter plasmids harboring 1,474 bp of rat Sgk1 5′ flanking sequence (–1,430 to +44) into IMCD cells by transient transfection. One of these reporters contained entirely wild-type sequence, while the other harbored a selective mutation (M3) in the TonE. As shown in Figure 5B, the wild-type sequence showed evidence of substantial osmosensitive association with NFAT5 in the ChIP assay; mutation of the TonE-binding site resulted in near complete loss of this association.
ChIP analysis of the Sgk1 promoter. (A) NaCl and/or sucrose treatment increases Sgk1 promoter activity. IMCD cells were exposed to 75 mM NaCl or 150 mM sucrose for 24 hours. Extracts were generated, cross-linked with formaldehyde, fragmented by sonication, and immunoprecipitated with antibody directed against NFAT5 or RNA polymerase II (Pol II) or with nonimmune IgG. Associated DNA fragments were amplified for Sgk1 promoter sequence or unrelated promoter sequence positioned 3.2 kb upstream from the Sgk1 gene (top panel). Products were then size-fractionated on a 1% agarose gel. Signal generated from input chromatin (i.e., without immunoprecipitation) is shown in the bottom panel. The lanes were run on the same gel but were noncontiguous. (B) TonE mutation dramatically reduces osmotically induced NFAT5 binding to the Sgk1 promoter in IMCD cells. The experiment was carried out using transfected wild-type Sgk1 promoter or the M3 promoter mutant as described in Methods. Twenty-four hours after transfection, cells were challenged with NaCl or sucrose for 24 hours and then processed as described above. Representative results are shown (n = 3).
To link the NFAT5 protein functionally to the osmotic induction of Sgk1 promoter activity, we used 2 different approaches to inhibit NFAT5 activity. Introduction of U6-N5 ex8, a small inhibitory RNA vector that specifically targets the NFAT5 endogenous transcript (46), completely reversed the osmotic induction of the Sgk1 promoter, while the empty U6 vector was devoid of activity (Figure 6). Similarly, introduction of a dominant negative NFAT5 expression vector (NFAT5-DN) along with the _Sgk1_-Luc reporter led to significant inhibition of the osmotic response. Collectively, these findings, along with the results presented above, support the hypothesis that NFAT5 is the relevant transcription factor that binds to the Sgk1 TonE and mediates the osmotic induction of that gene’s promoter in IMCD cells.
Selective inhibition of NFAT5 leads to blockade of the hypertonic induction of the Sgk1 gene promoter. A scrambled shRNA as a negative control (U6-N), an shRNA construct that targets NFAT5 (U6-N5 ex8), and a dominant negative mutant (NFAT5-DN) directed against NFAT5 inhibit the osmotic induction of the cotransfected Sgk1 promoter. The experiment is described in Methods. **P < 0.01 versus without NaCl (n = 3).
A previous study by Rozansky et al. (47) demonstrated that sgk1 gene expression can also be activated by reductions in extracellular tonicity (i.e. hypoosmolality) in A6 cells. To evaluate this phenomenon in cultured IMCD cells and its dependence on the TonE site described above, we introduced the wild-type _Sgk1_-Luc or the TonE-mutated _Sgk1_-Luc into primary cultures of rat IMCD cells and exposed them to either a hyper- or a hypotonic environment. As shown in Figure 7, relative to the cells cultured in isotonic medium, cells exposed to either the hypo- or hypertonic medium showed a significant increase in expression of the wild-type Sgk1 promoter. However, while mutation of the TonE, as expected, led to a near complete inhibition of the hypertonic stimulation of promoter activity, it had virtually no effect on the hypotonic stimulation. This implies that the hypotonic stimulation of Sgk1 gene transcription operates over signaling pathways that are largely independent of those used for the hypertonic induction, including NFAT5 and the TonE.
Hypotonic induction of the Sgk1 promoter does not require TonE. Wild-type –1,430 _Sgk1_-Luc (1 μg) or M3-_Sgk1_-Luc (1 μg) together with actin–β-galactosidase (0.2 μg) were transfected into IMCD cells. After 24 hours, cells were exposed to hypotonic medium (Hypo; 220 mOsm/kg water) or hypertonic medium (Hyper; 75 mM NaCl; final osmolality, 446 mOsm/kg water) for 24 hours. Pooled data (n = 3) representing normalized _Sgk1_-Luc activity are shown. **P < 0.01 versus control.
As mentioned above, expression of both Sgk1 and NPR-A has been shown to increase in IMCD cells exposed to increased extracellular tonicity. To establish the link between Sgk1 and the induction of NPR-A gene expression, we employed an siRNA approach. Cultured rat IMCD cells were transfected with one of 3 siRNA sequences, each of which was designed to target the rat Sgk1 gene coding sequence, prior to exposure of the cells to increased extracellular tonicity (75 mM NaCl). As shown in Figure 8A, exposure to hypertonic culture media resulted in a 4- to 5-fold increase in Sgk1 mRNA expression in these cells. This increment was partially reduced by siSgk1A and siSgk1B and nearly completely eliminated by cotransfection with siSgk1C. Similarly, as shown in the Western blot in Figure 8B, siSgk1C had little effect on basal levels of Sgk1 protein, but it nearly completely reversed the NaCl-dependent increment in Sgk1. The same level of extracellular tonicity (75 mM NaCl) led to a 3-fold increase in NPR-A mRNA levels (Figure 8C) and a 4-fold increment in NPR-A protein levels (Figure 8D) in IMCD cells. Transfection of these cells with siSGK1C inhibited the increase in NPR-A protein and mRNA by approximately 60%. Collectively, these data indicate that the osmotic induction of Sgk1 is a major contributor to the osmotic stimulation of NPR-A gene expression in IMCD cells.
Sgk1 siRNA blocks NaCl-induced Sgk1 and NPR-A gene transcription and translation. Three different Sgk1 siRNAs (siSgk1A, siSgk1B, siSgk1C) (A) and a negative control siRNA (C) were individually transfected into IMCD cells for 48 hours. Three hours prior to isolation of RNA, cells were treated with or without 75 mM NaCl. Total RNA was obtained for measurement of Sgk1/Gapdh mRNA by real-time PCR. (B) IMCD cells were transfected with control siRNA and siSgk1C for 41 hours and then exposed to control medium and 75 mM NaCl for 7 hours. Afterward, cells were harvested, and total protein was assayed for Sgk1 and GAPDH expression by Western blot analysis. Sgk1 expression was normalized to GAPDH protein levels. (C) siSgk1C or control siRNA was individually transfected into IMCD cells for 28 hours. Subsequently, cells were incubated with or without 75 mM NaCl for 20 hours. Total RNA was collected, and NPR-A/Gapdh mRNA levels were measured by real-time PCR. (D) In a separate experiment, after 24 hours of transfection with siSgk1C or control siRNA, cells were treated with or without 75 mM NaCl for 24 hours. Membrane proteins were prepared as described in Methods and used to assay NPR-A/β-tubulin expression by Western blot analysis. All experiments were repeated 3–4 times. Representative immunoblotting results are shown. **P < 0.01, *P < 0.05 versus control; †P < 0.01 versus NaCl alone.
To extend this observation to an in vivo model, we subjected male Sprague-Dawley rats to 24 hours of water deprivation, a procedure that is known to increase urine osmolality (4, 8, 10, 26, 43) and renal medullary sodium concentration (4, 42, 43). As shown in Table 1, water deprivation resulted in a significant increase in urine osmolality (~40% increase over the baseline urine osmolality). There was a modest, but not statistically significant, increase in plasma osmolality, and no significant change in systolic or diastolic blood pressure in these 2 groups. The water-restricted animals lost approximately 10% of their body weight over the 24-hour period, while weights in the control animals were stable.
Effect of 24-hour water restriction on plasma aldosterone levels, plasma and urine osmolality, blood pressure, and body weight (before and after dehydration)
Water restriction was associated with a statistically significant increase in urinary sodium excretion (Figure 9A), confirming the presence of the dehydration natriuresis observed by others in rats, rabbits, sheep, mice, dogs, and humans (1–11) following short-term water restriction. This was accompanied by a significant increase in Sgk1 mRNA (Figure 9B) and Sgk1 protein (Figure 9C) in the renal IMCD. Because Sgk1 gene expression is known to be regulated by aldosterone, we measured plasma aldosterone levels at the end of the water deprivation period. As reported in Table 1, there was no change in plasma aldosterone levels after 24-hour water deprivation, a finding that is consistent with previous dehydration studies (3, 9, 11). Since some studies (48, 49) have shown that increased Sgk1 expression is linked to increased epithelial sodium channel (ENaC) expression, we asked whether dehydration-induced Sgk1 expression was associated with increased ENaC expression. As shown in Figure 9D, water restriction modestly increased expression of the α subunit of ENaC (αENaC). However, this modest elevation in αENaC expression was accompanied by increased sodium excretion rather than sodium absorption (see Figure 9A), implying that Sgk1-mediated activation of natriuresis plays a dominant physiological role in this special circumstance.
Water restriction increases urine sodium excretion, Sgk1 mRNA, and protein levels in rat IMCD. (A) Urine from control rats or rats dehydrated (Dehy) for 24 hours was collected twice (9 and 24 hours into dehydration), and urine from the final 15 hours was used to measure volume, sodium concentration, and urine Na excretion rate (UNa V). Pooled data (n = 10) are shown. (B) Rats were euthanized, and renal medullas were isolated for RNA and protein preparation. Real-time PCR was performed for analysis of Sgk1 and Gapdh mRNA levels. Western blot analysis was used to measure Sgk1 and GAPDH protein levels (C) and αENaC and GAPDH protein levels (D). Representative results and pooled data (n = 6) are presented. **P < 0.01, *P < 0.05 versus control.
The increase in Sgk1 expression was accompanied by an increase in NPR-A mRNA (Figure 10A) and protein (Figure 10B), which, in turn, were linked to a 3-fold increase in urinary cyclic GMP excretion (Figure 10C). Cyclic GMP is the biological end product of NPR-A activation (i.e., ligand-dependent guanylyl cyclase activation).
Dehydration activates the natriuretic peptide system in rat IMCD. (A) Renal medullas from control rats or rats subjected to 24 hours of dehydration were used to prepare total RNA and membrane protein. NPR-A and Gapdh mRNAs were measured by real-time PCR. Pooled data (n = 8) are shown. (B) NPR-A and β-tubulin in membrane protein fractions were determined by immunoblot. NPR-A levels are normalized for β-tubulin expression. Representative results and pooled data (n = 10) are shown. (C) Urine of control rats or water-restricted rats from the final 15 hours of dehydration was used to measure urinary cGMP excretion rates (UcGMP V). Pooled data (n = 10) are shown. (D) After 24 hours of dehydration, the left atria were isolated for preparation of total RNA. ANP and Gapdh mRNA levels were measured by real-time PCR. Pooled data (n = 8) are presented. (E) Plasma was used to determine ANP levels. The data are derived from 8 animals in each group. (F) Renal medullas were used to measure urodilatin and Gapdh mRNA by real-time PCR. Pooled data (n = 8) are shown. (G) Urine was used to measure urodilatin excretion rates (Urodilatin V) in 8 animals in each group. **P < 0.01, *P < 0.05 versus control.
To investigate a possible role for the ligand itself (i.e., ANP) in contributing to this increase in urinary cyclic GMP, we measured ANP mRNA transcript levels in the atria of the control versus water-restricted rats, as well as plasma ANP levels in the 2 groups. As shown in Figure 10, D and E, water restriction reduced, rather than increased, expression of the atrial ANP gene and circulating levels of ANP.
The kidney itself is also thought to be a source of ANP gene expression. Cells of the distal nephron have been shown to harbor proANP gene transcripts and ANP immunoreactivity (50), and ANP immunoreactivity has been identified in the urine (51). Urinary ANP, termed urodilatin, is a product of the renal ANP gene with a 4-amino-acid N-terminal extension (supplied from the proANP precursor) linked to the core ANP peptide (52). As shown in Figure 10F, expression of proANP mRNA in the inner medulla is increased following 24 hours of water restriction, as is the excretion of urodilatin in urine (Figure 10G) (urodilatin concentration in urine: control, 4.37 ± 0.64 pg/μl vs. dehydrated, 9.33 ± 1.01 pg/μl; P < 0.01), suggesting that locally generated ligand as well as increased expression of the NPR-A receptor (see above) contribute to the increased urinary excretion of sodium in the acutely dehydrated state.
To confirm the mechanistic link between dehydration natriuresis and increased NPR-A expression, we examined the effect of 24-hour water restriction in wild-type and Npr1–/– (i.e., NPR-A gene–deleted) mice. As noted previously (53), Npr1–/– mice display significant hypertension compared with control littermates (systolic BP, 112.8 ± 10.6 vs. 103.8 ± 6.9 mmHg, P < 0.01; diastolic BP, 96.8 ± 12.5 vs. 83.3 ± 8.1 mmHg, P < 0.05). Twenty-four hours of water deprivation resulted in weight loss (12% in wild-type and 13% in Npr1–/– mice) and increased urine osmolality in wild-type and Npr1–/– mice compared with their respective control littermates with free access to water (wild-type, 3,123.4 ± 606.4 vs. 1,817.9 ± 401.5 mOsm/kg H2O, P < 0.01; Npr1–/–, 2,415.6 ± 549.1 vs. 1,568.8 ± 578.5 mOsm/kg H2O, P < 0.01). This was accompanied by a significant increase in Sgk1 protein expression in IMCD (Figure 11A), suggesting that hyperosmolality activates Sgk1 expression in this nephron segment in both wild-type and Npr1–/– mice. Urinary sodium excretion was markedly increased during 24 hours of water restriction in wild-type mice (Figure 11B); however, this natriuresis did not occur in Npr1–/– mice (Figure 11B). As expected, dehydration dramatically increased NPR-A expression in wild-type mice; there was no NPR-A protein expression in Npr1–/– mice (Figure 11C). NPR-A activity, assessed as urinary cyclic GMP excretion, was significantly increased in dehydrated wild-type but not in dehydrated Npr1–/– mice (Figure 11D). These findings support a mechanistic link between dehydration-induced natriuresis and NPR-A signaling activity.
Effect of water restriction on the Sgk1/NPR-A signaling pathway in wild-type and Npr1–/– mice. (A) Control and 24-hour water-restricted mice were euthanized, and IMCD was isolated for total protein preparation. Sgk1 and GAPDH protein levels were measured by Western blot analysis. Representative blot and pooled data (n = 5) are shown. (B) Urine from control mice and mice dehydrated for 24 hours was collected in the final 15 hours. Urine volume and sodium were measured (n = 9). (C) Isolated renal inner medulla was used to prepare membrane protein. NPR-A expression was determined by Western blot analysis and normalized to β-tubulin content. Representative results and pooled data (n = 4) are presented. (D) Urinary cGMP excretion was measured during the final 15 hours of water restriction (n = 9). **P < 0.01 versus the relevant control.