Original Article -ATP-Sensitive K؉ Channel Signaling in Deficient Maria S. Remedi, Joseph C. Koster, Brian L. Patton, and Colin G. Nichols

As the rate-limiting controller of metabolism, glu- gous inactivating GK mutations cause permanent neonatal cokinase represents the primary ␤-cell “glucose sensor.” diabetes in humans (12,13), which is characterized by Inactivation of both glucokinase (GK) alleles results in perinatal and low birth weight (14). The permanent neonatal diabetes; inactivation of a single allele disease usually requires treatment through the causes maturity-onset diabetes of the young type 2 (MODY- .؊/؊ patient’s lifetime, starting within the first month of life 2). Similarly, mice lacking both alleles (GK ) exhibit Targeted disruption of the neuroendocrine knockout of severe neonatal diabetes and die within a week, whereas ؉/؊ GK (nGK) gene in mice also causes diabetes (15). These heterozygous GK mice exhibit markedly impaired glu- ␤ cose tolerance and diabetes, resembling MODY-2. Glucose mice lack glucokinase in pancreatic -cells and neurons metabolism increases the cytosolic [ATP]-to-[ADP] ratio, but maintain normal liver glucokinase activity. Heterozy- ؉ which closes ATP-sensitive K channels (KATP channels), gous null mice show early-onset mild diabetes caused by -leading to membrane depolarization, Ca2؉ entry, and insu- an impaired insulin-secretory response to glucose, resem lin exocytosis. Glucokinase insufficiency causes defective bling MODY-2. Importantly, homozygous nGK-deficient KATP channel regulation, which may underlie the impaired mice show severe perinatal diabetes, as in permanent secretion. To test this prediction, we crossed mice lacking .(؉/؊ neonatal diabetes, and die within a few days of birth (15 neuroendocrine glucokinase (nGK ) with mice lacking ϩ -؊/؊ ATP-sensitive K channels (KATP channels) couple glu KATP channels (Kir6.2 ). Kir6.2 knockout rescues peri- -natal lethality of nGK؊/؊, although nGK؊/؊Kir6.2؊/؊ ani- cose metabolism to cellular electrical activity and there mals are postnatally diabetic and still die prematurely. fore play a critical role in excitation-secretion coupling in -nGK؉/؊ animals are diabetic on the Kir6.2؉/؉ background the pancreatic ␤-cell. Glucose metabolism raises the cyto but only mildly glucose intolerant on the Kir6.2؊/؊ back- solic [ATP]-to-[ADP] ratio, which causes closure of the ؉ ؊ ground. In the presence of glutamine, isolated nGK / K channel and depolarization of the ␤-cell membrane. Kir6.2؊/؊ islets show improved insulin secretion compared ATP ؉/؊ ؉/؉ Membrane depolarization, in turn, leads to opening of with nGK Kir6.2 . The significant abrogation of 2ϩ -؊/؊ ؉/؊ voltage-dependent Ca channels and a rise in intracellu nGK and nGK phenotypes in the absence of KATP 2ϩ demonstrate that a major factor in glucokinase deficiency lar [Ca ], which triggers insulin vesicle exocytosis (16) (Fig. 1). Glucose metabolism has additional downstream is indeed altered KATP signaling. The results have implica- tions for understanding and therapy of glucokinase-related actions, including generation of GTP, which may activate diabetes. Diabetes 54:2925–2931, 2005 “KATP-independent” pathways of insulin secretion. Na- ively, both electrical triggering of secretion through KATP channels and nonelectrical effects of glucose metabolism through ATP generation could be affected by glucokinase lucose metabolism in pancreatic ␤-cells is nec- deficiency. The relative importance of each component is essary to stimulate insulin secretion (1). Glu- not clear, and it might be presumed that the consequences cokinase, which phosphorylates glucose in the of failure to generate ATP should be much more devastat- Gfirst, rate-limiting reaction, is a key component ing than simply blocking electrical signaling. in the secretory pathway (2,3), and mutations of the GK It has previously been shown that glucose fails to close Ϫ/Ϫ ␤ gene are diabetogenic in humans (4). Heterozygous inac- KATP channels in nGK -cells (17). Sulfonylureas act by tivating mutations cause maturity-onset diabetes of the closing the KATP channel independently of glucose metab- young type 2 (MODY-2), an autosomal, dominantly inher- olism, resulting in membrane depolarization and conse- ited form of diabetes characterized by an early age of quent stimulation of insulin secretion (18). Consistent with onset and pancreatic ␤-cell dysfunction (5–11). Homozy- the idea that the primary downstream consequence of nGK deficiency is simply failure to close KATP channels, the sulfonylurea tolbutamide was effective at both inhibiting From the Department of Cell Biology and Physiology, Washington University Ϫ/Ϫ School of Medicine, St. Louis, Missouri. KATP channels and stimulating insulin secretion in nGK Address correspondence and reprint requests to Colin G. Nichols, the ␤-cells (17). These results imply that the defect in insulin Department of Cell Biology and Physiology, Washington University School of secretion in glucokinase deficiency results essentially Medicine, 660 S. Euclid Ave., St. Louis, MO 63110. E-mail: cnichols@cellbio. wustl.edu. from the defective regulation of KATP channels, rather than Received for publication 3 May 2005 and accepted in revised form 7 July from nonelectrical consequences of altered metabolism. 2005. Further indirect support for this hypothesis comes from: DMEM, Dulbecco’s modified Eagle medium; GSIS, glucose-stimulated insu- ϩ 1) the demonstration that mice with ATP-insensitive ␤-cell lin secretion; KATP channel, ATP-sensitive K channel; MODY-2, maturity- onset diabetes of the young type 2. KATP channels also die from neonatal diabetes (19) and 2) © 2005 by the American Diabetes Association. the recent dramatic demonstration that gain-of-function 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 KATP mutations also cause permanent neonatal diabetes in with 18 U.S.C. Section 1734 solely to indicate this fact. humans (20–24). A critical prediction is that the conse-

DIABETES, VOL. 54, OCTOBER 2005 2925 KATP AND GLUCOKINASE-DEFICIENCY DIABETES

Isolation of pancreatic islets and ␤-cells. Experiments using animals were performed in accordance with the relevant laws and institutional guidelines and were approved by the Washington University animal studies committee. Anesthetized mice were killed by cervical dislocation. Pancreata were re- moved and injected with Hank’s balanced salt solution containing type XI collagenase (0.5 mg/ml), pH 7.4. (Sigma, St. Louis, MO). Pancreata were digested for 7 min at 37°C then hand-shaken and washed three times in cold Hank’s solution (29). Islets were isolated by hand under a dissecting micro- scope and pooled. Islets were maintained overnight in CMRL-1640 (5.6 mmol/l glucose) supplemented with FCS (10%), penicillin (100 units/ml), and strep- tomycin (100 mg/ml). Blood glucose and insulin levels. Blood glucose was assayed using a glucose dehydrogenase–based enzymatic assay and quantitated using a Glu- cometer Elite XL meter (Bayer, Elkhart, IN). Blood insulin levels were assayed on 15 ␮l of mouse serum using an enzyme-linked immunosorbent assay kit with a rat insulin standard according to manufacturer’s procedure (Crystal Chem, Downers Grove, IL). Intraperitoneal glucose tolerance tests were performed on 5-week-old mice after a 16-h fast. Animals were injected intraperitoneally with glucose (1 g/kg). Blood was isolated from the tail vein at the times indicated and assayed for glucose as described above. Intraperi- toneal insulin tolerance tests were performed on 4-month-old mice after a 6-h fast. Animals were injected intraperitoneally with insulin (0.5 unit/kg). Blood FIG. 1. Glucokinase knockout can disrupt both electrical and nonelec- was isolated from the tail vein at the times indicated and assayed for glucose trical signaling. Glucose metabolism generates ATP, which stimulates as described above. insulin secretion through KATP-dependent (1) and KATP-independent Insulin release experiments. After overnight incubation, pancreatic islets ؊/؊ (2) pathways. In Kir6.2 islets, the KATP-dependent pathway (1) will (10 per well in 12-well plates) were preincubated in glucose-free Dulbecco’s be permanently “on,” and glucose-dependent regulation will then modified Eagle medium (DMEM) supplemented with 3 mmol/l of glucose for ,occur via K -independent (2) pathways. In nGK؊/؊Kir6.2؉/؉ islets ATP 2 h. The release assay was initiated by supplementing the DMEM glucose-free glucose metabolism is lost, and both pathways (1 and 2) will be “off.” ϩ ؊/؊ ؊/؊ media with D -glucose (1, 7, 16.7, or 23 mmol/l) as indicated. For amino acid In nGK Kir6.2 islets, the KATP-independent pathway (2) will be stimulation of insulin release, Krebs-Ringer buffer supplemented with D- “off,” but the KATP-dependent pathway (1) should remain permanently on.” PKC, protein kinase C; VDCC, voltage-dependent Ca2؉ channel. glucose and glutamine (10 mmol/l) was used in place of DMEM. Islets were“ incubated for 60 min at 37°C, and medium was removed and assayed for insulin content using a rat insulin radioimmunoassay (Linco, St. Charles, MO). quences of nGK deficiency should be abrogated in animals Each experiment was repeated in triplicate. that lack KATP channels. Mice lacking Kir6.2-dependent Statistics. Data are presented as means Ϯ SE. Differences among the four KATP channel activity in all tissues have been generated by genotypes were tested using ANOVA and post hoc Duncan’s test. A difference disruption of the pore-forming Kir6.2 subunit (25). K was defined as significant when P Ͻ 0.05. Nonsignificant differences are not Ϫ Ϫ ATP knockout mice (Kir6.2 / ) show a complex phenotype indicated. with transient hyperinsulinemia and as ne- onates, which rapidly progress to hypoinsulinemia but ϩ Ϫ euglycemia as adults (25–28). By crossing nGK / mice RESULTS Ϫ/Ϫ with Kir6.2 (KATP channel knockout) mice, we can test Neonatal survival of nGK knockout mice on the the critical prediction and separate electrical and metabolic Kir6.2 knockout background. Of 57 live births from consequences of nGK deletion. The results provide a striking cross-breeding of heterozygous nGKϩ/Ϫ mice on the Kir6.2 demonstration of the importance of the electrical signal in wild-type background, only 2 mice were positively geno- transducing glucose metabolism to insulin secretion. typed as nGKϪ/Ϫ (Fig. 2A). Both were found dead within the 1st week (genotyped postmortem) after birth (Fig. 2B). RESEARCH DESIGN AND METHODS Conversely, on the Kir6.2 knockout background, nGKϪ/Ϫ Previously, a global neuroendocrine knockout of GK (nGK) was achieved by pups were born at the expected ratio (1:2:1 for nGKϩ/ removal of upstream promoter elements (15). This was expected to result in Ϫ Ϫ ϩ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ ϩ / / / / / abolition of GK expression in ␤-cells and also in gut and brain (15). Knockout Kir6.2 , nGK Kir6.2 , and nGK Kir6.2 ,re- of the Kir6.2 gene was achieved by targeted introduction of the neomycin- spectively) (Fig. 2A). These double-knockout mice Ϫ/Ϫ resistant gene into the coding region of Kir6.2 (25). Heterozygous nGKϩ/Ϫ survived much longer than nGK on the Kir6.2 wild-type mice on a C57/BL6J background (gift from Dr. Yasuo Terauchi) (15) were bred background, with a median survival time of ϳ10 days, and to Kir6.2Ϫ/Ϫ on an outbred background (gift from Dr. Susumo Seino) (25) to ϳ30% surviving past weaning (21 days) (Fig. 2B). As is generate double-heterozygous F1 mice. F1 mice were backcrossed to each Ϫ/Ϫ Ϫ/Ϫ Ϫ/Ϫ ϩ/ϩ shown in Fig. 2C, double-knockout nGK Kir6.2 mice other to generate F2 mice of expected genotype. nGK Kir6.2 and ϩ/Ϫ Ϫ/Ϫ Ϫ/Ϫ ϩ/Ϫ were significantly smaller than both nGK Kir6.2 and nGK Kir6.2 were not identified in the litters, probably because of ϩ/ϩ Ϫ/Ϫ perinatal lethality. For all experiments described, we backcrossed F2 het- nGK Kir6.2 mice at 4 weeks of age. By 5 weeks of ϩ Ϫ ϩ ϩ Ϫ Ϫ Ϫ/Ϫ Ϫ/Ϫ erozygous nGK / mice (either on Kir6.2 / or Kir6.2 / backgrounds) to the age, nGK Kir6.2 mice weighed ϳ50% (4.78 Ϯ 0.17g) same genotypes, thus generating F3 litters of three genotypes (nGKϩ/ϩ, that of nGKϩ/ϩKir6.2Ϫ/Ϫ (8.34 Ϯ 0.24g) or nGKϩ/Ϫ ϩ Ϫ Ϫ Ϫ ϩ ϩ Ϫ Ϫ nGK / , and nGK / on Kir6.2 / or Kir6.2 / backgrounds). Mice were typed Kir6.2Ϫ/Ϫ (9.02 Ϯ 0.17g.) littermates. Nevertheless, it is for nGK and Kir6.2 genes as previously described by Terauchi et al. (15) and striking that these double-knockout mice, which com- Miki et al. (25). ␤ Previously, Kir6.2[AAA] mice expressing a dominant-negative Kir6.2 muta- pletely lack -cell glucokinase activity, can survive beyond tion (132GFG1343AAA) mutation under control of the rat insulin 1 promoter weaning. Ϫ/Ϫ were generated (29). For experiments shown in Fig. 7, these Kir6.2[AAA] mice On this Kir6.2 background, fasting blood glucose was (on C57BL6J background) were bred to nGKϩ/Ϫ mice, also on C57BL6J only slightly higher in nGKϪ/Ϫ than in nGKϩ/ϩ or nGKϩ/Ϫ background, to generate the relevant crosses. mice (Fig. 3A). However, fed glucose (Fig. 3B) was dra- To control for effects of genetic background, comparisons between geno- matically higher in nGKϪ/Ϫ mice, suggesting that these types were made with littermates in every experiment, and n values of at least three are reported in each paired genotype. Thus, it is reasonable to conclude mice either still do not secrete enough insulin to lower that any effects of nonmutated chromosomes will be randomized within all blood glucose values or that they develop insulin resis- genotypes. tance (see below).

2926 DIABETES, VOL. 54, OCTOBER 2005 M.S. REMEDI AND ASSOCIATES

FIG. 3. High fasting glucose in neuroendocrine glucokinase-defective mice. A: Fasting glucose levels from 6-week-old mice. No nGK؊/؊ Kir6.2؉/؉ animals survived to this age. B: Fed glucose levels from 6-week-old mice. In nGK؊/؊Kir6.2؊/؊ mice, glucose ranged from 23 mmol/l to >33 mmol/l (i.e., above the limit of detection). C: Plasma insulin under fed conditions showed no significant differences between .mice. *P < 0.05 8–4 ؍ groups. Data are the means ؎ SE, n ϩ Ϫ FIG. 2. Kir6.2؊/؊ rescues the nGK؊/؊ phenotype from perinatal lethal- heterozygous nGK / mice show nonprogressive mild ity. A: Number of nGK mice (F3) born of each genotype on the Kir6.2؉/؉ ؊/؊ ؉/؉ diabetes, as in human MODY-2. At birth, the blood glucose or Kir6.2 backgrounds. Expected ratios are 1:2:1 for nGK , ϩ/ϩ ϩ/Ϫ nGK؉/؊, and nGK؊/؊, respectively. Only 2 mice from 57 births were levels of nGK and nGK mice were similar (3.5 Ϯ 0.4 ,typed nGK؊/؊ on Kir6.2 wild-type background (expected ϳ14). A total and 3.9 Ϯ 0.2 mmol/l, respectively), but by 12 weeks of 10 mice (expected ϳ10) from 39 births were typed nGK؊/؊ on ϩ/Ϫ ϩ/ϩ ؊/؊ nGK Kir6.2 mice showed a twofold increase in both Kir6.2-null background. Unk, unknown genotype. B: Survival of nGK ϩ/ϩ ؊/؊ fasting and fed blood glucose compared with nGK mice. Two nGK mice that were born on the Kir6.2 wild-type back- ϩ/ϩ .(ground died within 5 days. Mean survival time for nGK؊/؊Kir6.2؊/؊ was Kir6.2 mice (Fig. 3A and B ϳ10 days, and ϳ30% survived past weaning (21 days). C: Littermate ؊/؊ ؊/؊ ؉/؊ ؊/؊ By contrast, on the Kir6.2 knockout background, overt nGK Kir6.2 and nGK Kir6.2 mice at 4 weeks of age. diabetes, as evidenced by elevated fasting glucose levels, is ameliorated in nGKϩ/ϪKir6.2Ϫ/Ϫ mice, although fed ,nGK؉/؊ animals are diabetic on the Kir6.2؉/؉ back- glucose levels are elevated (Fig. 3A and B). Plasma insulin ground but not on the Kir6.2؊/؊ background. Het- measured under fed conditions, was slightly elevated in erozygous nGKϩ/Ϫ mice were born at close to expected nGKϩ/Ϫ mice versus nGKϩ/ϩ on both Kir6.2ϩ/ϩ and ratios on both the Kir6.2ϩ/ϩ and Kir6.2Ϫ/Ϫ backgrounds Kir6.2Ϫ/Ϫ backgrounds (Fig. 3C). (Fig. 2A) and developed apparently normally. As previ- Enhancement in glucose tolerance of heterozygous ϩ/ϩ ously reported, on the wild-type Kir6.2 background, nGK-deficient mice in the absence of KATP channel

DIABETES, VOL. 54, OCTOBER 2005 2927 KATP AND GLUCOKINASE-DEFICIENCY DIABETES

mice (Fig. 5), similar to a previous report (25). Insulin sensitivity was significantly enhanced in nGKϩ/Ϫ mice on the Kir6.2ϩ/ϩ background, whereas nGKϩ/ϪKir6.2Ϫ/Ϫ mice showed intermediate sensitivity (Fig. 5). These results exclude enhanced insulin sensitivity as a possible mech- anism for the improved glucose tolerance of nGKϩ/Ϫ mice in the absence of KATP, and they point to enhanced insulin secretion as the more likely mechanism (see below). Absence of KATP channel activity may restore insulin -secretion from nGK؊/؊ pancreatic islets. To gain fur ther insight into the cellular basis for differential pheno- types of double-knockout animals, islets were isolated from each of the different genotypes and phenotypically examined. Morphologically, islet appearance and size were not different between the various genotypes (for example, islet diameter was 175.5 Ϯ 8.1 ␮m for nGKϩ/Ϫ Kir6.2ϩ/ϩ and 183.3 Ϯ 7.3 ␮m for nGKϩ/ϪKir6.2Ϫ/Ϫ, n ϭ 35 islets). Glucose-stimulated insulin secretion (GSIS) was

.؉ ؊ assessed in isolated islets from all five available genotypes FIG. 4. Kir6.2 knockout abrogates nGK / diabetes. Blood glucose As shown previously, the response of nGKϩ/Ϫ islets was concentration versus time after injection of 1 g/kg glucose. Glucose ϩ ϩ ϩ ϩ / tolerance tests on nGK؉/؉ and nGK؉/؊ animals on the Kir6.2 wild-type mildly reduced compared with nGK / , on the Kir6.2 mice. background (15). On the Kir6.2Ϫ/Ϫ background, GSIS was 9–3 ؍ and Kir6.2 knockout backgrounds. Data are means ؎ SE, n P < 0.05 for nGK؉/؉Kir6.2؉/؉ vs. all other genotypes; †P < 0.05* .for nGK؉/؉Kir6.2؉/؉ vs. nGK؉/؊Kir6.2؉/؉ and nGK؉/؊Kir6.2؊/؊ mice; greatly reduced in the presence or absence of nGK (Fig P < 0.05 for nGK؉/؉Kir6.2؉/؉ vs. nGK؉/؊Kir6.2؉/؉;‡P < 0.05 for 6A), and there was no enhancement of release from** nGK؉/؊Kir6.2؉/؉ vs. nGK؉/؊Kir6.2؊/؊. nGKϩ/Ϫ or nGKϪ/Ϫ, relative to the release from the same nGK genotypes on the Kir6.2ϩ/ϩ background. These data activity. To characterize physiological responses to glu- thus reiterate the perplexing result that Kir6.2Ϫ/Ϫ (25) and cose load, surviving mice underwent intraperitoneal glu- SUR1Ϫ/Ϫ (26) mice are mildly glucose tolerant, yet have cose tolerance testing at 5 weeks of age. As previously only minimal GSIS. Moreover, these results do not help to reported, blood glucose before and after glucose load was explain the enhanced glucose tolerance and survival of ϩ/Ϫ ϩ/ϩ significantly higher in nGK mice on the Kir6.2 nGK-deficient mice on this Kir6.2Ϫ/Ϫ background. Recent Ϫ/Ϫ background, and Kir6.2 mice showed a very mild studies have begun to shed light on the potential mecha- ϩ/ϩ impairment in glucose tolerance with respect to Kir6.2 nisms in SUR1Ϫ/Ϫ mice (30–32). In particular, glutamine mice (Fig. 4). Importantly, on the Kir6.2Ϫ/Ϫ background, has been shown to preferentially stimulate KATP-indepen- there was dramatically improved glucose tolerance of dent secretion from SUR1Ϫ/Ϫ islets (31). This effect has ϩ/Ϫ nGK mice (Fig. 4), to the extent that they were not not been examined in Kir6.2Ϫ/Ϫ islets, but if the pheno- ϩ/ϩ significantly different from nGK mice on the same types of SUR1Ϫ/Ϫ and Kir6.2Ϫ/Ϫ indeed both result from background. loss of KATP, we would expect the same KATP-independent The glucose-lowering effect of intraperitoneal insulin stimulation of Kir6.2Ϫ/Ϫ islets. As shown in Fig. 6B, injection (0.5 unit/kg) was assessed using insulin tolerance glutamine markedly stimulates secretion from both ϩ/ϩ Ϫ/Ϫ tests on 6-h fasted mice. nGK Kir6.2 mice were nGKϩ/ϩ and nGKϩ/Ϫ islets on the Kir6.2Ϫ/Ϫ background ϩ/ϩ ϩ/ϩ slightly more sensitive to insulin than nGK Kir6.2 relative to the wild-type background. Given likely physio- logical levels of glutamine, this may well cause enhanced insulin secretion in vivo and explain the improved glucose tolerance of nGKϩ/Ϫ on the Kir6.2Ϫ/Ϫ background (31) (see below). ␤ -Cell–specific reduction of KATP channel activity enhances glucose tolerance of nGK؉/؊ mice. Although the above data are consistent with a specific loss of KATP channels in ␤-cells being responsible for abrogation of the nGK-deficiency phenotype, it is a caveat that the Kir6.2 knockout is global. We previously generated ␤-cell–spe- cific dominant-negative Kir6.2[AAA] mice using an insulin promoter to drive transgene expression (29). Kir6.2[AAA] ϳ ␤ mice, which lack KATP in 70% of -cells, hypersecrete insulin and have slightly enhanced glucose tolerance (29). In the current study, we bred these mice with nGKϩ/Ϫ Kir6.2ϩ/ϩ mice to generate double-heterozygous (nGKϩ/Ϫ ؊ ؊ FIG. 5. Kir6.2 / does not significantly increase insulin sensitivity from RESEARCH DESIGN AND METHODS ؉ ؊ Kir6.2[AAA]) mice (see ). As nGK / . Intraperitoneal insulin tolerance test results from the four shown in Fig. 7, these mice have a marked improvement in genotypes on 4-month-old mice. Blood glucose measurements versus ϩ/Ϫ time after intraperitoneal bolus insulin injection (0.5 unit/kg body wt) glucose tolerance compared to nGK alone. Although -mice. *P < 0.05 for the improvement is not as dramatic as in the mice com 9–6 ؍ after6hoffasting. Data are means ؎ SE, n ؉ ؉ ؉ ؉ ؉ ؉ ؉ ؉ nGK / Kir6.2 / vs. all other genotypes; †P < 0.05 for nGK / Kir6.2 / pletely lacking K (Fig. 4), this result is consistent with vs. nGK؉/؊Kir6.2؉/؉ and nGK؉/؊Kir6.2؊/؊ mice; **P < 0.05 for nGK؉/؉ ATP ␤ ؉ ؉ ؉ ؊ ؉ ؉ ؉ ؊ ؉ ؉ Kir6.2 / vs. nGK / Kir6.2 / ;‡P < 0.05 for nGK / Kir6.2 / vs. -cell KATP channels indeed being the major players in .nGK؉/؊Kir6.2؊/؊. rescuing the nGK-deficiency phenotype

2928 DIABETES, VOL. 54, OCTOBER 2005 M.S. REMEDI AND ASSOCIATES

FIG. 7. nGK؉/؊Kir6.2[AAA] mice have enhanced glucose tolerance -compared with nGK؉/؊Kir6.2؉/؉ diabetic mice. Blood glucose concen tration versus time after injection of 1.0 g/kg glucose. Glucose toler- ance tests on nGK؉/؉ and nGK؉/؊ animals on the Kir6.2 wild-type and ؎ Kir6.2[AAA] dominant-negative backgrounds. Data are the means -mice. *P < 0.05 for nGK؉/؉Kir6.2؉/؉ vs. all other geno 5–4 ؍ SE, n types; †P < 0.05 for nGK؉/؉Kir6.2؉/؉ vs. nGK؉/؊Kir6.2؉/؉ and nGK؉/؊ Kir6.2[AAA] mice; ‡P < 0.05 for nGK؉/؊Kir6.2؉/؉ vs. nGK؉/؊ Kir6.2[AAA].

the relative importance of electrical (KATP-dependent) versus nonelectrical (KATP-independent) consequences of neuroendocrine glucokinase deficiency remains unclear. KATP channels are under tight metabolic control, and any perturbation of metabolism that leads to decreased gener- FIG. 6. GSIS and glutamine-stimulated insulin secretion from nGK- ation of ATP is expected to have pronounced inhibitory deficient islets. A: GSIS from isolated islets of 8-week-old mice. The islets from all five genotypes were incubated for 60 min at 1, 7, 16.7, effects on excitation-secretion coupling because of en- ؍ ؎ and 23 mmol/l glucose (means SE, n 5–8 mice) in DMEM. GSIS is hanced KATP activity. Studies of the electrical activity of similarly suppressed in nGK؉/؉, nGK؉/؊, and nGK؊/؊ islets on Kir6.2؊/؊ ؉ ؉ ؉ ؉ nGK-deficient mice indeed support the idea that reduced background. *P < 0.05 for nGK / Kir6.2 / vs. all other genotypes at the same glucose concentration; **P < 0.05 for nGK؉/؉Kir6.2؉/؉ vs. glycolytic metabolism specifically causes failure of KATP nGK؉/؊Kir6.2؉/؉ at the same glucose concentration. B: Relative glu- channels to close in response to glucose, contributing to tamine-stimulated insulin release from isolated islets on 8-week-old the observed reduction in insulin secretion (17). Consis- mice. Islets were placed in batches of 10 in Krebs-Ringer buffer media supplemented with 1 mmol/l glucose alone or 1, 7, and 16.7 mmol/l tent with this conclusion, it was reported that oral sulfo- glucose with the addition of 10 mmol/l L-glutamine. Data are means ؎ nylurea treatment could suppress perinatal death in .mice. *P < 0.05 for nGK؉/؉Kir6.2؉/؉ or nGK؉/؊Kir6.2؉/؉ vs. nGKϪ/Ϫ mice (15), presumably by restoring insulin release 3 ؍ SE, n -nGK؉/؉Kir6.2؊/؊ and nGK؉/؊Kir6.2؊/؊ at the same glucose concentra tion. Glucokinase is critical for liver regulation of glucose production (34). This could contribute to the whole animal phenotype in complex ways, and for this reason, we DISCUSSION examined a tissue-specific model in which the liver-spe- KATP-dependent versus KATP-independent pathways cific isoform of glucokinase is expressed normally, and of insulin secretion: relative role in glucokinase- only the neuroendocrine-specific isoform is deleted (15). deficient diabetes. Heterozygous loss-of-function muta- We specifically examined the consequences of nGK defi- tions in the glucokinase (GK) gene are the most common ciency on the wild-type and KATP-null backgrounds to cause of MODY-2, a relatively mild form of . elucidate the relative importance of the two signaling Conversely, homozygous or compound heterozygous GK pathways. Perinatal lethality of nGKϪ/Ϫ was absolute on deficiency mutations underlie permanent neonatal diabe- the Kir6.2ϩ/ϩ background; only 2 pups were detected of an tes, a more severe perinatal diabetes (13). Mimicking expected ϳ16 pups, from heterozygous nGKϩ/ϪKir6.2ϩ/ϩ human permanent neonatal diabetes, homozygous neu- backcrosses (Fig. 2), and both died within 5 days. In roendocrine GK knockout mice die shortly after birth of contrast, double-knockout mice (nGKϪ/ϪKir6.2Ϫ/Ϫ) were severe diabetes (15,33). Consistent with a MODY-2 pheno- born at the expected ratio. They exhibited low body type, mice lacking one allele of neuroendocrine GK exhibit weight, severe diabetes, and died prematurely, but perina- mild hyperglycemia with normal basal glucose turnover tal lethality is clearly abrogated (Fig. 2A and B). Amelio- rates, but they have impaired glucose tolerance and insulin ration of the nGKϪ/Ϫ phenotype on the Kir6.2Ϫ/Ϫ secretion during hyperglycemia (2,15). background can be explained by an increased, though still Glucokinase deficiency could affect insulin secretion by insufficient, level of circulating insulin. This striking dem- altering electrical and/or nonelectrical downstream conse- onstration of the nullifying effect of removal of the KATP quences of glucose metabolism (Fig. 1). To date, however, “brake” in vivo is further supported by the results with

DIABETES, VOL. 54, OCTOBER 2005 2929 KATP AND GLUCOKINASE-DEFICIENCY DIABETES heterozygous nGKϩ/Ϫ mice. On the Kir6.2ϩ/ϩ background, glucokinase deficiency. Most patients with heterozygous they have markedly elevated fed and fasting glucose GK mutations do not require pharmacological treatment, levels, and they are markedly glucose intolerant. By con- and the majority of cases are managed with diet alone. trast, on the Kir6.2Ϫ/Ϫ background, nGKϩ/Ϫ mice have However, during pregnancy, women with GK mutations near-normal fasting glucose levels and similar glucose are often treated with insulin to maintain normal blood tolerance to nGKϩ/ϩKir6.2Ϫ/Ϫ animals. The abolition of sugar levels, resulting in babies that are large for gesta- KATP significantly ameliorates the “MODY-2” phenotype of tional age because of the effect of insulin on fetal growth the nGKϩ/Ϫ mice. (35). In these cases, sulfonylurea treatment could presum- Complex phenotypes of Kir6.2-deficient ؋ nGK-defi- ably be a valid alternative treatment, given the normalizing cient mice. The initial prediction that, by removing the effect of KATP deficiency that we now demonstrate, be- electrical defect, the diabetic consequences of nGK defi- cause nGKϩ/ϪKir6.2ϩ/ϩ transgenic mice do respond to ciency should be significantly avoided on the Kir6.2Ϫ/Ϫ glibenclamide treatment in vitro (15). background was thus supported. However, the mechanis- Taken together, the results from nGK-deficient and tic consequences of the genetic alterations are not abso- KATP-deficient crosses show a marked improvement in lutely clear. The naive expectation of mice lacking KATP survivability and glucose tolerance of the former on the Ϫ/Ϫ channels was a persistent hyperinsulinemic phenotype KATP null background. Neonatal lethality of nGK because of unregulated secretion, but several studies have Kir6.2ϩ/ϩ mice is avoided on the Kir6.2Ϫ/Ϫ background, now shown that this expectation is not met (25–27). For demonstrating that this consequence is attributable to loss unexplained reasons (but see below), Kir6.2Ϫ/Ϫ mice are of metabolic coupling between glucose and insulin secre- normoglycemic, with slightly impaired glucose tolerance tion, probably through electrical signaling. As with the and reduced GSIS. Similarly, although glucose tolerance now dramatically improved treatment options for KATP- was greatly improved in nGKϩ/Ϫ mice on the Kir6.2Ϫ/Ϫ induced permanent neonatal diabetes (21,24), we suggest background, GSIS in isolated islets was still notably re- that there may be a role for sulfonylurea therapy in duced (Fig. 6A). However, as has recently been shown, the permanent neonatal diabetes resulting from glucokinase sensitizing effect of amino acids, in particular glutamine deficiency (13). A significant number of nGKϪ/ϪKir6.2Ϫ/Ϫ Ϫ/Ϫ (30,31), is much greater in islets that lack KATP (SUR1 ). mice survived beyond weaning. This should also allow Similarly, we find that Kir6.2Ϫ/Ϫ islets can show enhanced more extensive studies of the nonelectrical consequences insulin secretion at physiological glucose concentration (which are still ultimately lethal) of total glucokinase (compared with Kir6.2ϩ/ϩ mice) in the presence of stimu- deficiency. Such studies will have the potential to inform latory glutamine (Fig. 6B). This enhancement is seen in and further improve treatment options for diabetes result- both nGKϩ/ϩ and nGKϩ/Ϫ islets on this background. We ing from glucokinase deficiency. suggest that, in vivo, ambient glutamine levels may lead to such an enhancement of insulin secretion and that, in part, ACKNOWLEDGMENTS this underlies the abrogating effect of KATP deficiency on This work was supported by a mentor-based minority nGK-deficiency (31). postdoctoral fellowship (to M.S.R.) from the American Because we are examining a global neuroendocrine Diabetes Association and Takeda Pharmaceuticals, by the glucokinase deficiency, it is conceivable that the effects of National Institutes of Health Diabetes Research and Train- Kir6.2 knockout are mediated through extra-pancreatic ing Center (DK20579) at Washington University, and by consequences, although insulin tolerance tests (Fig. 5) National Institutes of Health grant DK69445 (to C.G.N.). show that nGKϩ/ϪKir6.2Ϫ/Ϫ mice are no more insulin ϩ/Ϫ ϩ/ϩ Kir6.2-deficient mice (25) and nGK-deficient mice (15) sensitive than nGK Kir6.2 mice, excluding enhanced were kindly provided by Dr. Susumo Seino and Dr. Yasuo insulin sensitivity as an explanation for the rescued phe- Terauchi, respectively. notype. Further evidence for a pancreatic basis for the Kir6.2-deficient rescue of nGK deficiency would be a REFERENCES ␤-cell–specific K knockout. A tissue-targeted knockout ATP 1. Ashcroft FM, Rorsman P: ATP-sensitive Kϩ channels: a link between has not been generated, although we have generated an B-cell metabolism and insulin secretion. Biochem Soc Trans 18:109–111, insulin promoter–driven dominant-negative Kir6.2[AAA] 1990 ϳ ␤ mouse (29) that lacks KATP in 70% of its -cells (29). The 2. Matschinsky FM, Glaser B, Magnuson MA: Pancreatic beta-cell glucoki- demonstration of marked improvement in glucose toler- nase: closing the gap between theoretical concepts and experimental ance in double-heterozygous (nGKϩ/ϪKir6.2[AAA]) mice, realities. Diabetes 47:307–315, 1998 compared with nGKϩ/Ϫ alone (Fig. 7), provides further 3. Randle PJ, Priestman DA, Mistry S, Halsall A: Mechanisms modifying ␤ glucose oxidation in diabetes mellitus. Diabetologia 37 (Suppl. 2):S155– support for -cell–specific KATP channels indeed being the S161, 1994 major players in rescuing the nGK-deficiency phenotype. 4. Vionnet N, Stoffel M, Takeda J, Yasuda K, Bell GI, Zouali H, Lesage S, Velho Relevance to glucokinase-deficiency diabetes in hu- G, Iris F, Passa P, et al.: Nonsense mutation in the glucokinase gene causes mans. The significant reduction in fasting glucose, slight early-onset non- insulin-dependent diabetes mellitus. Nature 356:721–722, yet significant decrease in fed glucose concentration, and 1992 marked enhancement in glucose tolerance in nGKϩ/Ϫ 5. Froguel P, Vaxillaire M, Sun F, Velho G, Zouali H, Butel MO, Lesage S, Ϫ/Ϫ ϩ/Ϫ ϩ/ϩ Vionnet N, Clement K, Fougerousse F, et al.: Close linkage of glucokinase Kir6.2 mice with respect to nGK Kir6.2 mice locus on chromosome 7p to early-onset non-insulin-dependent diabetes suggests that, at least in part, the MODY-2 phenotype is mellitus. Nature 356:162–164, 1992 [erratum Nature 357:607, 1992] abrogated on the KATP knockout background, highlighting 6. Page R, Hattersley A, Turner R: Beta-cell secretory defect caused by the importance of electrical signaling in the regulation of mutations in glucokinase gene (Letter). Lancet 340:1162–1163, 1992 insulin secretion. Decreased insulin secretion in response 7. Velho G, Froguel P, Clement K, Pueyo ME, Rakotoambinina B, Zouali H, to glucose from nGKϪ/ϪKir6.2Ϫ/Ϫ or nGKϩ/ϪKir6.2Ϫ/Ϫ is- Passa P, Cohen D, Robert JJ: Primary pancreatic beta-cell secretory defect caused by mutations in glucokinase gene in kindreds of maturity onset lets recapitulates the response observed in nGKϩ/ Ϫ/Ϫ diabetes of the young. Lancet 340:444–448, 1992 [see comments] ϩKir6.2 islets, thus indicating that the impaired GSIS is 8. Sakura H, Eto K, Kadowaki H, Simokawa K, Ueno H, Koda N, Fukushima due to the Kir6.2 knockout phenotype and not due to Y, Akanuma Y, Yazaki Y, Kadowaki T: Structure of the human glucokinase

2930 DIABETES, VOL. 54, OCTOBER 2005 M.S. REMEDI AND ASSOCIATES

gene and identification of a missense mutation in a Japanese patient with A, Lorini R, Cerutti F, Vanelli M, Barbetti F; Early Onset Diabetes Study early-onset non-insulin-dependent diabetes mellitus. J Clin Endocrinol Group of the Italian Society of Pediatric and Diabetology: Metab 75:1571–1573, 1992 [see comments] KCNJ11 activating mutations in Italian patients with permanent neonatal 9. Guazzini B, Gaffi D, Mainieri D, Multari G, Cordera R, Bertolini S, Pozza G, diabetes. Hum Mutat 25:22–27, 2005 Meschi F, Barbetti F: Three novel missense mutations in the glucokinase 23. Gloyn AL, Reimann F, Girard C, Edghill EL, Proks P, Pearson ER, Temple gene (G80S; E221K; G227C) in Italian subjects with maturity-onset diabe- IK, Mackay DJ, Shield JP, Freedenberg D, Noyes K, Ellard S, Ashcroft FM, tes of the young (MODY): mutations in brief no. 162. Online. Hum Mutat Gribble FM, Hattersley AT: Relapsing diabetes can result from moderately 12:136, 1998 activating mutations in KCNJ11. Hum Mol Genet 14:925–934, 2005 10. Ellard S, Beards F, Allen LI, Shepherd M, Ballantyne E, Harvey R, 24. Gloyn AL, Pearson ER, Antcliff JF, Proks P, Bruining GJ, Slingerland AS, Hattersley AT: A high prevalence of glucokinase mutations in gestational Howard N, Srinivasan S, Silva JM, Molnes J, Edghill EL, Frayling TM, diabetic subjects selected by clinical criteria. Diabetologia 43:250–253, Temple IK, Mackay D, Shield JP, Sumnik Z, van Rhijn A, Wales JK, Clark 2000 P, Gorman S, Aisenberg J, Ellard S, Njolstad PR, Ashcroft FM, Hattersley 11. Miller SP, Anand GR, Karschnia EJ, Bell GI, LaPorte DC, Lange AJ: AT: Activating mutations in the gene encoding the ATP-sensitive potassi- Characterization of glucokinase mutations associated with maturity-onset um-channel subunit Kir6.2 and permanent neonatal diabetes. N Engl J Med diabetes of the young type 2 (MODY-2): different glucokinase defects lead 350:1838–1849, 2004 to a common phenotype. Diabetes 48:1645–1651, 1999 25. Miki T, Nagashima K, Tashiro F, Kotake K, Yoshitomi H, Tamamoto A, 12. Njolstad PR, Sovik O, Cuesta-Munoz A, Bjorkhaug L, Massa O, Barbetti F, Gonoi T, Iwanaga T, Miyazaki J, Seino S: Defective insulin secretion and Undlien DE, Shiota C, Magnuson MA, Molven A, Matschinsky FM, Bell GI: enhanced insulin action in KATP channel-deficient mice. Proc Natl Acad Neonatal diabetes mellitus due to complete glucokinase deficiency. N Engl SciUSA95:10402–10406, 1998 J Med 344:1588–1592, 2001 26. Seghers V, Nakazaki M, DeMayo F, Aguilar-Bryan L, Bryan J: Sur1 13. Gloyn AL, Colclough K, Batten M, Allen LI, Beards F, Hattersley AT, Ellard knockout mice. A model for K(ATP) channel-independent regulation of S: Glucokinase (GCK) mutations in hyper- and hypoglycemia: maturity- insulin secretion. J Biol Chem 275:9270–9277, 2000 onset diabetes of the young, permanent neonatal diabetes, and hyperinsu- 27. Shiota C, Larsson O, Shelton KD, Shiota M, Efanov AM, Hoy M, Lindner J, linemia of infancy. Hum Mutat 22:417, 2003 Kooptiwut S, Juntti-Berggren L, Gromada J, Berggren PO, Magnuson MA: 14. Gloyn AL, Ellard S, Shield JP, Temple IK, Mackay DJ, Polak M, Barrett T, Sulfonylurea receptor type 1 knock-out mice have intact feeding-stimu- Hattersley AT: Complete glucokinase deficiency is not a common cause of lated insulin secretion despite marked impairment in their response to permanent neonatal diabetes. Diabetologia 45:290, 2002 glucose. J Biol Chem 277:37176–37183, 2002 15. Terauchi Y, Sakura H, Yasuda K, Iwamoto K, Takahashi N, Ito K, Kasai H, 28. Remedi MS, Koster JC, Markova K, Seino S, Miki T, Patton BL, McDaniel Suzuki H, Ueda O, Kamada N, et al.: Pancreatic beta-cell-specific targeted ML, Nichols CG: Diet-induced glucose intolerance in mice with decreased disruption of glucokinase gene: diabetes mellitus due to defective insulin ␤-cell ATP-sensitive Kϩ channels. Diabetes 53:3159–3167, 2004 secretion to glucose. J Biol Chem 270:30253–30256, 1995 29. Koster JC, Remedi MS, Flagg TP, Johnson JD, Markova KP, Marshall BA, 16. Ashcroft F, Rorsman P: Type 2 diabetes mellitus: not quite exciting Nichols CG: induced by targeted suppression of enough? Hum Mol Genet 13 (Spec. 1):R21–R31, 2004 KATP channels. Proc Natl Acad SciUSA99:16992–16997, 2002 17. Sakura H, Ashcroft SJ, Terauchi Y, Kadowaki T, Ashcroft FM: Glucose 30. Nenquin M, Szollosi A, Aguilar-Bryan L, Bryan J, Henquin JC: Both modulation of ATP-sensitive K-currents in wild-type, homozygous and triggering and amplifying pathways contribute to fuel-induced insulin heterozygous glucokinase knock-out mice. Diabetologia 41:654–659, 1998 secretion in the absence of sulfonylurea receptor-1 in pancreatic beta- 18. Ashcroft FM,Gribble FM: ATP-sensitive Kϩ channels and insulin secretion: cells. J Biol Chem 279:32316–32324, 2004 their role in health and disease (Review). Diabetologia 42:903–919, 1999 31. Li C, Buettger C, Kwagh J, Matter A, Daikhin Y, Nissim IB, Collins HW, 19. Koster JC, Marshall BA, Ensor N, Corbett JA, Nichols CG: Targeted Yudkoff M, Stanley CA, Matschinsky FM: A signaling role of glutamine in overactivity of beta cell K(ATP) channels induces profound neonatal insulin secretion. J Biol Chem 279:13393–13401, 2004 diabetes. Cell 100:645–654, 2000 32. Doliba NM, Qin W, Vatamaniuk MZ, Li C, Zelent D, Najafi H, Buettger CW, 20. Vaxillaire M, Populaire C, Busiah K, Cave H, Gloyn AL, Hattersley AT, Collins HW, Carr RD, Magnuson MA, Matschinsky FM: Restitution of Czernichow P, Froguel P, Polak M: Kir6.2 mutations are a common cause defective glucose-stimulated insulin release of sulfonylurea type 1 receptor of permanent neonatal diabetes in a large cohort of French patients. knockout mice by acetylcholine. Am J Physiol Endocrinol Metab 286: Diabetes 53:2719–2722, 2004 E834–E843, 2004 21. Sagen JV, Raeder H, Hathout E, Shehadeh N, Gudmundsson K, Baevre H, 33. Postic C, Shiota M, Magnuson MA: Cell-specific roles of glucokinase in Abuelo D, Phornphutkul C, Molnes J, Bell GI, Gloyn AL, Hattersley AT, glucose homeostasis. Recent Prog Horm Res 56:195–217, 2001 Molven A, Sovik O, Njolstad PR: Permanent neonatal diabetes due to 34. Magnuson MA: Tissue-specific regulation of glucokinase gene expression. mutations in KCNJ11 encoding Kir6.2: patient characteristics and initial J Cell Biochem 48:115–121, 1992 response to sulfonylurea therapy. Diabetes 53:2713–2718, 2004 35. Spyer G, Hattersley AT, Sykes JE, Sturley RH, MacLeod KM: Influence of 22. Massa O, Iafusco D, D’Amato E, Gloyn AL, Hattersley AT, Pasquino B, maternal and fetal glucokinase mutations in . Am J Tonini G, Dammacco F, Zanette G, Meschi F, Porzio O, Bottazzo G, Crino Obstet Gynecol 185:240–241, 2001

DIABETES, VOL. 54, OCTOBER 2005 2931