The phosphoinositide-dependent kinase-1 (PDK-1) activates the serum- and glucocorticoid-inducible kinase and protein kinase B isoforms, which, in turn, are known to stimulate the renal and intestinal Na+-dependent glucose transporter 1. The present study has been performed to explore the role of PDK-1 in electrogenic glucose transport in small intestine and proximal renal tubules. To this end, mice expressing ∼20% of PDK-1 (pdk1hm) were compared with their wild-type littermates (pdk1wt). According to Ussing chamber experiments, electrogenic glucose transport was significantly smaller in the jejunum of pdk1hm than of pdk1wt mice. Similarly, proximal tubular electrogenic glucose transport in isolated, perfused renal tubule segments was decreased in pdk1hm compared with pdk1wt mice. Intraperitoneal injection of 3 g/kg body wt glucose resulted in a similar increase of plasma glucose concentration in pdk1hm and in pdk1wt mice but led to a higher increase of urinary glucose excretion in pdk1hm mice. In conclusion, reduction of functional PDK-1 leads to impairment of electrogenic intestinal glucose absorption and renal glucose reabsorption. The experiments disclose a novel element of glucose transport regulation in kidney and small intestine.
- sodium-dependent glucose transporter 1
- phosphatidylinositol-3 kinase
- growth factors
- intestinal glucose transport
electrogenic Na+-dependent glucose transport represents the entry step across the apical brush-border membranes of the small intestine and the renal proximal tubule. Two major Na+-dependent electrogenic glucose transporters (SGLT) have been identified: SGLT-1 (SLC5A1) and SGLT-2 (SLC5A2). SGLT-2 is mainly expressed in kidney. SGLT-1 is the major intestinal electrogenic SGLT and also mediates high-affinity glucose reabsorption in the late proximal tubule of the kidney (46). Inborn defects in SGLT-1 cause intestinal glucose-galactose malabsorption with mild renal glucosuria (41, 46), and SGLT-2 mutations lead to isolated glucosuria (43), underlining the physiological importance of those carriers in epithelial glucose transport.
Despite their importance in glucose transport, only little is known about the regulation of these transporters. Several lines of evidence suggest, however, that SGLT-1 is regulated by kinases (8, 18, 19, 36, 44, 45).
According to in vitro experiments, PKB (6, 7) and the serum and glucocorticoid-inducible kinase (SGK) family members SGK1 (11), SGK2 (22), and SGK3 (22) are able to upregulate a variety of channels and transporters (27), including the renal and intestinal glucose transporter SGLT-1 (8).
The kinases are activated by IGF-I and insulin through the phosphatidylinositol-3 kinase and phosphoinositide-dependent kinase-1 (PDK-1) (1, 2, 9, 13, 21, 24, 33). Accordingly, lack of PDK-1 is expected to impair the function of the SGK and PKB isoforms and, in turn, the respective transport systems.
The present experiments have been performed to elucidate the significance of those in vitro studies for SGLT-1 activity in vivo. As the PDK-1 knockout mouse is not viable (30), we analyzed the PDK-1 hypomorphic mouse (pdk1hm), expressing only some 20% of PDK-1 activity compared with the wild-type littermates (pdk1wt) (30). SGLT-1 activity was studied as glucose-induced current (Iglc) in the renal proximal tubule and small intestine of pdk1hm and pdk1wt mice and indeed was found to be decreased in mice with reduced PDK-1 expression.
Generation and basic properties of PDK-1 hypomorphic mice have been described previously (30). Genotyping was made by PCR on tail DNA using PDK-1 and neo-R-specific primers, as previously described (30). Mice had free access to standard mouse diet (C1310, Altromin, Langen, Germany) and tap water. All animal experiments were conducted according to the guidelines of the American Physiological Society and the German law for the welfare of animals and were approved by the local authorities.
Ussing chamber experiments in small intestine.
For analysis of electrogenic intestinal glucose transport, animals were killed, the abdomen was opened, and the intestine was quickly removed. Jejunal segments (5–10 cm postpylorus) were mounted into a custom-made mini-Ussing chamber with an opening diameter of 0.99 mm and an opening area of 0.00769 cm2. Under control conditions, the serosal and luminal perfusate contained (in mM) 105 NaCl, 2 KCl, 1 MgCl2, 1.25 CaCl2, 0.4 KH2PO4, 1.6 K2HPO4, 5 Na pyruvate, 25 NaHCO3, and 20 mannitol. Where indicated, 20 mM glucose were added to the luminal perfusate at the expense of mannitol. All solutions were gassed with 95% O2-5% CO2 for at least 60 min until usage in the experiment, without altering pH. All substances were from Sigma (Taufkirchen, Germany) or Roth (Karlsruhe, Germany).
Electrogenic glucose transport in isolated, perfused, proximal straight tubules.
Experiments have been performed on 0.2- to 0.4-mm segments of proximal straight tubules following principally the method of Burg et al. (5). Modifications of the technique have been described in detail previously (15, 16). The luminal perfusion rate was <10 nl/min. The bath was continuously perfused at a rate of 20 ml/min and thermostated with a dual-channel feedback system (Hampel, Frankfurt, Germany). The bath temperature was kept constant at 38°C. The potential difference across the basolateral cell membrane (PDbl) was determined with and without glucose in the perfusate to stimulate electrogenic reabsorption, as described previously (42). The bath and luminal perfusates were composed of the following (in mM): 120 NaCl, 5 KCl, 20 NaHCO3, 1.3 CaCl2, 1 MgCl2, and 2 Na2HPO4. In the bath (in mM), 1 glucose, 18 mannitol, 2 glutamine, and 1 Na-lactate, and, in the lumen, 20 mannitol and 1 BaCl2 were added. The luminal addition of Ba2+ prevents the possibly PDK-1-dependent activation of apical K+ channels, which would lead to underestimation of Iglc. Where indicated (in mM), 20 mannitol were replaced by 20 glucose in the luminal perfusate. PDbl was measured with intracellular Ling Gerard electrodes (∼100 MΩ) by a high-impedance electrometer (FD223, WPI, Science Trading, Frankfurt, Germany) connected with the electrode via an Ag-AgCl half cell. An Ag-AgCl reference electrode was connected with the bath. Entry of positive charge by electrogenic transport is expected to depolarize the basolateral cell membrane. The magnitude of the depolarization depends on the magnitude of the induced current on the one hand and on the resistances of cell membranes and shunt on the other.
Glucose load and glucose excretion.
For determination of glucose tolerance and renal glucose excretion, unfasted mice were injected with 3 mg/g body wt glucose in a volume of 30 μl aqua ad injectabilia/g body wt intraperitoneally (ip). Control experiments in both genotypes were performed, injecting 30 μl 0.9% NaCl/g body wt. Blood glucose was measured after tail-vein bleeding using a glucometer (Accutrend, Roche, Mannheim, Germany) before and at 30, 60, and 120 min after the injection of glucose. Throughout the experiment, mice were placed in individual metabolic cages (Tecniplast, Hohenpeissenberg, Germany) for the collection of a spot urine sample after injection over the next 3 h, yielding 0.1–1.8 ml urine. Urinary glucose and creatinine concentrations were determined utilizing commercial enzymatic kits (gluco-quant, Roche Diagnostics, Mannheim, Germany, based on the hexokinase method, and creatinine PAP, Labor & Technik, Berlin, Germany, based on the creatininase method). For quantitative analysis of glucosuria, the ratio between glucose and creatinine concentration was calculated (in mg glucose/mg creatinine) to adjust for differences in urine dilution.
Preparation of brush-border membrane vesicles.
Brush-border membrane vesicles were prepared from whole mouse kidney, jejunum, and ileum using the Mg2+ precipitation technique, as described previously (4). After measurement of the total protein concentration (Biorad Protein kit), 20 μg of brush-border membrane protein were solubilized in Laemmli sample buffer, and SDS-PAGE was performed on 10% polyacrylamide gels. For immunoblotting, proteins were transferred electrophoretically from unstained gels to polyvinylidene difluoride membranes (Immobilon-P, Millipore, Bedford, MA). After blocking with 5% milk powder in Tris-buffered saline/0.1% Tween 20 for 60 min, the blots were incubated with affinity-purified rabbit anti-mouse SGLT-1 antibody (Chemicon, Temecula, CA) (1:3,000) and mouse monoclonal anti-actin (42 kDa, Sigma) 1:500, either for 2 h at room temperature or overnight at 4°C. After washing and subsequent blocking, blots were incubated with secondary antibodies conjugated with alkaline phosphatase or horseradish peroxidase (goat anti-rabbit 1:5,000 and donkey anti-mouse 1:5,000; Promega), for 1 h at room temperature. Antibody binding was detected with the enhanced chemiluminescence kit (Pierce) in the case of horseradish peroxidase-linked antibodies and with the CDP Star kit (Roche) for AP-linked antibodies before detection of chemiluminescence with the Diana III chemiluminescence detection system. Bands were quantified with the Aida Image Analyzer software (Raytest).
Data are provided as means ± SE; n represents the number of independent experiments. All data were tested for significance using the unpaired or paired Student t-test with Welch correction where applicable, and only results with P < 0.05 were considered statistically significant.
To determine PDK-1-dependent glucose transport, segments of jejunum from pdk1hm and pdk1wt mice were mounted into mini-Ussing chambers, and electrogenic glucose transport was determined utilizing electrophysiological analysis (Fig. 1).
The isoosmotic replacement of mannitol by glucose generated a lumen-negative shift of the transmural potential difference without significantly altering the transmural resistance. The glucose-induced alterations of transepithelial voltage and transepithelial resistance allowed the calculation of the Iglc. In proximal segments of pdk1hm, Iglc amounted to −943 ± 139 μA/cm2 (n = 10) for 20 mM glucose. In proximal segments of pdk1wt mice, Iglc approached −1,507 ± 210 μA/cm2 (n = 13). As summarized in Fig. 1B, the currents were significantly smaller in proximal jejunum obtained from pdk1hm than in the proximal jejunum of pdk1wt mice.
Similar observations were made in distal jenunal segments. In distal segments of pdk1hm, Iglc amounted to −785 ± 133 μA/cm2 (n = 11) for 20 mM glucose. In distal segments of pdk1wt mice, Iglc approached −1,291 ± 127 μA/cm2 (n = 12). Again, the currents were significantly smaller in pdk1hm than in pdk1wt mice (Fig. 1C).
According to intracellular impalements, the PDbl of isolated, perfused, straight proximal tubules (i.e., late aspects of proximal tubule) was in the absence of glucose not different between pdk1hm (−54.6 ± 1.3 mV, n = 7) and pdk1wt (−54.9 ± 2.4 mV, n = 6) mice (Fig. 2). Addition of 20 mM glucose to the luminal fluid replacing mannitol significantly decreased PDbl in both pdk1hm and pdk1wt mice, an effect, however, significantly smaller in pdk1hm than in pdk1wt mice (Fig. 2).
PDK-1 could regulate electrogenic glucose transport by enhancing expression and/or by increasing the activity of expressed protein. To explore the effect of PDK-1 on SGLT-1 expression, immunoblotting was performed in brush-border membranes of the proximal tubule, jejunum, and ileum. No significant difference in SGLT-1 protein abundance could be detected in the brush-border membrane fraction obtained from kidney, jejunum, and ileum from pdk1hm and pdk1wt mice (n = 5 for each genotype, Fig. 3).
Further experiments were performed to elucidate the plasma concentrations and renal excretion of glucose before and following a glucose load of 3 g/kg body wt. As illustrated in Fig. 4, intraperitoneal injection of glucose led to a transient increase of plasma glucose concentration, approaching similar values in pdk1hm and pdk1wt mice. The peak glucose concentration tended to be higher, and the subsequent decline of plasma glucose concentration tended to be slower, in pdk1hm than in pdk1wt mice, a difference, however, not statistically significant. To explore whether the filtered glucose load during the intraperitoneal glucose injection exceeded the maximal renal tubular glucose reabsorption, mice were placed in metabolic cages, and urinary glucose concentration was determined. To account for individual variations of urinary flow rate, the urinary excretion of glucose was divided by the respective creatinine excretion. As shown in Fig. 5 and Table 1, intraperitoneal glucose administration led to a transient glucosuria, which was markedly higher in pdk1hm than in pdk1wt mice.
According to the Ussing chamber experiments and electrophysiology of proximal renal tubules, electrogenic glucose transport is significantly smaller in pdk1hm than in pdk1wt mice. PDK-1 could regulate SGLT-1 through the downstream kinases PKB and SGK, which have previously been shown to stimulate the electrogenic glucose transporter SGLT-1 expressed in Xenopus oocytes (8). Moreover, these kinases also stimulate the voltage-gated K+ channel complex KCNE1/KCNQ1 (10), which contributes to the maintenance of the potential difference across the apical cell membrane of the renal proximal tubule (42), a critical driving force for electrogenic glucose transport (28, 29). K+ channels similarly maintain the driving force of intestinal glucose transport (37). The kinases further stimulate the Na+-K+-ATPase (17, 38), which is required to maintain the chemical driving force for Na+-coupled nutrient transport (28, 29, 37). Thus PDK-1 may regulate SGLT-1 activity, both directly and by modifying the respective driving forces.
Decreased expression of PDK-1 in the PDK-1 hypomorphic mice does not only affect glucose transport but may affect a variety of other transport systems dependent on the SGK and PKB isoforms (26). Again, those transport systems may be affected by a direct regulation and/or by modification of the driving forces.
Deranged transport may contribute to the smaller size of the PDK-1 hypomorphic mice reported earlier (30). Interestingly, the animals excrete slightly less creatinine (see Table 1), pointing to some decrease of creatinine formation. Recent experiments have disclosed the ability of SGK isoforms to stimulate the creatine transporter (39). As formation of creatinine requires the cellular uptake of creatine (34), impaired function of the creatine transporter is expected to decrease the formation and renal excretion of creatinine. The increased ratio of glucose over creatinine concentration in urine is, however, largely due to marked glucosuria rather than slightly decreased creatinine excretion.
The protein abundance of SGLT-1 in the brush-border membrane of the jejunum, ileum, and kidney was not significantly different between pdk1hm mice and pdk1wt mice. The scatter of the data does not, however, preclude effects of PDK-1 on SGLT-1 expression. Moreover, the residual PDK-1 activity in the hypomorphic mice may be sufficient to maintain expression. Irrespective of the underlying mechanism, the observations disclose that SGLT-1 activity in the brush-border membrane is regulated by PDK-1. The β-adrenergic regulation of glucose transport in rat small intestine has previously been shown to involve phosphorylation-dependent SGLT-1 stimulation (19), whereas IGF-mediated stimulation of intestinal glucose transport via SGLT-1 has been reported to involve increase of mRNA and protein abundance (25).
The effect of PDK-1 deficiency is only moderate but could be more profound at complete knockout of PDK-1. Complete knockout of SGK3 leads to a moderate decrease of intestinal SGLT-1 activity (36). Complete knockout of SGK1 abrogates the SGLT-1-stimulating effect of dexamethasone without appreciably affecting the basal activity of SGLT-1 (14).
The defective intestinal and renal glucose transport did not appreciably alter basal plasma glucose concentration. The increase of plasma glucose concentration following intraperitoneal glucose administration tended to be slightly higher in pdk1hm mice than in pdk1wt mice. The difference, however, did not reach statistical significance. Thus the ability of nonpolarized cells such as muscle or liver to accumulate glucose following an increase of plasma glucose concentration is not severely affected in the PDK-1 hypomorphic mouse. This observation may be surprising, in view of the substantial role of PKB (3, 12, 20, 23, 31, 35, 40) and SGK1 (32) in peripheral glucose uptake. Again, it should be kept in mind that the hyopomorphic mice do not completely lack PDK-1, which would not be compatible with survival. Thus the residual PDK-1 activity may be able to maintain almost normal glucose uptake into peripheral tissues, but may not be sufficient for full stimulation of intestinal and renal glucose transport.
In conclusion, the PDK-1 hypomorphic mice display moderate impairment of electrogenic epithelial glucose transport, an observation disclosing a novel player in the regulation of intestinal and renal nutrient transport.
This work was supported by grants from the Deutsche Forschungsgemeinschaft (La 315/4–7 and GRK 1302), the Bundesministerium für Wissenschaft u. Forschung, and the European Commission (LSHM-CT-2003–502852; EUGINDAT) EU.
The authors acknowledge the meticulous preparation of the manuscript by Lejla Subasic and Tanja Loch.
↵* F. Artunc, R. Rexhepaj, and H. Völkl 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