Excess salt intake decreases peripheral glucose uptake, thus impairing glucose tolerance. Stimulation of cellular glucose uptake involves phosphatidylinositide-3-kinase (PI-3K)–dependent activation of protein kinase B/Akt. A further kinase downstream of PI-3K is serum- and glucocorticoid-inducible kinase (SGK)1, which is upregulated by mineralocorticoids and, thus, downregulated by salt intake. To explore the role of SGK1 in salt-dependent glucose uptake, SGK1 knockout mice (sgk1−/−) and their wild-type littermates (sgk1+/+) were allowed free access to either tap water (control) or 1% saline (high salt). According to Western blotting, high salt decreased and deoxycorticosterone acetate (DOCA; 35 mg/kg body wt) increased SGK1 protein abundance in skeletal muscle and fat tissue of sgk1+/+ mice. Intraperitoneal injection of glucose (3 g/kg body wt) into sgk1+/+ mice transiently increased plasma glucose concentration approaching significantly higher values ([glucose]p,max) in high salt (281 ± 39 mg/dl) than in control (164 ± 23 mg/dl) animals. DOCA did not significantly modify [glucose]p,max in control sgk1+/+ mice but significantly decreased [glucose]p,max in high-salt sgk1+/+ mice, an effect reversed by spironolactone (50 mg/kg body wt). [Glucose]p,max was in sgk1−/− mice insensitive to high salt and significantly higher than in control sgk1+/+ mice. Uptake of 2-deoxy-d-[1,2-3H]glucose into skeletal muscle and fat tissue was significantly smaller in sgk1−/− mice than in sgk1+/+ mice and decreased by high salt in sgk1+/+ mice. Transfection of HEK-293 cells with active S422DSGK1, but not inactive K127NSGK, stimulated phloretin-sensitive glucose uptake. In conclusion, high salt decreases SGK1-dependent cellular glucose uptake. SGK1 thus participates in the link between salt intake and glucose tolerance.
Excessive salt intake may impede cellular glucose uptake in peripheral tissues and, thus, lead to delayed decrease of plasma glucose concentrations following a glucose load (1–5). Mechanisms accounting for the salt sensitivity of peripheral glucose uptake have remained ill defined. Glucose transport into insulin-responsive cells is accomplished in large part by insertion of the Na+-independent glucose carriers, GLUT1 (SLC2A1) and GLUT4 (SLC2A4), into the cell membrane (6–8). The stimulation of the transporters by insulin requires phosphatidylinositide-3-kinase (PI-3K) and is abrogated by pharmacological inhibitors (wortmannin and LY294002) or genetic knockout (PI-3K dominant-negative mutants) of the kinase (9–14). Downstream elements of PI-3K include the phosphoinositide-dependent kinase PDK-1, which in turn phosphorylates and thus activates the serine/threonine kinase Akt/protein kinase B (PKB) (15–17). The effect of PI-3K on GLUT4 trafficking is mediated by PKB (18,19). PKB is, however, at least in some cells, not required for the PI-3K–dependent trafficking of GLUT1 (18). Thus, some other PI-3K–dependent protein kinase is presumably involved in the regulation of GLUT1.
A further downstream molecule in the PI-3K signaling cascade is the serum- and glucocorticoid-inducible kinase (SGK)1 (20–22), which has previously been shown to stimulate several transport proteins (23), including the Na+-coupled glucose transporter SGLT1 (24). SGK1 was originally cloned as a glucocorticoid-sensitive gene from rat mammary tumor cells (25,26) and later as a human cell volume–regulated gene (27). SGK1 is strongly upregulated by mineralocorticoids (28–36) and participates in the stimulation of the renal epithelial Na+ channel ENaC by mineralocorticoids (30,33–38) and insulin (39–42).
The present study has been performed to explore whether SGK1 is involved in the link between salt intake and glucose tolerance. To this end, experiments were performed in wild-type animals, in gene-targeted mice lacking SGK1 (sgk1−/−), and in their wild-type littermates (sgk1+/+). Glucose tolerance and peripheral glucose uptake have been determined in animals with or without excessive salt intake. The results demonstrate that the effect of excessive salt intake on peripheral glucose uptake and alterations of plasma glucose concentration are dependent on, and mediated by, SGK1.
RESEARCH DESIGN AND METHODS
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.
Mice deficient in SGK1 (sgk1−/−) were generated and bred as previously described (43). In brief, a conditional targeting vector was generated from a 7-kb fragment encompassing the entire transcribed region on 12 exons. The neomycin resistance cassette was flanked by two loxP sites and inserted into intron 11. Exons 4–11, which code for the sgk1 kinase domain, were “floxed” by inserting a third loxP site into intron 3. A clone with a recombination between the first and third loxP site (type I recombination) was injected into C57BL/6 blastocytes. Male chimeras were bred to C57BL/6 and 129/SvJ females. Heterozygous SGK1-deficient mice were backcrossed to 129/SvJ wild-type mice (Charles River, Sulzfeld, Germany) for 10 generations and then intercrossed to generate homozygous SGK1 knockout mice (sgk1−/−) and their wild-type littermates (sgk1+/+). The genetic background of the animals was thus 129/SvJ. For the study, 4-month-old male SGK1 knockout (sgk1−/−) mice and their wild-type littermates (sgk1+/+) were selected. Where indicated, female SGK1 knockout (sgk1−/−) mice and their wild-type littermates (sgk1+/+) and unrelated (not littermate) wild-type male 129/SvJ mice were studied.
For determination of glucose tolerance, mice were starved overnight and glucose (3 g/kg body wt i.p.) was injected. Then, a drop of blood was drawn from the tail into a test strip of a glucometer (Accutrend; Roche, Mannheim, Germany) for measuring the blood glucose levels before and 15, 30, 45, 60, 75, 90, 120, 150, or 180 min after the injection.
In a further series of experiments, long-acting insulin (Novo Nordisk, Mainz, Germany) was injected (0.15 units/kg body wt i.p.) and plasma glucose concentrations were determined at 0, 15, 30, 45, 60, and 90 min as described above.
In vivo tissue glucose uptake during a glucose tolerance test was determined in male mice that have been starved for 16 h. 2-deoxy-d-[1,2-3H]glucose (2-DOG) was mixed with 20% of regular glucose (10 μCi/mouse, 3 g/kg body wt) and injected intraperitoneally. Blood glucose levels were determined to ensure adequate injection. After 120 min, the mice were killed and 100 mg of tissues were homogenized in 1 ml of water. Seven-percent ice-cold perchloric acid (800 μl) was added to homogenate (800 μl). The sample was then cleared by centrifugation, and 1 ml of the supernatant was neutralized for 30 min with 2.2 mol/l KHCO3. The precipitate was removed by centrifugation, and 500 μl of the supernatant was used to determine total 3H radioactivity.
To explore whether salt loading decreases SGK1 protein abundance, sgk1+/+ mice were divided into four groups (five mice per group) and treated separately with 1) control diet for 2 weeks, 2) control diet for 2 weeks plus 1% NaCl in drinking water for 2 weeks, 3) control diet for 2 weeks plus deoxycorticosterone acetate (DOCA; 35 mg/kg body wt s.c.) treatment for 2 days, and 4) control diet for 2 weeks plus DOCA (35 mg/kg body wt s.c.) and spironolactone (50 mg/kg body wt s.c.) treatment for 2 days. Mice were anesthetized with ketamine (60 mg/kg body wt i.p.) plus xylazine (10 mg/kg body wt i.p.). Skeletal muscle and fat tissues were removed and immediately shock frozen in liquid nitrogen. For Western blot analysis, the tissues were homogenized by using a homogenizer (Labortechnik, Mullheim, Germany). The muscle tissues were homogenized in lysis buffer containing 50 mmol/l Tris-HCl, 50 mmol/l NaF, pH 7, 50 mmol/l β-glycerophosphate, 10 mmol/l potassium phosphate, 10 mmol/l MgCl2, 1 mmol/l EDTA, 1 mmol/l EGTA, 1 mmol/l sodium vanadate, 0.1% Tween 20, and a protease inhibitor cocktail tablet (Complete mini EDTA-free; Roche). The fat tissues were homogenized in lysis buffer containing 20 mmol/l Tris, pH 7.4, 2 mmol/l EDTA, 137 mmol/l sodium chloride, 1% NP-40, 10% glycerol, 12 mmol/l β-glycerol phosphate, and a protease inhibitor cocktail tablet (Complete mini EDTA-free; Roche). The homogenates were centrifuged at 10,000g, 4°C for 30 min, and the supernatant was removed and used for Western blotting. Total proteins were measured using the Bradford assay. Total proteins (100 μg) were separated by SDS-PAGE (10% SDS Tris-Glycine), transferred to nitrocellulose membranes, blocked overnight in blocking buffer (5% fat-free milk in PBS containing 0.1% Tween) at 4°C, and incubated 1 h with a polyclonal anti-SGK1 antibody (46) diluted 1:1,000 in blocking buffer. The antibody was kindly provided by Nicola Perrotti, Catanzaro, Italy. After incubation with a horseradish peroxidase–conjugated anti-rabbit secondary antibody (diluted 1:1,000 in blocking buffer; Amersham, Freiburg, Germany), the bands were visualized with enhanced chemoluminescence (Amersham) according to the manufacturer’s instructions. Homogenates were also probed with a primary glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Santa Cruz, Heidelberg, Germany) antibody as a loading control. Densitometric analysis of SGK1 protein abundance was performed by using Quantity One software (1998 version; Bio-Rad Laboratories, Hercules, CA) and normalized using GAPDH.
To further explore whether SGK1 increases glucose transport, HEK-293 cells were seeded on six-well plates at 0.2 × 106 cells/well, and 24 h later, cells were transfected with 2 μg constitutively active S422DSGK1, inactive K127NSGK1, or empty vector (as a control) by using Lipofectamine (Invitrogen, Karlsruhe, Germany) according to the manufacturer′s instructions. For determination of transport, 2-DOG was used as the glucose analog. 2-DOG uptake was measured in HEK-293 cells 2 days after transfection by incubating the cells at 37°C for 5 min (linear range of uptake) in glucose-free Krebs Ringer HEPES buffer containing 0.3 mmol/l unlabeled 2-DOG and 0.1 μCi/well [3H]2-DOG with or without 100 μmol/l phloretin. Uptake was stopped by rapid aspiration of uptake solution and washing four times with ice-cold PBS containing 50 mmol/l unlabeled 2-DOG. Thereafter, cells were lysed with 10 mmol/l NaOH/0.1% Triton X-100 and radioactivity measured with a liquid scintillation counter. Protein concentration was determined by the Bradford method.
Data are provided as arithmetic means ± SE. Where indicated, n represents the number of independent experiments. All data were tested for significance using ANOVA or paired or unpaired Student’s t test, as applicable. Results with a P value <0.05 were considered statistically significant.
Glucose tolerance tests were performed in mice lacking SGK1 (sgk1−/−) and in their wild-type littermates (sgk1+/+). Before the experiments, the animals were fed a standard diet and allowed free access to either plain tap water (control) or 1% saline solution (high salt). As illustrated in Fig. 1, intraperitoneal injection of glucose (3 g/kg body wt) led to a transient increase of plasma glucose concentration approaching significantly higher values in high salt than in control sgk1+/+ mice. The delayed decrease of the plasma glucose concentrations after a glucose load points to impaired cellular uptake of glucose in saline-drinking animals. The effect of saline was observed in both male (Fig. 1A) and female (Fig. 1B) mice.
The increase of plasma glucose concentration was significantly larger in tap water–drinking sgk1−/− mice than in tap water–drinking sgk1+/+ mice and similar in tap–water drinking sgk1−/− mice and saline-drinking sgk1+/+ mice (Fig. 1A and B). The difference between sgk1−/− and sgk1+/+ mice was again apparent in both male (Fig. 1A) and female (Fig. 1B) mice. In sgk1−/− mice, the increase of plasma glucose concentration was not significantly different between animals drinking tap water and animals drinking saline. Thus, in contrast to sgk1+/+ mice, salt excess did not further affect glucose tolerance in sgk1−/− mice. As illustrated in Fig. 2, the decline of plasma glucose concentrations following intraperitoneal injection of insulin (0.15 units/kg body wt) was similarly blunted by salt excess in sgk1+/+ mice. Again, insulin sensitivity of plasma glucose concentration was significantly smaller in tap water–drinking sgk1−/− mice than in tap water–drinking sgk1+/+ mice. The salt sensitivity of the insulin-induced decrease of plasma glucose concentration was significantly smaller in sgk1−/− mice than in sgk1+/+ mice. Those observations strongly suggest that SGK1 mediates part of the insulin-induced cellular uptake of glucose and substantially contributes to, or even accounts for, the salt sensitivity of glucose tolerance.
As high salt intake decreases the plasma concentration of aldosterone, a known stimulator of SGK1 expression, we explored in sgk1+/+ mice, whether the effect of high salt intake on glucose tolerance was due to decreased mineralocorticoid action. Animals were treated with subcutaneous injection of the mineralocorticoid DOCA (dissolved in soybean oil, 35 mg/kg body wt) 4 h before the glucose tolerance test. The DOCA treatment did not significantly modify the increase of plasma glucose concentrations in control sgk1+/+ mice but significantly blunted the increase of plasma glucose concentrations in saline-drinking sgk1+/+ mice (Fig. 3). After treatment with DOCA, the increase of plasma glucose concentrations was similar in saline-drinking sgk1+/+mice as in sgk1+/+ mice drinking tap water (Fig. 3).
The subcutaneous injection of the mineralocorticoid receptor antagonist spironolactone (dissolved in soybean oil, 50 mg/kg body wt) did not significantly modify the increase of plasma glucose concentrations following glucose injection in control sgk1+/+ mice but reversed the effect of DOCA in sgk1+/+ mice on a high-salt diet. The increase of plasma glucose concentration was not significantly different in saline-drinking sgk1+/+ mice treated with DOCA plus spironolactone and saline-drinking sgk1+/+ mice without pharmacological treatment. The values were, however, significantly higher than the values of untreated tap water–drinking sgk1+/+ mice (Fig. 3).
Further experiments were performed on glucose uptake to identify the tissues accounting for the delayed cellular uptake of plasma glucose in salt-loaded wild-type animals or in animals lacking SGK1. To this end, radiolabeled [H3]2-DOG was injected into the peritoneal cavity and the radioactive tracer was determined in several tissues. As illustrated in Fig. 4, glucose uptake in muscle and fat tissue was significantly lower in saline-drinking sgk1+/+ mice compared with tap water–drinking sgk1+/+ mice. Moreover, during free access to tap water, glucose uptake into muscle, liver, and fat tissue was significantly smaller in sgk1−/− mice than in sgk1+/+ mice. Conversely, glucose uptake into kidney was significantly enhanced in sgk1−/− mice, an effect possibly due to enhanced plasma glucose concentrations.
Western blot analysis was performed to determine whether the decrease of 2-DOG uptake into skeletal muscle and fat tissue was indeed paralleled by a decrease of SGK1 protein expression following salt load. As illustrated in Fig. 5, SGK1 protein abundance was indeed significantly decreased by high-salt intake. Treatment of tap water–drinking mice with DOCA treatment significantly enhanced the SGK1 protein expression, an effect reversed by spironolactone (Fig. 5).
Those experiments pointed to the ability of SGK1 to stimulate cellular glucose uptake. To further test this possibility, constitutively active S422DSGK1 or, for comparison, the inactive mutant K127NSGK1, were expressed in HEK-293 cells and 2-DOG uptake determined. As illustrated in Fig. 6, the transfection with active S422DSGK1 led to a marked increase of 2-DOG uptake. In contrast, transfection with inactive K127NSGK1 decreased 2-DOG uptake. In the presence of phloretin, 2-DOG uptake was markedly reduced and no more sensitive to transfection with S422DSGK1 or K127NSGK1. HEK-293 cells mainly express GLUT1, which is sensitive to phloretin (47).
Excessive salt intake increases blood pressure, an effect paralleled by insulin resistance and/or glucose intolerance in salt-sensitive subjects (3,48). Our present data demonstrate that high salt intake leads to impaired glucose tolerance and impaired insulin sensitivity following an intraperitoneal glucose load. More importantly, the observations provide a cellular mechanism accounting for this pathophysiologically relevant phenomenon. A high-salt diet decreases SGK1 expression, thus disrupting SGK1-dependent glucose uptake into several tissues, including liver, fat, and skeletal muscle. The transfection experiments indeed confirm the ability of SGK1 to profoundly stimulate glucose uptake.
This observation discloses a novel function of SGK1, including the participation in the stimulation of cellular glucose uptake by insulin. Accordingly, SGK1 does not only integrate the effects of mineralocorticoids and insulin on renal tubular Na+ transport (39–42) but similarly affects glucose transport.
The contribution of SGK1 to the regulation of renal tubular Na+ transport is modest, and the SGK1 knockout mouse is seemingly normal and able to maintain normal renal salt excretion under standard diet. However, the elevated plasma aldosterone concentration in sgk1−/− mice points to the functional significance of SGK1-dependent regulation of renal salt conservation even under those dietary conditions. The enhanced plasma aldosterone concentration compensates for the lack of SGK1 and allows the maintenance of normal blood pressure. The maintenance of blood pressure in sgk1−/− mice is consistent with SGK1-independent regulation of renal Na+ reabsorption (43). In contrast, severe salt wasting is observed in the mineralocorticoid receptor knockout mice (49). The ENaC knockout mice survive only a few days after delivery (50). The contribution of SGK1 to renal Na+ reabsorption becomes crucial during salt depletion. The ability of mice lacking functional SGK1 to lower urinary NaCl excretion during salt restriction is impaired, leading to severe, and eventually lethal, volume loss despite decrease of glomerular filtration rate and increase of proximal tubular reabsorption (43).
Saline tended to decrease the insulin sensitivity of blood glucose levels in sgk1−/− mice. Even though the difference between tap water and saline-drinking sgk1−/− mice was not statistically significant, in view of the scatter of the data, the present experiments do not rule out additional SGK1-independent mechanisms leading to insulin resistance during sodium excess. Possible candidates include the other SGK or PKB/Akt isoforms. Clearly, the contribution of SGK1 to the regulation of cellular glucose uptake is modest and does not match the glucose transport stimulating effect of PKB (18,51,52). Particularly, PKB/Akt2 plays a key role in the stimulating effects of insulin on cellular glucose uptake (53–56). Recently, PKB/Akt2 has been shown to be required for the upregulation of skeletal muscle glucose transport during calorie restriction (56). However, cellular glucose uptake is not disrupted in PKB/Akt2-deficient mice, pointing to additional kinases serving a similar function. One of those kinases could well be SGK1, which may fully account for the deranged glucose tolerance during excessive salt intake.
Unlike the expression of PKB, the expression of SGK1 could be upregulated by a variety of hormones, mediators, and other regulators (16). Most importantly, SGK1 is under strong transcriptional control of mineralocorticoids (28–36). Moreover, SGK1 expression is markedly upregulated by cerebral ischemia (57) and may contribute to the enhancement of cellular glucose uptake in ischemic tissues. Along those lines, the enhanced cerebral SGK1 expression in enriched environment (58) may serve to adjust glucose uptake to the enhanced demand following stimulation of neuronal activity. As suggested in Fig. 4, the tissues with SGK1-sensitive glucose uptake may include the brain. Considering the virtually ubiquitous expression of SGK1 (27), it is likely that SGK1 participates in the regulation of nutrient uptake in a variety of further tissues.
Salt loading has previously been shown to decrease SGK1 expression (59), an effect at least partially mediated by a decrease of aldosterone concentration (28–34,36). Accordingly, the effect of salt excess on plasma glucose concentrations following a glucose load can be reversed by replacing the endogenous mineralocorticoids with exogenous DOCA. Moreover, the effect of DOCA is abrogated by the aldosterone receptor antagonist spironolactone. However, the increase of plasma glucose concentration following a glucose load is not significantly affected by spironolactone in the absence of DOCA. Possibly, treatment of tap water–drinking mice with spironolactone leads to volume depletion, which may lead in turn to mineralocorticoid independent stimulation of SGK1 expression (e.g., by increase of angiotensin II).
The present observations do not address the mechanisms of SGK1-dependent regulation of cellular glucose uptake. Clearly, expression of SGK1, but not of the inactive mutant, enhances glucose uptake into HEK-293 cells, indeed confirming the ability of SGK1 to stimulate glucose transport. Most recent in vitro studies (60) revealed the ability of SGK1 to enhance phloretin-sensitive glucose uptake into adipocytes.
The stimulating effect of SGK1 on glucose uptake into fat tissue would be expected to favor the development of obesity. As a matter of fact, a “gain-of-function” gene variant of SGK1 affecting as many as some 5% of a Caucasian population is associated with obesity (24). The obesity may eventually decrease insulin sensitivity despite the stimulating effect of SGK1 on glucose uptake. Moreover, the same gene variant is associated with increased blood pressure (61) presumably due to the salt-retaining property of SGK1 (43). The salt excess would again favor insulin resistance. Thus, SGK1 may participate in the pathophysiology of metabolic syndrome, a condition characterized by a variety of disorders including hypertension, obesity, and insulin resistance (62). Along those lines, metabolic syndrome shares several features with glucocorticoid excess (63), which should upregulate SGK1 expression (25). According to the present data, upregulation of SGK1 may blunt the peripheral insulin resistance caused by glucocorticoids. On the other hand, SGK1 mediates the inhibition of insulin release by glucocorticoids (64). Clearly, additional experiments are required to elucidate the role of SGK1 in glucose metabolism.
In conclusion, excessive salt intake impairs glucose tolerance and insulin-induced glucose uptake. The effect is reversed by application of mineralocorticoids and abrogated in SGK1 knockout mice. The observations point to a novel mechanism participating in the regulation of cellular glucose transport.
K.M.B. and A.M.H. contributed equally to this work.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact
This study was supported by the Deutsche Forschungsgemeinschaft (La 315/4-6, international graduate course), a Marie Curie Research Fellowship (QLGA-CT-2001-52014), and the Department of Veterans Affairs and the National Institutes of Health (DK56248 to V.V.).
We acknowledge the technical assistance of B. Noll and the meticulous preparation of the manuscript by T. Loch and L. Subasic. The constructs encoding active S422DSGK1 and inactive K127NSGK1 were kindly provided by Sir Philip Cohen.