Serum- and glucocorticoid-inducible kinase (SGK) 1 and SGK3 share the ability to upregulate several ion channels, including the epithelial Na+ channel. Whereas SGK1 is under genomic control of mineralocorticoids and glucocorticoids, SGK3 is constitutively expressed. The SKG1-knockout (sgk1−/−) mouse is seemingly normal when it is fed a standard diet, but its ability to retain NaCl is impaired when it is fed a salt-deficient diet. In the SGK3-knockout (sgk3−/−) mouse fed standard and salt-deficient diets, hair growth is strikingly delayed but NaCl excretion is normal. Thus the possibility was considered that SGK1 and SGK3 could mutually replace each other, thus preventing severe NaCl loss in sgk1−/− and sgk3−/− mice. We crossed SGK1- and SGK3-knockout mice and compared renal electrolyte excretion of the double mutants (sgk1−/−/sgk3−/−) with that of their wild-type littermates (sgk1+/+/sgk3+/+). Similar to sgk3−/− mice, the sgk1−/−/sgk3−/− mice display delayed hair growth. Blood pressure was slightly, but significantly (P < 0.03), lower in sgk1−/−/sgk3−/− (102 ± 4 mmHg) than in sgk1+/+/sgk3+/+ (114 ± 3 mmHg) mice, a difference that was maintained in mice fed low- and high-salt diets. Plasma aldosterone concentrations were significantly (P < 0.01) higher in sgk1−/−/sgk3−/− than in sgk1+/+sgk3+/+ mice fed control (511 ± 143 vs. 143 ± 32 pg/ml) and low-salt (1,325 ± 199 vs. 362 ± 145 pg/ml) diets. During salt depletion, absolute and fractional excretions of Na+ were significantly (P < 0.01) higher in sgk1−/−/sgk3−/− (1.2 ± 0.2 μmol/24 h g body wt, 0.12 ± 0.03%) than in sgk1+/+/sgk3+/+ (0.4 ± 0.1 μmol/24 h g body wt, 0.04 ± 0.01%) mice. The sgk1−/−/sgk3−/− mice share the delayed hair growth with sgk3−/− mice and the modestly impaired renal salt retention with sgk1−/− mice. Additional lack of the isoform kinase does not substantially compound the phenotype for either property.
- blood pressure
- epithelial sodium channel
serum- and glucocorticoid-inducible kinase (SGK) 1 was originally discovered as a gene genomically upregulated by glucocorticoids (26) and was later found to be a mineralocorticoid-inducible gene (6, 11, 15, 21, 39, 42, 46, 49, 52). SGK1 is expressed in the aldosterone-sensitive distal nephron (4, 4, 39, 39), and coexpression of SGK1 markedly enhances activity of the epithelial Na+ channel (ENaC) heterologously expressed in Xenopus oocytes (3, 10, 15, 37, 42, 53) and A6 cells (24), an effect at least partially due to phosphorylation of the ubiquitin ligase Nedd4-2 (17, 50), which reduces the affinity of the enzyme to the target protein (3, 37, 53). As a result, SGK1 inhibits ubiquitination of the ENaC and, thus, leads to enhanced ENaC protein abundance in the cell membrane. Also, SGK1 has been suggested to stimulate ENaC activity by direct phosphorylation of the carrier protein (18). SGK1 also enhances the activity of other renal transport systems, including the apical renal outer medullary K+ (ROMK) channel (43, 57), the Na+-K+-ATPase (29, 48, 58), the Na+-K+-2Cl− cotransporter NKCC2 (37), the epithelial Ca2+ channel TRPV5 (23, 44), the Na+/H+ exchanger NHE3 (56, 57), and the K+ channel KCNE1 (22).
The stimulatory effect of SGK1 on the ENaC (27), the Na+-K+-ATPase (29, 48, 58), and the epithelial Ca2+ channel TRPV5 (23, 44) in Xenopus oocytes is shared by its isoform SGK3, which has been discovered by homology screening (16, 33), and as “cytokine-independent survival kinase” (38). In contrast to SGK1, SGK3 appears not to be a transcriptional target of glucocorticoids or serum (35). All three kinases are, however, activated by insulin-like growth factor I and insulin through phosphatidylinositol 3-kinase and phosphoinositide-dependent kinase 1 (1, 2, 20, 28, 32, 34, 45). Accordingly, SGK1 is considered to participate in the regulation of renal Na+ excretion by aldosterone, insulin, and insulin-like growth factor I (7–9, 54).
The role of SGK1 in renal Na+ reabsorption is illustrated by the phenotype of the SGK1-knockout (sgk1−/−) mouse, which shows normal Na+ excretion and blood pressure when fed a normal-salt diet but does not show a sufficient decrease in renal Na+ excretion when fed a salt-deficient diet (55). However, the mild impairment of renal Na+ retention in the sgk1−/− mouse differs from the severe salt wasting of mineralocorticoid receptor-knockout mice (5) and the lethality of ENaC-knockout mice (31). Moreover, there was no evidence for renal salt wasting in the SGK3-knockout (sgk3−/−) mouse (40). Thus the following question arises: Does SGK1 could fully replace SGK3 in sgk3−/− mice, and does SGK3 partially replace SGK1 in sgk1−/− mice?
The present study was performed to explore whether animals lacking both SGK1 and SGK3 would be viable or would suffer from severe salt wasting.
All animal experiments were conducted according to the guidelines of the American Physiological Society as well as the German law for the welfare of animals and were approved by local authorities.
Gene-targeted mice deficient in SGK1 (sgk1−/−) (55) and SGK3 (sgk3−/−) (40) were crossed, and the offspring were genotyped by PCR on tail DNA using neomycin resistance specific primers as previously described (40, 55). Mice were reproduced by heterozygous crossing. The genetic background of the animals was a mix of Sv/J129 and C57BL/6.
Experiments were carried out on 12- to 16-mo-old homozygous double (SGK1/SGK3)-knockout animals (sgk1−/−/sgk3−/−, four females and 5 males) and their wild-type littermates (sgk1+/+/sgk3+/+, 4 females and 5 males) and, for comparison, age-matched SGK1-knockout animals (sgk1−/−) and their wild-type littermates (sgk1+/+). The mice were maintained on a standard diet and tap water before the experiment. For evaluation of renal excretion, sgk1−/−/sgk3−/− and sgk1+/+/sgk3+/+ mice were placed individually in metabolic cages over five consecutive days (Tecniplast, Hohenpeissenberg, Germany) for 24-h urine collection (51); they had free access to tap water and a control diet [C1000, 0.24% Na+ (wt/wt); Altromin, Heidenau, Germany] for 2 days followed by a low-salt diet [C1036, 0.015% Na+ (wt/wt); Altromin] for another 3 days. The inner wall of the metabolic cages was siliconized, and urine was collected under water-saturated oil.
The animals were lightly anesthetized with diethylether (Roth, Karlsruhe, Germany), and ∼150 μl of blood were withdrawn into heparinized capillaries by puncture of the retroorbital plexus. Blood losses were replaced with 200 μl of 0.9% NaCl subcutaneously, and the animals were allowed to recover for 2 wk.
Fecal dry weight was obtained after the samples were dried at ∼80°C for ∼3 h. Then the feces was prepared for determination of electrolyte content as follows. After they were dissolved in 5 ml of 0.75 M HNO3, the samples were shaken for 48 h to yield a homogenous creamy mass. The samples were then centrifuged at ∼2,500 g for 10 min, and 1 ml of the supernatant was again centrifuged at ∼10,000 g for 5 min. Aliquots from the second supernatant were stored at −20°C until analysis. Measured electrolyte concentrations were multiplied by 5 to yield the absolute electrolyte content of the feces in micromoles.
Plasma, urinary, and fecal concentrations of Na+ and K+ were measured by flame photometry (model AFM 5051; Eppendorf, Hamburg, Germany), and plasma aldosterone concentrations were quantified with a commercial RIA kit (Demeditec, Kiel, Germany). Plasma and urinary creatinine concentrations were measured using an enzymatic colorimetric method (creatinine PAP; Lehmann, Berlin, Germany).
Systolic arterial blood pressure was determined by the tail-cuff method (model 179; IITC Life Sciences, Woodland Hills, CA) while the animals were fed the control diet and after 2 wk on the low-salt diet, as well as 15 days after exposure to 1% NaCl in their drinking water and the control diet. As reviewed recently (41
Quantitative real-time PCR.
Automated disruption and homogenization of frozen renal tissue from sgk1+/+/sgk3+/+ (n = 5) and sgk1−/−/sgk3−/− (n = 5) mice was performed using the MagNa Lyser (Roche Diagnostics, Mannheim, Germany). Cleared cell lysate was transferred for further RNA purification (RNeasy Mini Kit, Qiagen, Hilden, Germany). Subsequently, 1 μg of total RNA was reverse transcribed to cDNA utilizing the reverse transcription system (Bioscience) with oligo(dT) primers according to the manufacturer's protocol. To determine murine SGK2 (mSGK2) mRNA levels, quantitative real-time PCR was carried out with the LightCycler System (Roche Diagnostics). PCR for mSGK2 were performed in a final volume of 20 μl containing 2 μl of cDNA, 2.4 μl of MgCl2 (3 μM), 1 μl of primer mix (0.5 μM both primers), 2 μl of cDNA Master SybrGreen I mix (Roche Molecular Biochemicals, Mannheim, Germany), and 12.6 μl of diethylpyrocarbonate-treated water. The transcript levels of the housekeeping gene GAPDH were determined for each sample using a commercial primer kit (Search LC, Heidelberg, Germany). PCR for GAPDH were performed in a final volume of 20 μl containing 2 μl of cDNA, 2 μl of primer mix (Search LC), 2 μl of cDNA Master SybrGreen I mix, and 14 μl of diethylpyrocarbonate-treated water. The target DNA was amplified during 35 cycles at 95°C for 10 s, 68°C for 10 s, and 72°C for 16 s, each with a temperature transition rate of 20°C/s and a secondary target temperature of 58°C with a step size of 0.5°C. Melting curve analysis was performed at 95°C at 0 s, 58°C at 10 s, and 95°C at 0 s to determine melting temperatures of primer dimers and the specific PCR products. Melting curve analysis confirmed the amplified products, which were then separated on 1.5% agarose gels to confirm the expected size (413 bp). Finally, results were calculated as a ratio of target gene to housekeeping gene GAPDH transcripts.
The following primers for mSGK2 (GenBank accession no. NM _013731) were used: 5′-CCA CAG ACT TTG ATT TCC TC-3′ (forward) and 3′-GGC AGT CCA AGA GAA TGT T-5′ (reverse). The primers showed no overlap with murine SGK1 (mSGK1, GenBank accession no. NM _011361) and murine SGK3 (mSGK3, GenBank accession no. NM_133220).
Values are means ± SE; n represents the number of independent experiments. All data were tested for significance with two-tailed paired or unpaired Student's t-test and Welch's correction where required using GraphPad InStat version 3.00 for Windows 95 (GraphPad Software, San Diego CA). P < 0.05 was considered statistically significant.
As shown in Fig. 1, hair growth was markedly delayed in sgk1−/−/sgk3−/− mice. The time course of hair growth of the sgk1−/−/sgk3−/− mice was virtually identical to that of the SGK3-knockout mouse described previously (40).
Body weight and food and fluid intake.
After birth, body weight gain was slightly slower in sgk1−/−/sgk3−/− than in sgk1+/+/sgk3+/+ mice, but, after ∼12 wk, both groups attained an almost identical body weight. Thus, at the time of the experiments, body weight was not significantly different between sgk1−/−/sgk3−/− and sgk1+/+/sgk3+/+ mice (Table 1). The low-salt diet did not significantly alter body weight (Table 1). Food intake per body weight was slightly, but significantly, greater in sgk1−/−/sgk3−/− than in sgk1+/+/sgk3+/+ mice fed the control (P < 0.0084) and low-salt (P < 0.0013) diets (Table 1). Food intake tended to increase after the low-salt diet, but the change did not reach statistical significance (P < 0.055 for sgk1+/+/sgk3+/+ and P < 0.057 for sgk1−/−/sgk3−/−). Fluid intake in sgk1−/−/sgk3−/− mice similarly tended to increase after a salt-deficient diet and tended to be greater in sgk1−/−/sgk3−− than in sgk1+/+/sgk3+/+ mice; again, the differences did not reach statistical significance. In sgk1−/− and sgk1+/+ mice fed the low-salt diet, fluid intake was significantly increased (Table 2).
Fecal dry weight and Na+ and K+ content.
Fecal dry weight was significantly higher in sgk1−/−/sgk3−/− mice fed the control diet; this difference disappeared when these animals were fed the low-salt diet (Table 1). As expected, fecal Na+ excretion significantly decreased in sgk1+/+/sgk3+/+ (P < 0.0002) and sgk1−/−/sgk3−/− (P < 0.0001) mice fed the low-salt diet. However, no significant differences were observed in fecal Na+ excretion between sgk1+/+/sgk3+/+ and sgk1−/−/sgk3−/− mice fed either diet. Fecal K+ excretion was similar in sgk1+/+/sgk3+/+ and sgk1−/−/sgk3−/− mice fed either diet (Table 1).
Hematocrit and plasma Na+ and K+ concentrations.
No significant differences were observed in hematocrit between sgk1+/+/sgk3+/+ and sgk1−/−/sgk3−/− mice fed either diet or between the control diet and the low-salt diet in either genotype (Table 1). Similarly, plasma concentrations of Na+ and K+ were not significantly modified by the diet in either genotype and were virtually identical in sgk1+/+/sgk3+/+ and sgk1−/−/sgk3−/− mice fed both diets (Table 1). Plasma K+ concentration tended to increase on initiation of the low-salt diet and tended to be higher in sgk1+/+/sgk3+/+ than in sgk1−/−/sgk3−/− mice (Table 1). None of these apparent differences reached statistical significance.
Plasma aldosterone concentration.
Plasma aldosterone concentration was significantly (P < 0.01) higher in sgk1−/−/sgk3−/− than in sgk1+/+/sgk3+/+ mice fed the standard diet (Fig. 2). Over 7 days, plasma aldosterone concentration markedly increased in sgk1−/−/sgk3−/− mice fed the low-salt diet (P < 0.01). The effect of the low-salt diet on plasma aldosterone concentration did not reach statistical significance (P = 0.17) in sgk1+/+/sgk3+/+ mice. Plasma aldosterone concentration was also higher in sgk1−/− than in sgk1+/+ mice fed the control diet (71 ± 35 vs. 42 ± 12 pg/ml, n = 5 each) and the low-salt diet (4,515 ± 949 vs. 3,395 ± 350 pg/ml, n = 5 each).
No significant differences were observed in creatinine clearance and urinary flow rate between sgk1+/+/sgk3+/+ and sgk1−/−/sgk3−/− mice fed either diet or between the control and the low-salt diet in either genotype (Table 1). Urinary sodium excretion decreased significantly in sgk1+/+/sgk3+/+ (P < 0.0002) and sgk1−/−/sgk3−/− (P < 0.0001) mice fed the low-salt diet. Under control (P < 0.05) and low-salt (P < 0.005) diet conditions, urinary Na+ excretion was significantly higher in sgk1−/−/sgk3−/− than in sgk1+/+/sgk3+/+ mice (Table 1). The fractional Na+ excretion was again significantly decreased by the low-salt diet in sgk1+/+/sgk3+/+ (P < 0.0001) and sgk1−/−/sgk3−/− (P < 0.0001) mice. Fractional Na+ excretion was significantly higher in sgk1−/−/sgk3−/− than in sgk1+/+/sgk3+/+ mice fed either diet. Absolute and fractional urinary K+ excretion tended to be higher in sgk1−/−/sgk3−/− than in sgk1+/+/sgk3+/+ mice, a difference reaching statistical significance when the animals were fed the low-salt diet (P < 0.03). The low-salt diet did not significantly influence absolute or fractional K+ excretion in sgk1+/+/sgk3+/+ mice, whereas it tended to increase K+ excretion in sgk1−/−/sgk3−/− mice (Table 1).
Systolic blood pressure.
Blood pressure was significantly (P < 0.05) lower in sgk1−/−/sgk3−/− than in sgk1+/+/sgk3+/+ mice fed the control diet (103 ± 4 vs. 114 ± 3 mmHg; Fig. 3). Neither the low-salt (2 wk) nor the high-salt diet significantly influenced systemic blood pressure. Blood pressure remained significantly lower in sgk1−/−/sgk3−/− than in sgk1+/+/sgk3+/+ mice fed low-salt (P < 0.05) and high-salt (P < 0.05) diets (Fig. 3). In sgk1−/− mice fed the control diet, blood pressure was similarly lower than in sgk1+/+ mice (106 ± 4 vs. 121 ± 4 mmHg, n = 8 each, P < 0.05).
Expression of SGK2.
Quantitative real-time PCR of liver and kidney did not reveal statistically significant differences between the genotypes in the levels of mRNA encoding SGK2 (ratio of SGK2 mRNA to GAPDH mRNA) in liver or kidney (Fig. 4).
The present observations show that sgk1−/−/sgk3−/− mice are viable and display a phenotype reflecting properties of sgk1−/− and sgk3−/− mice. On the one hand, the mice display the same delay of hair growth described previously for sgk3−/− mice (40). Moreover, similar to sgk3−/− mice (40), they tend to be smaller, despite significantly greater food intake. The discrepancy may be due to moderately impaired intestinal nutrient absorption in the intestine, similar to that shown in sgk3−/− mice (47).
On the other hand, the sgk1−/−/sgk3−/− mice display a moderate impairment of renal Na+ retention previously described in sgk1−/− mice (55). This impairment is evident because of enhanced renal excretion of Na+, particularly in animals fed a low-salt diet, the enhanced plasma aldosterone concentration, and the decreased blood pressure under anesthesia (55). The enhanced plasma aldosterone concentration partially overrides the lack of SGK1 and apparently prevents hypotension in nonanesthetized animals. The moderate salt wasting of the sgk1−/−/sgk3−/− mice is similar to that of sgk1−/− mice but by far less severe than that of mice lacking functional mineralocorticoid receptors (5) or mice lacking a functional ENaC (31). Mice without mineralocorticoid receptors suffer from severe renal salt wasting (5), and ENaC-knockout mice are not viable (31). The mild salt wasting of sgk1−/− mice is consistent with SGK1-independent regulation of renal Na+ reabsorption (55). In theory, the lack of severe salt wasting could have been due to the partial functional compensation by SGK3 in sgk1−/− mice. The present observations clearly demonstrate that this is not the case. Instead, the salt wasting in the sgk1−/−/sgk3−/− mice is not substantially more severe than that in sgk1−/− mice. Thus it appears safe to conclude that SGK3 does not play a major role in renal Na+ reabsorption and that it does not account for the SGK1-independent regulation of the ENaC. The third isoform, SGK2, similarly stimulated the ENaC in vitro (27) and, thus, is a further candidate for the SGK1-independent stimulation of the ENaC. However, SGK2 transcript levels are not upregulated in sgk1−/−/sgk3−/− mice.
Even though the influence of SGK1 on renal salt retention is only moderate, it may well play a significant role in deranged blood pressure regulation. Enhanced SGK1 expression has been observed in the salt-sensitive Dahl rat (25), and moderately enhanced blood pressure is observed in individuals carrying a variant of the SGK1 gene, affecting as many as 5% of unselected Caucasians (12, 12). In the same individuals, increased body mass index (19, 19) and a shortening of the Q-T interval (12, 13) have been observed. The increased body mass index may be partially due to enhanced stimulation of the intestinal glucose transporter SGLT1 (19), the accelerated cardiac repolarization due to enhanced activation of the cardiac K+ channel KCNE1 (14, 23). Thus excessive stimulation of carriers and channels by SGK1 could account for obesity, hypertension, and shortened cardiac action potential (36).
SGK1 has previously been shown to regulate the ROMK1 channel (57). Moreover, SGK1-dependent ENaC activity is expected to depolarize the apical cell membrane of principal cells, thus favoring K+ secretion. Presumably, both effects lead to slightly impaired renal K+ elimination by the sgk1−/− mice (30). Despite enhanced plasma aldosterone concentration, plasma K+ concentration is not lower in the sgk1−/−/sgk3−/− mice. However, neither plasma concentration nor renal excretion of K+ is significantly different between sgk1−/−/sgk3−/− mice and sgk1+/+/sgk3+/+ mice.
The phenotype of a gene-targeted mouse is influenced by the genetic background. We have been unable to breed sgk1−/− mice on a pure C57BL/6 background (unpublished observations). To minimize the bias of variable genetic background, comparisons were always made between littermates.
The observations that hair phenotype is not as pronounced in sgk1−/−/sgk3−/− as in sgk3−/− mice and that salt wasting is not as severe in sgk1−/−/sgk3−/− as in sgk1−/− mice indicate that there is little overlap of SGK1 and SGK3 in these functions. Thus, even though there is considerable overlap of targets (36, 46) and regulation (35), the kinases are apparently not serving identical functions. Additional experiments are needed to elucidate whether the lack of SGK1 and SGK3 has additive effects on other functions. If so, those functions are not required for viability, or they are sufficiently maintained in the absence of SGK1 and SGK3 that the mice do not show gross functional deficits.
This work was supported by grants from Deutsche Forschungsgemeinschaft and the German Ministry for Science and Technology (BMBF) to D. Kuhl and F. Lang.
↵* F. Grahammer and F. Artunc contributed equally to this work.
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.
- Copyright © 2006 the American Physiological Society