HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Seghers, V. Articles by Bryan, J. Search for Related Content PubMed PubMed Citation Articles by Seghers, V. Articles by Bryan, J. Social Bookmarking                      What's this?

J Biol Chem, Vol. 275, Issue 13, 9270-9277, March 31, 2000

Sur1 Knockout Mice
A MODEL FOR K_ATP CHANNEL-INDEPENDENT REGULATION OF INSULIN SECRETION^*

Victor Seghers^§, Mitsuhiro Nakazaki^§, Franco DeMayo, Lydia Aguilar-Bryan^¶, and Joseph Bryan

From the Departments of  Molecular and Cellular Biology and ^¶ Medicine, Baylor College of Medicine, Houston, Texas 77030

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Sur1 knockout mouse -cells lack K_ATP channels and show spontaneous Ca²⁺ action potentials equivalent to those seen in patients with persistent hyperinsulinemic hypoglycemia of infancy, but the mice are normoglycemic unless stressed. Sur1^-/- islets lack first phase insulin secretion and exhibit an attenuated glucose-stimulated second phase secretion. Loss of the first phase leads to mild glucose intolerance, whereas reduced insulin output is consistent with observed neonatal hyperglycemia. Loss of K_ATP channels impairs the rate of return to a basal secretory level after a fall in glucose concentration. This leads to increased hypoglycemia upon fasting and contributes to a very early, transient neonatal hypoglycemia. Whereas persistent hyperinsulinemic hypoglycemia of infancy underscores the importance of the K_ATP-dependent ionic pathway in control of insulin release, the Sur1^-/- animals provide a novel model for study of K_ATP-independent pathways that regulate insulin secretion.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

ATP-sensitive potassium channels (K_ATP channels)¹ are a unique combination of a K⁺ inward rectifier (either K_IR6.1 or K_IR6.2) and a sulfonylurea receptor (SUR1 or SUR2) transport ATPase superfamily members (1-3). These channels respond to changes in ATP/ADP and can couple metabolism to membrane electrical activity. SUR1 and K_IR6.2 comprise the K_ATP channels in pancreatic -cells that regulate the ionic pathway mediating glucose-stimulated insulin secretion by setting the resting membrane potential below the activation threshold for voltage-gated Ca²⁺ channels (4).

Mutations in human Sur1 or K_IR6.2 cause a recessive form of persistent hyperinsulinemic hypoglycemia of infancy (PHHI) characterized by oversecretion of insulin despite severe hypoglycemia (1, 5). Surprisingly, two recent studies involving disruption of K_ATP channels in mice produced a quite different picture. Targeted overexpression of a dominant-negative K_IR6.2 subunit, K_IR6.2_G132S, in -cells reduced channel activity producing animals that were hypoglycemic at birth but became increasingly hyperglycemic secondary to -cell death (6).

K_IR6.2 null mice, K_IR6.2^-/-, on the other hand, completely lack -cell K_ATP channels but exhibit a less severe phenotype (7). The K_IR6.2^-/- animals have nearly normal blood glucose levels, showing mild glucose intolerance when challenged with glucose. These animals are reported to release a small amount of insulin in response to glucose, whereas isolated, perifused islets show a small first phase of glucose-stimulated insulin secretion and no second phase. The normal blood glucose levels have been attributed to insulin hypersensitivity secondary to the loss of SUR2A/K_IR6.2 K_ATP channels in skeletal muscle.

SUR1 null mice, Sur1^-/-, unlike their K_IR6.2^-/- counterparts, are not insulin-hypersensitive. Isolated Sur1^-/- islets exhibit a pattern of glucose-stimulated insulin release consistent with regulation by an underlying K_ATP-independent pathway (or pathways), the nature of which is unknown (8-12). The Sur1^-/- mice are both significantly more hyperglycemic when glucose-loaded and significantly more hypoglycemic when fasted than the control animals. Potential physiologic mechanisms of regulation of blood glucose by K_ATP-independent insulin secretion are discussed.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Animals and Genotyping-- The 5'-end of the mouse Sur1 gene was cloned from a mouse genomic library constructed in  phage. The approved symbol for the Sur1 gene is Abcc8. A targeting vector was constructed that contained 5.5 kilobase pairs of DNA, including the 5'- and 3'-intronic sequence flanking a puromycin resistance cassette that replaced the second exon of Sur1. Thymidine kinase was included downstream of the 3' region of homology for negative selection. The vector was linearized and used to electroporate RW4 embryonic stem cells (Genome Systems, Inc., Palo Alto, CA). Transfected cells were selected with puromycin (3 µg/ml) and 1-(-2-deoxy-2-fluoro-1--D-arabinofuranosyl)-5-iodouracil (3.5 nM); colonies were picked after 12 days and screened for targeted disruption by Southern blotting. Correct recombination was verified by polymerase chain reaction from both the 5'- and 3'-ends of the targeted region. Embryonic stem cells were injected into C57BL/6 blastocysts using standard techniques. Chimeric males were selected and crossed with wild type C57BL/6 females to generate Sur1^+/- heterozygotes, which were bred to obtain Sur1^+/+, Sur1^+/-, and Sur1^-/- animals. Animals were maintained on a 12-h light/dark cycle and fed standard rodent chow.

Photolabeling-- Brains from control and Sur1^-/- mice were used to isolate membranes as described (13). Sur1 was identified by photolabeling with [¹²⁵I]azidoiodoglibenclamide as described previously (14).

Glucose and Insulin Measurements-- Blood glucose levels were measured using Precision Q·I·D glucose sensors (MediSense, Inc., Bedford, MA). Microtainer® serum separator tubes (Becton Dickinson, Franklin Lakes, NJ) were used to extract serum. Plasma insulin was measured with a rat insulin ELISA kit using mouse insulin as a standard (Crystal Chem, Inc., Chicago, IL). Medium insulin was measured using a rat insulin RIA kit (Linco Research, Inc., St. Charles, MO). Fed and fasted tail blood samples were taken from randomly feeding animals or following a 16-h fast. Intraperitoneal glucose and insulin tolerance tests were done on 12-16-week-old male mice following a 16-h fast. Animals were injected intraperitoneally with glucose (1.2 g/kg of body weight) or insulin (0.1 units/kg of body weight) solubilized in 0.9% NaCl. Animals were anesthetized with 65 mg/kg of body weight of sodium pentobarbital approximately 15 min prior to drawing blood from the retro-orbital venous sinus.

Analysis of the Isolated Islets and Islet Cells-- Animals were anesthetized using sodium pentobarbital as described above. Islets were isolated after digestion of the pancreas by intraductal injection of 1 mg/ml collagenase P (Roche Molecular Biochemicals) dissolved in Krebs-Ringer bicarbonate buffer (KRBB) solution as described elsewhere (15, 16). Collected islets were used directly or dispersed mechanically into single cells in a Ca²⁺-free KRBB solution for electrophysiological measurements. For batch experiments, islets were preincubated in RPMI 1640 medium containing 2.8 mM glucose, and then the islets (10 islets/well in 24-well plates on Transwell^TM permeable supports, 6.5 mm in diameter, 8.0 µm pore size; Corning Inc., Corning, NY) were incubated overnight in 0.75 ml of RPMI 1640 medium supplemented with 10% fetal bovine serum and test concentrations of glucose equilibrated with 5% CO₂/95% air, pH 7.4, at 37 °C. Perifusion was done following Komatsu et al. (17), with modifications. One hundred islets in a column of Bio-Gel P-10 (Bio-Rad) were continuously perifused at 37 °C with Hepes-NaHCO₃ KRBB with 1 mg/ml bovine serum albumin (equilibrated with 5% CO₂/95% air, pH 7.4) at a flow rate of 0.75 ml/min. After perifusion for 40 min under 2.8 mM glucose, solutions were changed as indicated in the figure. Test substances were added to the basal medium without adjustment of the final osmolarity.

Electrophysiology-- Dispersed islet cells were cultured for 1-3 days in RPMI 1640 medium containing 11.1 mM glucose with 100 µg/ml streptomycin, 100 IU/ml penicillin, and 10% fetal calf serum at 37 °C in 5% CO₂. Standard patch clamp techniques were used to record ion channel currents (18). The pipette solution used for cell-attached recording contained 140 mM KCl, 2 mM CaCl₂, and 11 mM HEPES, pH 7.2. The standard internal solution for whole-cell recording contained 50 mM KCl, 35 mM K₂SO₄, 2.0 mM MgCl₂, 11 mM EGTA, 1. mM0 CaCl₂, and 11 mM HEPES, and with or without ATP at the indicated concentrations, pH 7.2. The pipette solution for perforated patch recording contained 40 mM K₂SO₄, 50 mM KCl, 10 mM HEPES, 2.0 mM MgCl₂, and 0.5 mM EGTA, pH 7.2. Amphotericin B (240 µg/ml dissolved in 0.4% Me₂SO) was included in the pipette to perforate the membrane. Isolated islet cells were superfused with KRBB containing 2.8 mM glucose, 129 mM NaCl, 4.7 mM KCl, 2.0 mM CaCl₂, 1.2 mM MgCl₂, 1.2 mM KH₂PO₄, and 5.0 mM NaHCO₃, pH 7.4, for at least 30 min before recording. Currents were recorded using an AXOPATCH-1C (Axon Instruments Inc., Foster City, CA) amplifier and analyzed using pCLAMP (Axon Instruments Inc.). All experiments were performed at room temperature (22-25 °C). -Cells were identified by a combination of their morphology and the presence of ATP-sensitive potassium currents.

Statistics-- One-way analysis of variance was used to evaluate the significance of measurements. Tukey's post-test was used to compare pairs of group means. Statistical calculations were done using GraphPad Instat (GraphPad Software, San Diego, CA). We have summarized the results using mean values and standard deviations to accurately reflect the variability of the population.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Targeted Inactivation of the Sur1 Gene-- The Sur1 gene was inactivated by replacing the second exon in mouse embryonic stem cells with a puromycin resistance cassette (Fig. 1A). The targeting vector was transfected into 129/SvJ embryonic stem cells by electroporation. Puromycin resistant clones were isolated and screened by Southern blotting to identify homologous recombinants. Two clones were injected into C57 BL/6 blastocysts to generate germline chimeras. The identification of Sur1^+/+, Sur1^-/+ and Sur1^-/- mice is shown (Fig. 1, B and C). The absence of SUR1 was confirmed using neuronal membranes. Sur1 was identified by photolabeling with [¹²⁵I]azidoiodoglibenclamide as described previously (14) and is detectable in heterozygous, but not homozygous, Sur1^-/- animals (Fig. 1D).

View larger version (46K): [in this window] [in a new window]   Fig. 1.   Generation and screening of Sur1-/- animals. A, schematic of the first 5 exons of the 39-exon Sur1 gene. Exon 2 was replaced with the puromycin resistance gene, and a negative selection marker encoding thymidine kinase (TK) was placed downstream of the 3' region of DNA homology. B and C, Southern blotting and polymerase chain reaction (PCR) were used to identify Sur1+/+, Sur1+/-, and Sur1-/- mice. D, neuronal membranes from Sur1+/+, Sur1+/-, and Sur1-/- mice were isolated and photolabeled with 3 nM [125I]azidoiodoglibenclamide in the presence or absence of unlabeled 1 µM glibenclamide and then solubilized and separated by SDS-polyacrylamide gel electrophoresis.

The heterozygous and homozygous animals are viable, fertile, and phenotypically normal. The Sur1^-/- mice showed no obvious changes in birth weights, weight gains, or behavior to distinguish them from their wild type littermates.

Sur1^-/- -Cells Lack K_ATP Channels-- The loss of K_ATP channels in isolated pancreatic islet cells, predominantly -cells, was established using the patch clamp technique. In the whole cell recording mode, control cells show a marked increase in K⁺ current that peaks at approximately 2 min (3.21 ± 0.62 nS/pF) after membrane rupture and dialysis of intracellular nucleotides and then proceed to "rundown" in the absence of ATP (Fig. 2A). The Sur1^-/- cells exhibit no comparable K⁺ currents. Comparison of the peak values of the normalized conductances of control versus heterozygous -cells indicates that their densities of K_ATP channels are the same (Fig. 2A, inset), implying a mechanism for regulation of channel density although the number of genes has been halved. Inclusion of ATP (1 mM) in the patch pipette reduces the peak currents and rate of channel rundown.

View larger version (38K): [in this window] [in a new window]   Fig. 2.   Electrophysiology of Sur1+/-, Sur1+/-, and Sur1-/- -cells. A, K+ currents increase in Sur1+/+ but not Sur1-/- -cells as cytoplasmic ATP falls during whole-cell recording, as shown in a plot of normalized conductance versus time. The holding potential was -70 mV; test voltage pulses, 100 ms duration of 10 mV, were applied every 3 s. The control currents peaked at about 2 min and then proceeded to rundown. No currents were detected in the Sur1-/- -cells. The inset shows that the normalized conductance from control and heterozygous Sur1± -cells are equivalent. Inclusion of 1 mM ATP in the pipette solution reduces the peak currents and rate of rundown. The error bars are ± S.D.; n = 10 cells. B, Sur1-/- -cells exhibit spontaneous action potentials. Recording from Sur1+/+ control cells in cell-attached mode illustrates the presence of KATP channels in the patch, whereas Sur1-/- cells exhibited spontaneous action potentials that were suppressed by the L-type Ca2+ channel blocker nifedipine but not by diazoxide. The holding potential was 0 mV; cells were incubated in 2.8 mM glucose.

Sur1^-/- -Cells Show Spontaneous Electrical Activity-- The electrical activity of Sur1^+/+ and Sur1^-/- -cells in low glucose solutions is markedly different. K_ATP channels are active in control cells (downward deflections in Fig. 2B, top trace); the average resting membrane potential was -62.4 ± 12.3 mV, n = 48. By contrast, the membrane potential of the Sur1^-/- cells fluctuated between -45 and -25 mV (82%, n = 28; average value, -32.6 ± 6.2 mV, n = 28) as a result of spontaneous action potentials (Fig. 2B, bottom traces) that were not suppressed by diazoxide but were inhibited by the Ca²⁺ channel blocker nifedipine.

Sur1^-/- Islet Histology Is Nearly Normal-- Terminal deoxynucleotidyl transfer-mediated nick end labeling assays on control and Sur1^-/- adult islets showed no differences in apoptosis. Staining for insulin and somatostatin showed no differences, whereas glucagon staining revealed that the peripheral distribution of -cells was disrupted with an increased number of glucagon positive cells within the interior of the central -cell mass (data not shown). The cause of this redistribution is unclear, but it has been reported for the K_IR6.2^-/- mice (7) and for the -cell specific knockout of the insulin receptor mice (19), among others.

Sur1^-/- Islets Exhibit an Altered Response to Glucose Stimulation-- Short term perifusion of islets revealed marked differences in glucose-stimulated insulin secretion. Control islets showed a brisk first phase of insulin secretion after an up-shift in glucose concentration that was entirely absent in the Sur1^-/- islets (Fig. 3A). The Sur1^-/- islets exhibited a rise in insulin output that we interpret as an attenuated second phase release. This second release, measured as the average rate of release during the plateau period between 18 and 38 min, was approximately 35% of the control value. The control islets in 16.7 mM glucose exhibited a marked hypersecretion of insulin in response to tolbutamide, whereas the Sur1^-/- islets showed no increased secretion. Short term static incubation experiments confirmed this lack of response to tolbutamide.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Sur1-/- islets exhibit changes in both the kinetics and extent of insulin secretion. Glucose-stimulated insulin release from control (circles) and Sur1-/- (squares) islets during continuous perifusion. A, in a short term experiment, the concentration of glucose (G) was increased from 2.8 to 16.7 mM between 10 and 50 min. Tolbutamide (Tlb), 0.3 mM, was added to the medium starting at 40 min. B, in long term perifusion experiments, the concentration of glucose was increased from 2.8 to 16.7 mM between 0 and 500 min. Note the break in the time line and scale change at 25 min. The longer time values are the mean values of seven 1-min time points flanking the given time, and the error bars are ±S.D. C, the islets in B were switched to 2.8 mM glucose to follow their rates of return to a basal secretory level. The data were normalized to an initial value of 1 by dividing by the average value immediately before switching to 2.8 mM glucose. The results from two separate experiments are plotted; the lines are the best fits of a monoexponential function to the combined data sets. The time course of insulin output after changes in glucose concentration was assayed from 100 Sur1+/+ or Sur1-/- islets during continuous perifusion.

Glucose-stimulated insulin release from control and Sur1^-/- islets over an 8-h period is summarized in Fig. 3B. Note the broken time line and the change of scale on the y axis. The Sur1^-/- islets showed about a 4-fold increase in their rate of insulin release within 5 min; both the control and Sur1^-/- islets exhibited a slow, parallel increase in insulin output through 500 min. The rate of insulin release of the Sur1^-/- islets at 400 or 500 min was approximately 30% of the controls. The results imply regulation of insulin release by a slow, K_ATP-independent, glucose-stimulated mechanism.

To determine whether K_ATP channels play a role in returning insulin secretion to a basal level after a down-shift in glucose, the control and Sur1^-/- islets were switched to 2.8 mM glucose. Insulin output from the control islets dropped rapidly, t_1/2 = ~5 min. By contrast, insulin output from Sur1^-/- islets fell slowly, t_1/2 = ~67 min, averaged for these two experiments (Fig. 3C). The results indicate K_ATP channels, presumably by lowering the resting membrane potential, play a major role in wild type -cells in "switching off" insulin release when glucose levels fall. The Sur1^-/- results show that the K_ATP-independent mechanism(s) has a much slower response time.

Sur1^-/- Islets Show Glucose-responsive Insulin Secretion-- The glucose-responsiveness of isolated islets was tested in extended static incubations at various glucose concentrations. Insulin release from both Sur1^-/- and control islets, integrated over 20 h, increased with glucose concentration (Fig. 4), with half-maximal stimulation between 9 and 10 mM. Insulin output at the lowest glucose concentrations was indistinguishable, whereas the maximum output from the Sur1^-/- islets was ~60% of the control value (285.1 ± 60.2 versus 494.5 ± 88.9, p < 0.0005). This insulin secretion requires Ca²⁺ influx and is sensitive to the Ca²⁺ channel blocker nifedipine (Fig. 4, inset), implying that insulin output is coupled with glucose concentration via a Ca²⁺-sensitive mechanism. In Fig. 4, the hatched bar centered at ~6.5 mM glucose indicates the range of glucose values in randomly feeding adult control and Sur1^-/- mice (see Fig. 6).

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Extended culture of control and Sur1-/- islets demonstrates glucose-dependent insulin secretion. Islets were cultured for 20 h at the indicated glucose concentrations before assay of released insulin. The control (circles) and Sur1-/- islets (squares) exhibited glucose-stimulated insulin release, but secretion from the Sur1-/- islets was impaired relative to control islets. At 16.7 mM glucose, the values were 494.5 ± 88.9 versus 285.1 ± 60.2 ng/ml/10 islets/20 h (n = 10; mean ± S.D.; p < 0.0005), respectively. The lines are best fits to a logistic equation; the half-maximal values are 9.0 and 9.8 mM for the control and Sur1-/- groups, respectively. For comparative purposes, the hatched bar indicates the range of blood glucose concentrations seen in randomly feeding adult animals. The inset demonstrates that secretion in response to 16.7 mM glucose is sensitive to nifedipine (1 µM). The error bars are ±S.D.

Sur1^-/- Mice Are Mildly Glucose Intolerant-- In response to a glucose challenge the Sur1^-/- animals exhibit mild glucose intolerance (Fig. 5A); their peak glucose values are higher and their clearance times longer versus control mice (15.1 ± 3.2 versus 13.5 ± 2.6 mM at 15 min, p < 0.01, and 12.8 ± 2.0 versus 7.4 ± 0.8 mM at 60 min, p < 0.001). Consistent with the lack of first phase release from isolated islets, the Sur1^-/- animals showed little increase in insulin output after intraperitoneal glucose injection, in marked contrast to the robust increase seen in control animals (Fig. 5B). These results suggest that mice have both insulin-induced and insulin-independent pathways for glucose disposal.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Sur1-/- animals display impaired glucose tolerance but are not insulin-hypersensitive. Control and Sur1-/- male mice (12-16 weeks of age) were injected intraperitoneally with glucose (1.2 g/kg of body weight) following a 16-h fast. A, blood glucose levels of the Sur1-/- animals were significantly higher than those of the controls at 30 and 60 min (15.1 ± 3.2 mM versus 13.5 ± 2.6 mM at 15 min; n = 13 and 10; p < 0.01; and 12.8 ± 2.0 mM versus 7.4 ± 0.8 mM at 60 min; n = 14 and 9; p < 0.001). B, control animals release insulin in response to a glucose challenge, whereas the Sur1-/- animals do not. C, intraperitoneal injection of porcine insulin, 0.1 unit/kg of body weight, produced an equivalent glucose lowering effect in similarly fasted animals. The error bars are ±S.D.

In contrast to what has been reported for the K_IR6.2^-/- mice (7), the Sur1^-/- animals showed no increased insulin sensitivity. Following intraperitoneal injection of insulin. the glucose values of both groups dropped with no significant differences (Fig. 5C).

Sur1^-/- Mice Exhibit Glucose-stimulated Insulin Secretion-- The blood glucose levels of random-fed adult Sur1^-/- mice were normal, and their plasma insulin/blood glucose ratios did not differ significantly from those of the controls (Fig. 6, A and B). By contrast, Sur1^-/- animals fasted for 16 h were significantly more hypoglycemic than similarly fasted control animals (2.76 ± 0.65 versus 3.63 ± 0.44 mM, p < 0.05 for males, and 2.76 ± 0.94 versus 3.82 ± 0.52 mM, p < 0.05, for females). This result is consistent with the inability of Sur1^-/- islets to restrict their insulin output rapidly when glucose levels fall. The data in Fig. 3C suggest the plasma insulin levels of the Sur1^-/- mice will be higher than those of the controls for nearly 3 h. Although the insulin levels are expected to equalize before the end of the fast, the Sur1^-/- mice appear to be unable to compensate for the increased insulin and bring their glucose levels back to the fasted control level. The fasted plasma insulin/blood glucose ratios support this interpretation; both male and female Sur1^-/- mice have higher mean ratios than controls (0.33 ± 0.13 versus 0.15 ± 0.07 for males and 0.24 ± 0.11 versus 0.15 ± 0.05 ng/ml/mM for females), although only the males are significantly different (p < 0.001).

View larger version (36K): [in this window] [in a new window]   Fig. 6.   Adult and neonatal glucose and insulin levels. Glucose and insulin values were measured in male and female adult mice, 12-16 weeks of age, in the fed and fasted states. A, the blood glucose levels of control and knockout animals drop in response to fasting, but the Sur1-/- animals have lower values (males: 2.76 ± 0.65 versus 3.63 ± 0.44 mM, p < 0.05, n = 20 or 19; females: 2.76 ± 0.94 versus 3.82 ± 0.52 mM, p < 0.05, n = 16 or 15). There were no significant differences by one-way analysis of variance in the glucose values of the control versus Sur1-/- animals in the fed state (p > 0.05 for +/+ versus -/- males and for +/+ versus -/- females, n = 21, 20, 15, and 16, respectively; the average value was 6.53 ± 0.99 mM, n = 72). B, the insulin/glucose ratios of the Sur1-/- animals were higher than the controls, although only the difference for males was significant (0.33 ± 0.13 versus 0.146 ± 0.07 ng/ml/mM, p < 0.001, n = 20 or 19). The insulin/glucose ratios for randomly feeding animals were not significantly different (p > 0.05, for +/+ versus -/- males and for +/+ versus -/- females, n = 21, 20, 15, and 16, respectively; the average value was 0.172 ± 0.085 ng/ml/mM, n = 69). Error bars are ±S.D. Neonatal glucose and insulin levels were measured in control and Sur1-/- mice at 1 and 5 days of age and plotted against body weights. C, the 1- and 2-day blood glucose values show little correlation with body weight (r = -0.22, -0.31, and 0.048, n = 10, for 1-day control animals (open circles), 1-day Sur1-/- animals (open squares), and 2-day Sur1-/- animals (gray squares), respectively). The mean 1-day blood glucose values for the control (3.73 ± 0.53 mM) and Sur1-/- pups (1.73 ± 0.49 mM) are both hypoglycemic relative to the mean adult level (6.53 ± 0.99 mM), but the Sur1-/- pups have lower glucose values, normalized against body weight, than controls (1.12 ± 0.32 versus 2.56 ± 0.44 mM/gm, n = 10, p < 0.001; see inset). The 5-day values for the Sur1-/- pups (filled squares) exhibit a strong correlation with body weight (r = 0.91, n = 12), whereas the control values (filled circles) exhibit only a weak correlation (p = 0.55, n = 14). Comparison of the blood glucose values normalized against body weight (inset) show that the hypoglycemia of the Sur1-/- pups resolves within 36-48 h, and by 5 days, the Sur1-/- pups are significantly hyperglycemic relative to controls (2.11 ± 0.28, n = 12 versus 2.82 ± 0.25 mM/gm, p < 0.001, n = 14). D, comparison of the 1-day plasma insulin values shows that the Sur1-/- pups are hyperinsulinemic relative to control animals (3.62 ± 2.16 versus 1.58 ± 1.28 ng/ml, p < 0.01, n = 26 and 10, respectively). Within 5 days, the plasma insulin values of the Sur1-/- animals (0.72 ± 0.28 ng/ml) were lower than the controls (1.29 ± 0.88 ng/ml), but the difference is not significant at the p = 0.05 level (n = 12 and 14, respectively). The inset compares the plasma insulin/blood glucose ratios (see text). The error bars in the insets are ±S.D.

Sur1^-/- Animals Exhibit Transient Neonatal Hypoglycemia-- Loss of -cell K_ATP channel activity and appearance of spontaneous Ca²⁺ action potentials in PHHI patients is associated with unregulated insulin release (20) and severe neonatal hypoglycemia (21). The clinical phenotype of PHHI patients strongly suggested that the Sur1^-/- mice would show severe, persistent hypoglycemia. This was not the case. The Sur1^-/- pups were hypoglycemic 1 day after birth (1.73 ± 0.49 mM), and their insulin/blood glucose ratios were inappropriately high versus control values (2.31 ± 1.56 versus 0.42 ± 0.33 ng/ml/mM, p < 0.001), but this hypoglycemia reversed during the second day, and the insulin/glucose ratio reached a normal value (Fig. 6, C and D).

The blood glucose values of the 5-day Sur1^-/- pups display a correlation with body weight that was not apparent in the controls, and the 5-day Sur1^-/- pups were significantly hyperglycemic versus controls (8.29 ± 1.71 versus 6.23 ± 0.47 mM, p < 0.001). The Sur1^-/- insulin/blood glucose ratio was half that of the controls, although the difference is not significant (0.09 ± 0.04 versus 0.20 ± 0.13 ng/ml/mM, p > 0.05). This hyperglycemia in the 5-day Sur1^-/- pups is consistent with the reduced insulin output of their islets at high glucose concentrations (Fig. 4).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The interaction of SUR and K_IR6.x subunits is tightly integrated with both required to form K_ATP channels (1, 2). On the other hand, K⁺ currents have been reported when K_IR6.2 subunits are overexpressed using strong promoters (22, 23) or when the endogenous ER retention signals described by Zerangue et al. (24) are removed (25-27). Although homomeric K_IR6.2 channels have the same conductance as wild type K_ATP channels and exhibit low sensitivity to inhibitory ATP, they differ in all other respects. We saw no indication that K_IR6.2 subunits reach the cell surface without SUR1, and we expected that their presence in the plasma membrane, unregulated by SUR1, would result in hyperpolarization, reduced insulin secretion, and profound hyperglycemia. There was no indication that other K⁺ channels arise to compensate for the loss of SUR1/K_IR6.2 channels.

The loss of K_ATP channel activity leads to a specific -cell "electrophysiological" phenotype now described in three systems: human PHHI (21, 28), K_IR6.2^-/- (7), and Sur1^-/- -cells. This phenotype is characterized 1) by the loss of K⁺ currents activated by reduced intracellular nucleotides (shown for all three systems), 2) by an elevated membrane potential (also shown for all three systems), and 3) by spontaneous Ca²⁺ action potentials (shown for PHHI and Sur1^-/- -cells) that increase [Ca²⁺]_i (shown for PHHI and K_IR6.2^-/- -cells). This K_ATP^-/- -cell phenotype agrees with the "classical ionic mechanism," or K_ATP-dependent pathway, proposed to rapidly control glucose-stimulated insulin secretion (4, 29).

The loss of K_ATP channel activity produces remarkably different alterations in glucose homeostasis in PHHI neonates versus K_ATP^-/- mice. Although PHHI is a heterogeneous disorder (1, 30), the complexity of which is increased further by genetic imprinting (31, 32), loss of function mutations in both Sur1 and K_IR6.2 have been identified in homozygous individuals whose hyperinsulinemia requires near total pancreatectomy to control the resulting hypoglycemia (1). Although insulin secretory data from isolated PHHI -cells is scanty, the available results imply continuous secretion at a high rate regardless of glucose concentration (20).

K_ATP^-/- mice, both K_IR6.2^-/- (7) and Sur1^-/-, exhibit a far less dramatic phenotype. Neither K_IR6.2^-/- mice nor Sur1^-/- mice are persistently hypoglycemic as neonates, and as adults, both groups are normoglycemic under fed conditions. The two phenotypes differ subtly. The K_IR6.2^-/- mice are not significantly hypoglycemic versus controls following a 16-h fast, whereas the Sur1^-/- mice are (Fig. 6). The two groups exhibit a similar mild glucose intolerance. Fasting plasma insulin values in the K_IR6.2^-/- and Sur1^-/- animals are not significantly different from those of fasting control animals. The plasma insulin values increase to control levels when Sur1^-/- animals are fed, indicating some mechanism for regulating insulin release in the absence of K_ATP channels; plasma insulin data have not been reported for fed K_IR6.2^-/- mice. In addition to being unable to properly regulate their insulin output when blood glucose falls during a fast, the Sur1^-/- animals cannot properly regulate when blood glucose levels rise. This results in a peculiar plasma glucose profile, with the Sur1^-/- animals being significantly more hyperglycemic when glucose-loaded, e.g. the 5-day neonatal Sur1^-/- pups, and significantly more hypoglycemic when fasted, relative to control animals. The two K_ATP^-/- animals differ markedly in their sensitivity to insulin. Although we observed no difference in insulin sensitivity (Fig. 5C), Miki et al. (7) report that the K_IR6.2^-/- animals are insulin-hypersensitive, suggest that this is secondary to loss of SUR2A/K_IR6.2 ATP-sensitive K⁺ channels in skeletal muscle, and propose that the hypersensitivity accounts for the normoglycemia. This proposal suggests that the plasma insulin values of the K_IR6.2^-/- animals should be significantly lower than controls; otherwise, they should be hypoglycemic. The availability of Sur2A^-/- animals should allow a test of this hypothesis.

The secretion profiles of isolated K_IR6.2^-/- and Sur1^-/- islets also differ. The Sur1^-/- islets show no first phase insulin secretion in response to a shift from 2.8 to 16.7 mM glucose, whereas the K_IR6.2^-/- animals do show some first phase release, although it is small (Fig. 3B in Miki et al. (7)). By contrast, the K_IR6.2^-/- islets exhibit no second phase release, whereas the Sur1^-/- islets show a second phase response, although it is attenuated, measuring only 35% of the control value (Fig. 3A). In extended perifusion experiments, the Sur1^-/- and control islets exhibit a parallel increase in insulin output. Extended static incubation experiments demonstrate that insulin output from both control and Sur1^-/- islets is glucose-dependent and suppressible by the Ca²⁺ channel blocker nifedipine (Fig. 4). Furthermore, the insulin output of the Sur1^-/- islets is quite similar to control islets in the range of blood glucose values determined in randomly fed animals (hatched bar in Fig. 4). The reason(s) for the attenuated insulin output from the Sur1^-/- islets is unclear, it but could reflect a role for SUR1 apart from K_ATP channels (33).

In the case of Sur1^-/- mice, the loss of K_ATP channel activity reveals an underlying control of what would commonly be considered the 2nd phase of insulin secretion by a K_ATP-independent mechanism or pathway. Several K_ATP-independent pathways have been identified previously by pharmacologic means. One method used the potassium channel opener diazoxide to hold K_ATP channels in an open state, while increasing the external K⁺ concentration to depolarize the -cell membrane, activate voltage-gated Ca²⁺ channels, and increase [Ca²⁺]_i (10, 34). A second method, analogous to the K_ATP^-/- mouse models, used sulfonylureas to block K_ATP channels (8, 35). Again, [Ca²⁺]_i was elevated secondary to blocking K_ATP channels. With these strategies, it was possible to show that glucose metabolism augmented Ca²⁺-stimulated insulin release. A third route required simultaneous activation of protein kinases A and C and did not require an increase in [Ca²⁺]_i but was GTP-dependent (36-39). The available data are not sufficient to determine which pathway, or perhaps a novel pathway, is responsible for the regulation of insulin release in Sur1^-/- mice. Suppression of insulin release from Sur1^-/- islets by nifedipine is consistent with a requirement for Ca²⁺ influx, suggesting that a Ca²⁺-dependent pathway is involved. On the other hand, as pointed out during the review of this paper, insulin release via the Ca²⁺-dependent, K_ATP-independent pathway may be more rapid and of greater magnitude than observed from Sur1^-/- islets. We have been unable to find data on the response time of the Ca²⁺-dependent, K_ATP-independent pathway after lowering glucose for comparison with the slow response of Sur1^-/- islets.

The results of our whole animal and isolated islet experiments are summarized and integrated in Fig. 7 in the form of a stimulus-response experiment, the stimulus being a step up, then down, in glucose concentration. During the "up-shift" in glucose metabolism, wild type islets respond by releasing a bolus of insulin constituting first phase insulin release. Loss of this first phase release in the Sur1^-/- animals causes their mild glucose intolerance because they are unable to quickly increase the basal rate of glucose disposal via an insulin-dependent mechanism (see Fig. 5). It will be of interest to determine whether the loss of first phase insulin release will lead to type II diabetes as the Sur1^-/- animals age, as is seen in non-insulin-dependent diabetes mellitus (40). Continued glucose infusion is expected to result in higher "steady-state" glucose levels in Sur1^-/- animals because their insulin output is reduced at higher blood glucose concentrations relative to controls (see Figs. 3, 4, and 5B). We observe hyperglycemia in well fed 5-day-old Sur1^-/- mice because they are unable to secrete sufficient insulin to increase the rate of glucose disposal. Finally, without K_ATP channels, Sur1^-/- -cells cannot quickly reduce their insulin output when blood glucose levels fall. We suggest this impaired ability to restrict insulin output, a consequence of being unable to repolarize their -cells, accounts for the increased hypoglycemia that we observed in fasting adult animals and may play a similar role in the 1-day-neonatal pups.

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Generalized blood glucose and insulin responses to up- and down-shifts in glucose concentrations. The top panel illustrates hypothetical changes in glucose concentration either as a result of glucose infusion into mice or through a change in perifusion medium. The middle panel shows the resulting insulin output from control (solid line) or Sur1-/- islets (dotted line) in response to the up- and down-shifts in the upper panel. The bottom panel illustrates the proposed changes in blood glucose levels that would be observed in response to the changes in insulin output. Loss of first phase secretion in the Sur1-/- animals results in higher glucose levels and slower clearance of a glucose load, whereas the reduced insulin output from Sur1-/- islets results in a higher steady-state glucose level. The impaired ability of the Sur1-/- islets to restrict insulin output during a fall in glucose concentration leads to more pronounced hypoglycemia in the knockout animals (for example, during fasting).

A key question is why the glucose levels in PHHI patients are persistently low, whereas the K_ATP^-/- mice are normoglycemic. There is a significant body of literature on differences in insulin secretory responses between mouse and rat (and human) -cells. When rat islets are shifted to high glucose, they show a first phase of insulin secretion followed by a larger, rising second phase. Mouse islets exhibit an equivalent first phase peak followed by an elevated plateau, but no rising second phase. This differential behavior has been described by a number of authors, including Lenzen (41), Berglund (42), Ma et al. (43), and Zawalich et al. (44). Our working hypothesis to explain the species differences is that K_ATP^-/- -cells are continuously in the second phase of insulin secretion. Insulin output is greater for PHHI -cells than K_ATP^-/- mouse -cells, consistent with normal human -cells, which, like rat -cells, have a large rising second phase of insulin secretion. This hypothesis requires that basal insulin output from PHHI -cells at low concentrations of glucose should be elevated, in contrast to the Sur1^-/- -cells, the basal output of which is comparable to controls. The Sur1^-/- mice offer a potential background for testing this hypothesis by engineering transgenic animals to search for component(s) that contribute to the large rising second phase of insulin secretion in rats and humans and for compounds that modulate second phase insulin release.

	ACKNOWLEDGEMENTS

We acknowledge the technical assistance of Li-Zhen Song, Maria Shlyapobersky, Gabriela Gonzalez, and Jie Wang; the help of the staff at the Center for Comparative Medicine; and the laboratory of Dr. Tim McDonnell for doing terminal deoxynucleotidyl transfer-mediated nick end labeling assays.

	FOOTNOTES

^* Supported by National Institutes of Health Grant DK 50750 (to J. B.).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.

^§ These authors contributed equally to this work.

To whom correspondence should be addressed: Dept. of Molecular and Cellular Biology, Rm. 108C, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030. Tel.: 713-798-4007; Fax: 713-790-0545; E-mail: jbryan@bcm.tmc.edu.

	ABBREVIATIONS

The abbreviations used are: K_ATP channels, ATP-sensitive K⁺ channels; SUR, sulfonylurea receptor; PHHI, persistent hyperinsulinemic hypoglycemia of infancy; K_IR, potassium inward rectifier; KRBB, Krebs-Ringer bicarbonate buffer.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				D. D. De Leon, C. Li, M. I. Delson, F. M. Matschinsky, C. A. Stanley, and D. A. Stoffers Exendin-(9-39) Corrects Fasting Hypoglycemia in SUR-1-/- Mice by Lowering cAMP in Pancreatic {beta}-Cells and Inhibiting Insulin Secretion J. Biol. Chem., September 19, 2008; 283(38): 25786 - 25793. [Abstract] [Full Text] [PDF]

				M. Abdulhadi-Atwan, J. D. Bushman, S. Tornovsky-Babaey, A. Perry, A. Abu-Libdeh, B. Glaser, S.-L. Shyng, and D. H. Zangen Novel De Novo Mutation in Sulfonylurea Receptor 1 Presenting as Hyperinsulinism in Infancy Followed by Overt Diabetes in Early Adolescence Diabetes, July 1, 2008; 57(7): 1935 - 1940. [Abstract] [Full Text] [PDF]

				C. Li, I. Nissim, P. Chen, C. Buettger, H. Najafi, Y. Daikhin, I. Nissim, H. W. Collins, M. Yudkoff, C. A. Stanley, et al. Elimination of KATP Channels in Mouse Islets Results in Elevated [U-13C]Glucose Metabolism, Glutaminolysis, and Pyruvate Cycling but a Decreased {gamma}-Aminobutyric Acid Shunt J. Biol. Chem., June 20, 2008; 283(25): 17238 - 17249. [Abstract] [Full Text] [PDF]

				L. Aguilar-Bryan and J. Bryan Neonatal Diabetes Mellitus Endocr. Rev., May 1, 2008; 29(3): 265 - 291. [Abstract] [Full Text] [PDF]

C. G. Nichols Alchemy in the Soup: Transforming Metabolic Signals to Excitability Sci. Signal., October 30, 2007; 2007(410): pe59 - pe59. [Abstract] [Full Text] [PDF]