Efficacy of glucokinase activation on glycemic control is limited to a short-term period. One reason might be related to excess glucose signaling by glucokinase activation toward β-cells. In this study, we investigated the effect of glucokinase haploinsufficiency on glucose tolerance as well as β-cell function and mass using a mouse model of type 2 diabetes. Our results showed that in db/db mice with glucokinase haploinsufficiency, glucose tolerance was ameliorated by augmented insulin secretion associated with the increase in β-cell mass when compared with db/db mice. Gene expression profiling and immunohistochemical and metabolomic analyses revealed that glucokinase haploinsufficiency in the islets of db/db mice was associated with lower expression of stress-related genes, greater expression of transcription factors involved in the maintenance and maturation of β-cell function, less mitochondrial damage, and a superior metabolic pattern. These effects of glucokinase haploinsufficiency could preserve β-cell mass under diabetic conditions. These findings verified our hypothesis that optimizing excess glucose signaling in β-cells by inhibiting glucokinase could prevent β-cell insufficiency, leading to improving glucose tolerance in diabetes status by preserving β-cell mass. Therefore, glucokinase inactivation in β-cells, paradoxically, could be a potential strategy for the treatment of type 2 diabetes.
Introduction
The number of people worldwide with diabetes is almost half a billion and it is predicted to increase 25% by 2030 and 51% by 2045 (1). Thus, the establishment of a promising therapeutic strategy for diabetes is an urgent task. Type 1 diabetes and type 2 diabetes are characterized by insufficient insulin secretion due to reduced numbers and/or function of pancreatic β-cells (2–6); thus, preserving, expanding, and improving β-cell function and/or mass represent fundamental therapeutic approaches to the treatment of diabetes (7).
Glucokinase, which phosphorylates glucose to form glucose-6-phosphate, is expressed in organs with an integrated role in glucose sensing (8). In pancreatic β-cells, glucose phosphorylation with glucokinase is the rate-limiting step in insulin secretion. Glucokinase activation ultimately stimulates insulin secretion, whereas its inactivation leads to its impairment. Mice with glucokinase haploinsufficiency in pancreatic β-cells exhibit mild hyperglycemia associated with impaired insulin secretion (9). In liver, glucose phosphorylation by glucokinase initiates glucose uptake and promotes glycogen synthesis, leading to lowering of the blood glucose level. Because of its central role in glucose homeostasis, glucokinase was targeted for drug development aiming to augment its activity (10,11). Since 2003, numerous glucokinase activators (GKAs) have been developed, and their ability to lower blood glucose levels has been demonstrated in several animal models of type 2 diabetes (10–16). Moreover, we (17,18) and others (19,20) demonstrated that GKAs have can induce β-cell proliferation in addition to augmenting insulin secretion in pancreatic β-cells and glucose uptake in liver.
However, phase 2 clinical trials of some GKAs showed their unsustained effects on glycemic control (21–23). One of the plausible reasons for their unsustainability could be related to a toxic effect of GKAs on pancreatic β-cells. This hypothesis was elegantly demonstrated using a mouse model of genetic activation of β-cell–specific glucokinase (24). That is, the augmentation of glucose signaling by glucokinase activation led to an initial increase in insulin secretion and β-cell proliferation, but was subsequently associated with β-cell failure (11). This initial augmentation and subsequent β-cell insufficiency were also observed in type 2 diabetes (24).
These observations suggest that optimizing excess glucose signaling in β-cells by reducing glucokinase activity could prevent β-cell failure due to excess glucose influx, leading to improved glucose tolerance in diabetes by preserving β-cell quality and quantity. To test this paradoxical hypothesis, we investigated the effect of glucokinase haploinsufficiency in pancreatic β-cells on glucose tolerance and β-cell function and mass in a mouse model of diabetes.
Research Design and Methods
Animals
Leprdb/+ (db/+) mice were purchased from Oriental Yeast Co. (Tokyo, Japan). These were crossed with Gcktm1Tka/Gck+ (Gck+/−) mice (9), generating Gck+/−db/+ mice. These mice were subsequently crossed to generate Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice. The mutant Gck allele in these mice affects the expression of the neuroendocrine isoform of glucokinase but not that of the hepatic isoform (9). Animals were given free access to drinking water and food. The room temperature was maintained at 25°C. Standard chow (MF; Oriental Yeast Co. Ltd.) was provided. Mice were housed two to three animals per cage for all experiments under controlled ambient conditions with a 12-h light/dark cycle (lights on at 7 a.m.) and maintained in accordance with standard animal care procedures based on institutional guidelines. This study was approved by the Animal Use Committee of Hokkaido University Graduate School of Medicine and was conducted in compliance with the Animal Use Guidelines of Hokkaido University.
Measurement of Biochemical Parameters
Blood glucose level was determined using a Glutestmint portable glucose meter (Sanwa Chemical Co., Nagoya, Japan). Insulin levels were measured using an insulin ELISA kit (Morinaga Institute of Biological Science, Yokohama, Japan). Plasma levels of free fatty acids, triglycerides, and ALT were assayed by enzymatic methods (Wako Pure Chemical Industries Ltd., Osaka, Japan).
Oral Glucose Tolerance Test
For the oral glucose tolerance test (OGTT), mice were fasted for 16 hours before being orally loaded with glucose (0.5 mg/g body wt). Blood samples were collected at 0, 15, 30, 60, and 120 min after glucose administration to determine the blood glucose levels.
Insulin Tolerance Test
Insulin tolerance tests were performed under fasting conditions. Human regular insulin (4 units/kg body wt) was injected intraperitoneally, and blood samples were collected before and at 30, 60, 90, and 120 min after the injection.
β-Cell Morphology and Immunohistochemistry
Isolated pancreata were immersion fixed in 4% paraformaldehyde. The tissue was routinely processed for paraffin embedding, and 5-μm sections mounted on glass slides were immunostained with goat anti–rabbit insulin antibody (diluted 1:1,000; Proteintech Group Inc., Rosemont, IL). We analyzed two sections of each pancreas that were 50 μm apart. β-Cell area was calculated using a BZ-II analyzer (Keyence Co., Osaka, Japan). β-Cell mass was estimated for each animal by determining the β-cell area as a proportion of total pancreatic area per animal and multiplying this proportion by the pancreas weight. For immunofluorescence, tissue sections were incubated overnight at 4°C with the primary antibodies listed in Supplementary Table 1. After rinsing with PBS, tissues were incubated with secondary antibodies for 60 min (diluted 1:200) using previously described procedures (25) (Supplementary Table 1). Immunofluorescence images were acquired using a BZ-II analyzer (Keyence Co.) according to the manufacturer’s instructions.
Islet Isolation and Glucose-Stimulated Insulin Secretion Assay
Islets were isolated using collagenase from Clostridium histolyticum (Sigma-Aldrich, St. Louis, MO) according to the manufacturer’s instructions. Krebs-Ringer-HEPES buffer (KRBH) was prepared as described previously (17). Isolated islets were preincubated for 30 min at 37°C in 500 µL KRBH with 2.8 mmol/L glucose. Subsequently, islets were transferred for a 90-min incubation in 500 µL KRBH in low (5.6 mmol/L) or high (22 mmol/L) glucose concentrations. After incubation, medium was collected and stored at −20°C. To determine the insulin content, isolated islets were incubated in acid ethanol, and the insulin concentration in the assay buffer was determined using an insulin ELISA kit (Morinaga Institute of Biological Science).
Real-time Quantitative PCR Analysis
Total RNA was extracted with an RNeasy Mini Kit (Qiagen, Hilden, Germany) and used as the starting material for cDNA preparation. Real-time PCR was performed in duplicate using a 7500 Fast Real Time PCR system with power SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA). Regarding the islet experiments, data were normalized to GAPDH expression. The primer sequences used are listed in Supplementary Table 2.
Microarray Analysis
Total RNA from isolated islets was extracted with an RNeasy Mini Kit (Qiagen). The mRNA expression profiles were determined using a Clariom S Mouse Gene 2.0 ST array (Thermo Fisher Scientific Inc., Waltham, MA). Differentially expressed genes were defined as genes that showed at least a 1.5-fold change.
Metabolome Analysis
Approximately 5 mg of frozen isolated islets were plunged into 450 µL of 50% acetonitrile/Milli-Q water containing internal standards (H3304–1002; Human Metabolome Technologies, Inc., Tsuruoka, Japan) at 0°C to inactivate enzymes. The tissue was homogenized twice at 1,500 rpm for 120 s using a tissue homogenizer (Micro Smash MS100R; Tomy Digital Biology Co., Ltd., Tokyo, Japan), and the homogenate was centrifuged at 2,300g and 4°C for 5 min. Subsequently, 400 µL of the upper aqueous layer was centrifugally filtered through a 5-kDa cutoff filter (Millipore, Billerica, MA) at 9,100g and 4°C for 120 min to remove proteins. The filtrate was centrifugally concentrated and resuspended in 25 µL of Milli-Q water for capillary electrophoresis mass spectrometry analysis. Metabolome measurements were carried out through a facility service at Human Metabolome Technologies, Inc. Hierarchical cluster analysis was performed using the software PeakStat and SampleStat (Human Metabolome Technologies, Inc.). Detected metabolites were plotted on metabolic pathway maps using Visualization and Analysis of Networks containing Experimental Data (VANTED) software (26).
Statistics
Results are expressed as means ± SD. Differences between two groups were assessed using Student t test. Individual comparisons between more than two groups were performed using ANOVA followed by a post hoc Bonferroni test. Survival times were assessed by log-rank test. P < 0.05 was considered statistically significant.
Data and Resource Availability
The data sets are available from the corresponding author upon reasonable request.
Results
Effect of Glucokinase Haploinsufficiency in db/db Mice on Glucose Tolerance and β-Cell Mass
To examine whether glucokinase inhibition affects glucose tolerance in vivo, Leprdb/+ (db/+) mice were crossed with Gck+/− mice (9), generating Gck+/−db/+ mice. Then we crossed and generated Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice, and compared their metabolic phenotypes. Gck+/+db/db and Gck+/−db/db mice gained significantly more body weight than did Gck+/+db/+ or Gck+/−db/+ mice (Fig. 1A and Supplementary Fig. 1A). Although Gck+/+db/+ and Gck+/−db/+ mice had similar weight gain, Gck+/−db/db mice had slightly but significantly greater weight gain compared with Gck+/+db/db mice (Fig. 1A and Supplementary Fig. 1A). The fed glucose levels of Gck+/−db/+ mice were higher than those of Gck+/+db/+ mice, as described previously (9,27,28), whereas those of Gck+/−db/db mice conversely decreased after 13 weeks of age and remained significantly lower than those of Gck+/+db/db mice after 20 weeks of age (Fig. 1B and Supplementary Fig. 1B). There was no difference in food intake between the Gck+/+db/db and Gck+/−db/db mice (Supplementary Fig. 2A).
Effects of glucokinase haploinsufficiency on metabolic changes and glucose tolerance. A and B: Changes in (A) body weight and (B) blood glucose levels in the Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice (n = 15–18) (*P < 0.05; **P < 0.01 vs. Gck+/+db/db). C: Blood glucose levels during the OGTT in the Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice after a 16-h fast at 23 weeks of age (n = 12–18). D: Area under the curve of glucose excursion during the OGTT in the Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice after a 16-h fast at 23 weeks of age (n = 12–18) (*P < 0.05 vs. Gck+/+db/db). E: Fasting plasma insulin level in the Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice after a 16-h fast at 23 weeks of age (n = 8–21). F: Glucose-stimulated insulin secretion in isolated islets from the Gck+/+db/db and Gck+/−db/db mice at 23 weeks of age (n = 9–11). G: The survival time of the Gck+/+db/db and Gck+/−db/db mice (n = 10–17). Values are means ± SD. P values were determined using ANOVA followed by post hoc Bonferroni test in A–F. Survival times were assessed by log-rank test in G. *P < 0.05; **P < 0.01.
Effects of glucokinase haploinsufficiency on metabolic changes and glucose tolerance. A and B: Changes in (A) body weight and (B) blood glucose levels in the Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice (n = 15–18) (*P < 0.05; **P < 0.01 vs. Gck+/+db/db). C: Blood glucose levels during the OGTT in the Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice after a 16-h fast at 23 weeks of age (n = 12–18). D: Area under the curve of glucose excursion during the OGTT in the Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice after a 16-h fast at 23 weeks of age (n = 12–18) (*P < 0.05 vs. Gck+/+db/db). E: Fasting plasma insulin level in the Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice after a 16-h fast at 23 weeks of age (n = 8–21). F: Glucose-stimulated insulin secretion in isolated islets from the Gck+/+db/db and Gck+/−db/db mice at 23 weeks of age (n = 9–11). G: The survival time of the Gck+/+db/db and Gck+/−db/db mice (n = 10–17). Values are means ± SD. P values were determined using ANOVA followed by post hoc Bonferroni test in A–F. Survival times were assessed by log-rank test in G. *P < 0.05; **P < 0.01.
Furthermore, the OGTT results when the mice were 23 weeks old indicated that Gck+/−db/db mice were significantly more glucose tolerant than were the Gck+/+db/db mice (Fig. 1C and D). There were no difference in insulin sensitivity measured by the insulin tolerance test, lipid profile, or hepatic function between the Gck+/+db/db and Gck+/−db/db mice (Supplementary Fig. 2B–E). Gck+/−db/db mice had significantly higher plasma insulin levels than did the Gck+/+db/db mice (Fig. 1E). Ex vivo glucose-stimulated insulin secretion of isolated islets from Gck+/−db/db mice significantly increased compared with that of Gck+/+db/db mice (Fig. 1F). Interestingly, the survival time of Gck+/−db/db mice was significantly longer than that of Gck+/+db/db mice (Fig. 1G).
Subsequently, we evaluated pancreatic β-cell mass of these mice when they were 24 weeks old. Histological analysis showed that β-cell area in Gck+/−db/db mice was significantly greater than that in Gck+/+db/db mice, although Gck+/+db/+ and Gck+/−db/+ mice had similar β-cell area (Fig. 2A and B). There were no differences in pancreatic weight between the Gck+/+db/+ and Gck+/−db/+ mice and between the Gck+/+db/db and Gck+/−db/db mice (Fig. 2C). As a result, β-cell mass in Gck+/−db/db mice significantly increased compared with that of the other three groups (Fig. 2D). On the other hand, there was no difference in the ratio of the number of α-cells to the total combined number of α- and β-cells in the pancreatic islets of the Gck+/+db/db and Gck+/−db/db mice (Supplementary Fig. 3). Taken together, these results show glucose tolerance in Gck+/−db/db mice was ameliorated by augmentation of insulin secretion associated with increasing β-cell mass, compared with Gck+/+db/db mice.
Effect of glucokinase haploinsufficiency on β-cell morphology in mice aged 24 weeks. A: Immunohistological analysis of pancreatic islets from the four groups; Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice. β-Cells are stained brown. Black scale bars represent 500 μm. B–D: Quantitation of (B) β-cell area, (C) pancreatic weight, and (D) β-cell mass in the Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice (n = 9–15). Values are means ± SD. P values were determined using ANOVA followed by post hoc Bonferroni test. **P < 0.01.
Effect of glucokinase haploinsufficiency on β-cell morphology in mice aged 24 weeks. A: Immunohistological analysis of pancreatic islets from the four groups; Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice. β-Cells are stained brown. Black scale bars represent 500 μm. B–D: Quantitation of (B) β-cell area, (C) pancreatic weight, and (D) β-cell mass in the Gck+/+db/+, Gck+/−db/+, Gck+/+db/db, and Gck+/−db/db mice (n = 9–15). Values are means ± SD. P values were determined using ANOVA followed by post hoc Bonferroni test. **P < 0.01.
Effect of Glucokinase Haploinsufficiency in db/db Mice on Exhaustive Changes in Gene Expression
To explore a striking difference in gene expression of islets from Gck+/+db/db and Gck+/−db/db mice, DNA microarray analysis was performed using islets isolated from these mice when they were 10 weeks old. Of note, when aged 10 weeks, fed blood glucose levels and glucose tolerance evaluated by OGTT in Gck+/−db/db mice were equivalent to those in Gck+/+db/db mice (Fig. 1B and Supplementary Fig. 4A). β-Cell mass was also comparable between the two groups of mice (Supplementary Fig. 4B). Microarray analysis revealed 1,484 genes were differentially expressed 1.5 fold, with statistical significance in islets isolated from Gck+/−db/db versus Gck+/+db/db mice. Among these, 432 were upregulated and 1,052 were downregulated. Pathway analyses showed significant changes in seven upregulated and 22 downregulated pathways (Table 1). In particular, upregulated pathways included the Ptf1a-related regulatory pathway, such as NK6 homeobox 1 (Nkx6.1) and pancreatic and duodenal homeobox 1 (Pdx1), which play important roles in β-cell function and maturation, whereas the downregulated pathways included metabolic stress and β-cell damage-related pathways, such as the oxidative stress and inflammatory response pathway.
Pathway analysis from the microarray data
Pathway Name . | No. of Changed Genes . | Total Genes . | Z score . | P value . | Gene symbol . |
---|---|---|---|---|---|
Ratio > 1.5 | |||||
Mm_Amino_Acid_metabolism_WP662_71177 | 8 | 95 | 4.56 | 7.14E−04 | Hnmt Hadh Hibadh Ddc Pycr1 Auh Mut Pcx |
Mm_GPCRs_Class_B_Secretin-like_WP456_69131 | 3 | 22 | 3.96 | 0.012 | Adgrl3 Gipr Glp1r |
Mm_Ptf1a_related_regulatory_pathway_WP201_69113 | 2 | 11 | 3.89 | 0.025 | Pdx1 Nkx6–1 |
Mm_G1_to_S_cell_cycle_control_WP413_84705 | 4 | 61 | 2.60 | 0.035 | E2f1 Ccnd2 Cdc45 Cdkn1a |
Mm_G_Protein_Signaling_Pathways_WP232_89955 | 5 | 91 | 2.44 | 0.036 | Pde7a Pde4b Gng12 Akap13 Akap5 |
Mm_Non-odorant_GPCRs_WP1396_69993 | 10 | 266 | 2.13 | 0.041 | Adgrb3 P2ry1 Gabbr2 Adgrl3 Gipr Ackr2 Sstr3 Glp1r Mc5r Adra2a |
Mm_Metapathway_biotransformation_WP1251_69747 | 6 | 133 | 2.13 | 0.049 | Hnmt Akr1b10 Chst11 Gal3st1 Gpx2 Hs3st6 |
Ratio < 0.67 | |||||
Mm_Cytoplasmic_Ribosomal_Proteins_WP163_78425 | 34 | 79 | 10.40 | 1.38E−11 | Rpl37a Rpl7 Rpl7a Rpl35 Rps3a1 Rpl34 Rps20 Rps8 Rpl9 Rpl37rt Rps5 Rps19 Rps16 Rpl27a Rplp2 Rps11 Rps13 Rps15a Rpl13 Rps25 Rpl36a Rpl39 Rps4× Rps12 Rpl26 Rps27a Rpl23a Rps7 Rpl37 Rpl3 Rpl35a Rps28 Rps14 Fau |
Mm_Focal_Adhesion_WP85_82887 | 40 | 184 | 5.88 | 3.22E−06 | Col3a1 Lamb3 Col5a2 Fn1 Lamc1 Col5a1 Lamc3 Pdgfc Jun Hgf Pdgfra Spp1 Pxn Actb Col1a2 Cav2 Cav1 Met Vwf Itgam Dock1 Col4a2 Pdgfd Lamb2 Figf Flna Lama4 Igf1 Lama2 Col6a2 Vtn Col1a1 Erbb2 Itgb4 Rhob Itgb8 Shc3 Mylk Lama3 Pdgfrb |
Mm_PodNet-_protein-protein_interactions_in_the_podocyte_WP2310_82879 | 56 | 315 | 5.26 | 9.80E−06 | Sulf1 Cfh Lamc1 Tgfb2 Olfm1 Angptl2 Notch2 Cyr61 Hspg2 Pdpn Agrn Spp1 Pxn Slc29a4 Arpc1b Actb Smurf1 Met Cxcl12 Plaur Tgfb1 Cd151 Bcam Fcgrt Nr2f2 Sh2d4a Nrp1 Palld Myo1e Lamb2 Tagln Smad3 Tgfbr2 Ctgf Igf1 Tcf21 Col18a1 Myh10 Itgb4 Wwc1 Cxcl16 Egln3 Tgfb3 Nid1 Dbn1 Ctsl Scel Bmp4 Lgals1 Angpt1 Igsf5 Crim1 Epas1 Ezr Notch3 Plce1 |
Mm_XPodNet_-_protein-protein_interactions_in_the_podocyte_expanded_by_STRING_WP2309_82878 | 113 | 825 | 4.44 | 2.24E−05 | Sulf1 Nrp2 Fn1 Cfh Lamc1 Tgfb2 Vim Olfm1 Lamc3 Angptl2 Enkur Zeb2 Eng Notch2 Ngf Csf1 Cyr61 Lyn Nr4a3 Hspg2 Pdpn Agrn Cdk6 Ptpn13 Spp1 Pxn Slc29a4 Arpc1b Eln Cldn4 Actb Smurf1 Cav1 Met Fbln2 Bhlhe40 Cxcl12 Vwf Plaur Tgfb1 Gab2 Arrb1 Htra1 Cd151 Cdc42ep5 Bcam Zfp36 Fcgrt Nr2f2 Nrip3 Sh2d4a Cdh5 Sntb2 Plcg2 Nrp1 Palld Gab1 Anxa2 Myo1e Lamb2 Tagln Smad3 Rab8b Tgfbr2 Trf L1cam Ctgf Gja1 Igf1 Cdk17 Tcf21 Lama2 Hk1 Col18a1 Syt1 Gnb2l1 Myh10 Tax1bp3 Vtn Itgb4 Wwc1 Cxcl16 Abcc3 Pecam1 Dock4 Egln3 Tgfb3 Nid1 Actn2 Dbn1 Ctsl Scel Bmp4 Ajuba Sorbs3 Cdc42ep1 Lgals1 Angpt1 Plec Cldn8 Igsf5 Ltbp1 Crim1 Epas1 Ezr Notch3 Pdgfrb Csf1r Rin1 Ffar4 Plce1 Sorbs1 Blnk |
Mm_Prostaglandin_Synthesis_and_Regulation_WP374_69204 | 11 | 31 | 5.07 | 5.14E−04 | Hsd11b1 Ptgs1 S100a6 S100a10 Ptger3 Anxa3 Tbxas1 Hpgd Anxa2 Ednrb Anxa1 |
Mm_Endochondral_Ossification_WP1270_87973 | 16 | 62 | 4.53 | 6.24E−04 | Ddr2 Tgfb2 Fgfr3 Spp1 Mgp Tgfb1 Serpinh1 Sox6 Fgfr1 Timp3 Igf1 Enpp1 Sox9 Mef2c Ctsl Slc38a2 |
Mm_Complement_Activation_Classical_Pathway_WP200_72061 | 7 | 16 | 4.78 | 2.13E−03 | C1qb C1qc C1ra C1s1 C7 Masp1 C4b |
Mm_Spinal_Cord_Injury_WP2432_87679 | 19 | 96 | 3.59 | 3.52E−03 | Selp Fcgr2b Vim Rhoc Tlr4 Tgfb1 Rgma Zfp36 Pirb Mmp12 Icam1 Gja1 Sox9 Rhob Rtn4r Egr1 Aqp4 Sema6a Anxa1 |
Mm_Complement_and_Coagulation_Cascades_WP449_71733 | 14 | 61 | 3.72 | 4.73E−03 | Cfh Tfpi Serping1 C1qb C1qc C1ra Vwf C1s1 Plaur C7 Pros1 Masp1 C4b Cfb |
Mm_Macrophage_markers_WP2271_69962 | 5 | 10 | 4.47 | 6.10E−03 | Cd52 Cd163 Lyz2 Cd68 Cd14 |
Mm_TYROBP_Causal_Network_WP3625_90841 | 13 | 58 | 3.48 | 7.46E−03 | Dpyd C1qc Spp1 Tyrobp Kcne3 Itgam Igsf6 Elf4 Cxcl16 Npc2 Lhfpl2 Slc7a7 Zfp36l2 |
Mm_Focal_Adhesion-PI3K-Akt-mTOR-signaling_pathway_WP2841_89989 | 44 | 318 | 2.86 | 0.011 | Col3a1 Lamb3 Col5a2 Fn1 Lamc1 Col5a1 Lamc3 Pfkfb3 Pdgfc Ngf Csf1 Slc2a1 Hgf Fgfr3 Pdgfra Spp1 Ppargc1a Col1a2 Met Vwf Itgam Fgfr2 Col4a2 Fgfr1 Pdgfd Lamb2 Figf Lama4 Igf1 Kitl Lama2 Col6a2 Vtn Col1a1 Itgb4 Itgb8 Fgfr4 Osmr Angpt1 Epas1 Lama3 Pdgfrb Csf1r Ppp2r2b |
Mm_Oxidative_Stress_WP412_89965 | 7 | 27 | 3.01 | 0.021 | Cat Mgst1 Hmox1 Mt1 Nfix Junb Cyba |
Mm_Alpha6-Beta4_Integrin_Signaling_Pathway_WP488_72049 | 13 | 66 | 2.95 | 0.021 | Lamb3 Lamc1 Vim Met Cd151 Lamb2 Smad3 Ar Lama2 Erbb2 Itgb4 Plec Lama3 |
Mm_Striated_Muscle_Contraction_WP216_87693 | 10 | 45 | 3.02 | 0.023 | Des Vim Myl9 Tpm3 Tpm2 Tpm4 Tpm1 Dmd Actn2 Acta2 |
Mm_Primary_Focal_Segmental_Glomerulosclerosis_FSGS_WP2573_89874 | 13 | 70 | 2.71 | 0.026 | Vim Jag1 Tlr4 Agrn Plaur Tgfb1 Cd151 Myo1e Lamb2 Vtn Itgb4 Ctsl Plce1 |
Mm_Matrix_Metalloproteinases_WP441_69114 | 7 | 29 | 2.79 | 0.028 | Mmp23 Mmp2 Mmp15 Mmp12 Timp3 Timp2 Mmp14 |
Mm_Inflammatory_Response_Pathway_WP458_78438 | 7 | 30 | 2.68 | 0.032 | Col3a1 Fn1 Lamc1 Col1a2 Lamb2 Vtn Col1a1 |
Mm_Microglia_Pathogen_Phagocytosis_Pathway_WP3626_87399 | 9 | 41 | 2.83 | 0.034 | Vav3 Lyn C1qb C1qc Arpc1b Tyrobp Itgam Plcg2 Cyba |
Mm_Factors_and_pathways_affecting_insulin-like_growth_factor_(IGF1)-Akt_signaling_WP3675_90178 | 7 | 31 | 2.58 | 0.037 | Igfbp5 Pdk1 Pld1 Ppargc1a Smad3 Igf1 Igfbp4 |
Mm_Irinotecan_Pathway_WP475_69144 | 3 | 7 | 3.08 | 0.045 | Bche Abcg2 Ces1d |
Mm_Statin_Pathway_WP1_73346 | 5 | 19 | 2.58 | 0.046 | Pltp Abca1 Scarb1 Apoe Lrp1 |
Pathway Name . | No. of Changed Genes . | Total Genes . | Z score . | P value . | Gene symbol . |
---|---|---|---|---|---|
Ratio > 1.5 | |||||
Mm_Amino_Acid_metabolism_WP662_71177 | 8 | 95 | 4.56 | 7.14E−04 | Hnmt Hadh Hibadh Ddc Pycr1 Auh Mut Pcx |
Mm_GPCRs_Class_B_Secretin-like_WP456_69131 | 3 | 22 | 3.96 | 0.012 | Adgrl3 Gipr Glp1r |
Mm_Ptf1a_related_regulatory_pathway_WP201_69113 | 2 | 11 | 3.89 | 0.025 | Pdx1 Nkx6–1 |
Mm_G1_to_S_cell_cycle_control_WP413_84705 | 4 | 61 | 2.60 | 0.035 | E2f1 Ccnd2 Cdc45 Cdkn1a |
Mm_G_Protein_Signaling_Pathways_WP232_89955 | 5 | 91 | 2.44 | 0.036 | Pde7a Pde4b Gng12 Akap13 Akap5 |
Mm_Non-odorant_GPCRs_WP1396_69993 | 10 | 266 | 2.13 | 0.041 | Adgrb3 P2ry1 Gabbr2 Adgrl3 Gipr Ackr2 Sstr3 Glp1r Mc5r Adra2a |
Mm_Metapathway_biotransformation_WP1251_69747 | 6 | 133 | 2.13 | 0.049 | Hnmt Akr1b10 Chst11 Gal3st1 Gpx2 Hs3st6 |
Ratio < 0.67 | |||||
Mm_Cytoplasmic_Ribosomal_Proteins_WP163_78425 | 34 | 79 | 10.40 | 1.38E−11 | Rpl37a Rpl7 Rpl7a Rpl35 Rps3a1 Rpl34 Rps20 Rps8 Rpl9 Rpl37rt Rps5 Rps19 Rps16 Rpl27a Rplp2 Rps11 Rps13 Rps15a Rpl13 Rps25 Rpl36a Rpl39 Rps4× Rps12 Rpl26 Rps27a Rpl23a Rps7 Rpl37 Rpl3 Rpl35a Rps28 Rps14 Fau |
Mm_Focal_Adhesion_WP85_82887 | 40 | 184 | 5.88 | 3.22E−06 | Col3a1 Lamb3 Col5a2 Fn1 Lamc1 Col5a1 Lamc3 Pdgfc Jun Hgf Pdgfra Spp1 Pxn Actb Col1a2 Cav2 Cav1 Met Vwf Itgam Dock1 Col4a2 Pdgfd Lamb2 Figf Flna Lama4 Igf1 Lama2 Col6a2 Vtn Col1a1 Erbb2 Itgb4 Rhob Itgb8 Shc3 Mylk Lama3 Pdgfrb |
Mm_PodNet-_protein-protein_interactions_in_the_podocyte_WP2310_82879 | 56 | 315 | 5.26 | 9.80E−06 | Sulf1 Cfh Lamc1 Tgfb2 Olfm1 Angptl2 Notch2 Cyr61 Hspg2 Pdpn Agrn Spp1 Pxn Slc29a4 Arpc1b Actb Smurf1 Met Cxcl12 Plaur Tgfb1 Cd151 Bcam Fcgrt Nr2f2 Sh2d4a Nrp1 Palld Myo1e Lamb2 Tagln Smad3 Tgfbr2 Ctgf Igf1 Tcf21 Col18a1 Myh10 Itgb4 Wwc1 Cxcl16 Egln3 Tgfb3 Nid1 Dbn1 Ctsl Scel Bmp4 Lgals1 Angpt1 Igsf5 Crim1 Epas1 Ezr Notch3 Plce1 |
Mm_XPodNet_-_protein-protein_interactions_in_the_podocyte_expanded_by_STRING_WP2309_82878 | 113 | 825 | 4.44 | 2.24E−05 | Sulf1 Nrp2 Fn1 Cfh Lamc1 Tgfb2 Vim Olfm1 Lamc3 Angptl2 Enkur Zeb2 Eng Notch2 Ngf Csf1 Cyr61 Lyn Nr4a3 Hspg2 Pdpn Agrn Cdk6 Ptpn13 Spp1 Pxn Slc29a4 Arpc1b Eln Cldn4 Actb Smurf1 Cav1 Met Fbln2 Bhlhe40 Cxcl12 Vwf Plaur Tgfb1 Gab2 Arrb1 Htra1 Cd151 Cdc42ep5 Bcam Zfp36 Fcgrt Nr2f2 Nrip3 Sh2d4a Cdh5 Sntb2 Plcg2 Nrp1 Palld Gab1 Anxa2 Myo1e Lamb2 Tagln Smad3 Rab8b Tgfbr2 Trf L1cam Ctgf Gja1 Igf1 Cdk17 Tcf21 Lama2 Hk1 Col18a1 Syt1 Gnb2l1 Myh10 Tax1bp3 Vtn Itgb4 Wwc1 Cxcl16 Abcc3 Pecam1 Dock4 Egln3 Tgfb3 Nid1 Actn2 Dbn1 Ctsl Scel Bmp4 Ajuba Sorbs3 Cdc42ep1 Lgals1 Angpt1 Plec Cldn8 Igsf5 Ltbp1 Crim1 Epas1 Ezr Notch3 Pdgfrb Csf1r Rin1 Ffar4 Plce1 Sorbs1 Blnk |
Mm_Prostaglandin_Synthesis_and_Regulation_WP374_69204 | 11 | 31 | 5.07 | 5.14E−04 | Hsd11b1 Ptgs1 S100a6 S100a10 Ptger3 Anxa3 Tbxas1 Hpgd Anxa2 Ednrb Anxa1 |
Mm_Endochondral_Ossification_WP1270_87973 | 16 | 62 | 4.53 | 6.24E−04 | Ddr2 Tgfb2 Fgfr3 Spp1 Mgp Tgfb1 Serpinh1 Sox6 Fgfr1 Timp3 Igf1 Enpp1 Sox9 Mef2c Ctsl Slc38a2 |
Mm_Complement_Activation_Classical_Pathway_WP200_72061 | 7 | 16 | 4.78 | 2.13E−03 | C1qb C1qc C1ra C1s1 C7 Masp1 C4b |
Mm_Spinal_Cord_Injury_WP2432_87679 | 19 | 96 | 3.59 | 3.52E−03 | Selp Fcgr2b Vim Rhoc Tlr4 Tgfb1 Rgma Zfp36 Pirb Mmp12 Icam1 Gja1 Sox9 Rhob Rtn4r Egr1 Aqp4 Sema6a Anxa1 |
Mm_Complement_and_Coagulation_Cascades_WP449_71733 | 14 | 61 | 3.72 | 4.73E−03 | Cfh Tfpi Serping1 C1qb C1qc C1ra Vwf C1s1 Plaur C7 Pros1 Masp1 C4b Cfb |
Mm_Macrophage_markers_WP2271_69962 | 5 | 10 | 4.47 | 6.10E−03 | Cd52 Cd163 Lyz2 Cd68 Cd14 |
Mm_TYROBP_Causal_Network_WP3625_90841 | 13 | 58 | 3.48 | 7.46E−03 | Dpyd C1qc Spp1 Tyrobp Kcne3 Itgam Igsf6 Elf4 Cxcl16 Npc2 Lhfpl2 Slc7a7 Zfp36l2 |
Mm_Focal_Adhesion-PI3K-Akt-mTOR-signaling_pathway_WP2841_89989 | 44 | 318 | 2.86 | 0.011 | Col3a1 Lamb3 Col5a2 Fn1 Lamc1 Col5a1 Lamc3 Pfkfb3 Pdgfc Ngf Csf1 Slc2a1 Hgf Fgfr3 Pdgfra Spp1 Ppargc1a Col1a2 Met Vwf Itgam Fgfr2 Col4a2 Fgfr1 Pdgfd Lamb2 Figf Lama4 Igf1 Kitl Lama2 Col6a2 Vtn Col1a1 Itgb4 Itgb8 Fgfr4 Osmr Angpt1 Epas1 Lama3 Pdgfrb Csf1r Ppp2r2b |
Mm_Oxidative_Stress_WP412_89965 | 7 | 27 | 3.01 | 0.021 | Cat Mgst1 Hmox1 Mt1 Nfix Junb Cyba |
Mm_Alpha6-Beta4_Integrin_Signaling_Pathway_WP488_72049 | 13 | 66 | 2.95 | 0.021 | Lamb3 Lamc1 Vim Met Cd151 Lamb2 Smad3 Ar Lama2 Erbb2 Itgb4 Plec Lama3 |
Mm_Striated_Muscle_Contraction_WP216_87693 | 10 | 45 | 3.02 | 0.023 | Des Vim Myl9 Tpm3 Tpm2 Tpm4 Tpm1 Dmd Actn2 Acta2 |
Mm_Primary_Focal_Segmental_Glomerulosclerosis_FSGS_WP2573_89874 | 13 | 70 | 2.71 | 0.026 | Vim Jag1 Tlr4 Agrn Plaur Tgfb1 Cd151 Myo1e Lamb2 Vtn Itgb4 Ctsl Plce1 |
Mm_Matrix_Metalloproteinases_WP441_69114 | 7 | 29 | 2.79 | 0.028 | Mmp23 Mmp2 Mmp15 Mmp12 Timp3 Timp2 Mmp14 |
Mm_Inflammatory_Response_Pathway_WP458_78438 | 7 | 30 | 2.68 | 0.032 | Col3a1 Fn1 Lamc1 Col1a2 Lamb2 Vtn Col1a1 |
Mm_Microglia_Pathogen_Phagocytosis_Pathway_WP3626_87399 | 9 | 41 | 2.83 | 0.034 | Vav3 Lyn C1qb C1qc Arpc1b Tyrobp Itgam Plcg2 Cyba |
Mm_Factors_and_pathways_affecting_insulin-like_growth_factor_(IGF1)-Akt_signaling_WP3675_90178 | 7 | 31 | 2.58 | 0.037 | Igfbp5 Pdk1 Pld1 Ppargc1a Smad3 Igf1 Igfbp4 |
Mm_Irinotecan_Pathway_WP475_69144 | 3 | 7 | 3.08 | 0.045 | Bche Abcg2 Ces1d |
Mm_Statin_Pathway_WP1_73346 | 5 | 19 | 2.58 | 0.046 | Pltp Abca1 Scarb1 Apoe Lrp1 |
Regarding glucose metabolism, we found the expression of some genes that encode glycolytic enzymes, including aldolase B (Aldob), enolase 1 (Eno1), and glucose phosphate isomerase 1 (Gpi1), was lower in Gck+/−db/db mice than in Gck+/+db/db mice (Supplementary Table 4). In contrast, the expression of genes that encode enzymes involved in the tricarboxylic acid (TCA) cycle, including pyruvate carboxylase (Pcx), isocitrate dehydrogenase 3, γ (Idh3g), isocitrate dehydrogenase 2 (Idh2), and aconitase 2 (Aco2), was significantly higher or tended to be higher in Gck+/−db/db mice. Moreover, the expression of the gene encoding pyruvate dehydrogenase kinase, isoenzyme 1 (Pdk1), which limits the entry of pyruvate into the TCA cycle, was significantly lower in Gck+/−db/db mice than in Gck+/+db/db mice (Table 1 and Supplementary Tables 3 and 4). These results suggest glucokinase haploinsufficiency is associated with greater expression of β-cell–associated transcription factors, less expression of metabolic stress-related factors, and altered expression of glucose metabolism–related factors in db/db mice.
Effect of Glucokinase Haploinsufficiency in db/db Mice on β-Cell Proliferation and Death, and β-Cell–Associated Transcription Factors
Although there was no difference in β-cell mass between the Gck+/+db/db and Gck+/−db/db mice aged 10 weeks (Supplementary Fig. 4B), insulin content in isolated islets of Gck+/−db/db mice was significantly higher than in Gck+/+db/db mice (Fig. 3A). Moreover, the β-cell proliferation rate evaluated by Ki67 staining was significantly greater in Gck+/−db/db mice (Fig. 3B). There were no differences in the β-cell apoptotic rate, evaluated using fluorometric TUNEL staining, or in the expression of islet progenitor cell markers, including neurogenin 3 (Neurog3) and aldehyde dehydrogenase 1a3 (Aldh1a3), between the two groups of mice at 10 weeks of age (Supplementary Fig. 5). However, the rate of TUNEL-positive β-cells in Gck+/−db/db mice tended to be lower than in Gck+/+db/db mice at 24 weeks of age (Supplementary Fig. 6).
Effect of glucokinase haploinsufficiency on β-cell proliferation and transcription factors in db/db mice aged 10 weeks. A: Islet insulin content in the Gck+/+db/db and Gck+/−db/db mice (n = 8). B: Proliferation rate of β-cells assessed by Ki67 staining in the Gck+/+db/db and Gck+/−db/db mice (n = 4–5). C: Gene expression levels of Ki67, Ccna2, Ccnb1, Ccnd1, Ccnd2, and Ccnd3 in the Gck+/+db/db and Gck+/−db/db mice by real-time quantitative PCR (n = 8–11). D: Gene expression levels of Nkx6.1, Mafa, Pdx1, Ins1, and Ins2 in the Gck+/+db/db and Gck+/−db/db mice by real-time quantitative PCR (n = 8–11). E: Representative insulin (green) and MafA (red) immunofluorescent staining of pancreas sections from the Gck+/+db/db and Gck+/−db/db mice. Yellow scale bars represent 100 μm. F: Quantitation of MafA and insulin double-positive cells relative to insulin-positive cells (%) in the Gck+/+db/db and Gck+/−db/db mice (n = 4–6). G: Representative insulin (green) and Nkx6.1 (red) immunofluorescent staining of pancreas sections from the Gck+/+db/db and Gck+/−db/db mice. Yellow scale bars represent 100 μm. H: Quantitation of Nkx6.1 and insulin double-positive cells relative to insulin-positive cells (%) in the Gck+/+db/db and Gck+/−db/db mice (n = 4–6). Values are means ± SD. P values were determined using Student t test. *P < 0.05; **P < 0.01.
Effect of glucokinase haploinsufficiency on β-cell proliferation and transcription factors in db/db mice aged 10 weeks. A: Islet insulin content in the Gck+/+db/db and Gck+/−db/db mice (n = 8). B: Proliferation rate of β-cells assessed by Ki67 staining in the Gck+/+db/db and Gck+/−db/db mice (n = 4–5). C: Gene expression levels of Ki67, Ccna2, Ccnb1, Ccnd1, Ccnd2, and Ccnd3 in the Gck+/+db/db and Gck+/−db/db mice by real-time quantitative PCR (n = 8–11). D: Gene expression levels of Nkx6.1, Mafa, Pdx1, Ins1, and Ins2 in the Gck+/+db/db and Gck+/−db/db mice by real-time quantitative PCR (n = 8–11). E: Representative insulin (green) and MafA (red) immunofluorescent staining of pancreas sections from the Gck+/+db/db and Gck+/−db/db mice. Yellow scale bars represent 100 μm. F: Quantitation of MafA and insulin double-positive cells relative to insulin-positive cells (%) in the Gck+/+db/db and Gck+/−db/db mice (n = 4–6). G: Representative insulin (green) and Nkx6.1 (red) immunofluorescent staining of pancreas sections from the Gck+/+db/db and Gck+/−db/db mice. Yellow scale bars represent 100 μm. H: Quantitation of Nkx6.1 and insulin double-positive cells relative to insulin-positive cells (%) in the Gck+/+db/db and Gck+/−db/db mice (n = 4–6). Values are means ± SD. P values were determined using Student t test. *P < 0.05; **P < 0.01.
In addition, to assess fibrosis in these pancreata, Azan staining was performed. This showed fibrosis in and around the islets of Gck+/+db/db, but this was less marked in Gck+/−db/db mice (Supplementary Fig. 7). This greater proliferation and lesser cell damage could explain the greater β-cell mass in Gck+/−db/db mice at 24 weeks of age (Fig. 2D). In addition, quantitative real-time PCR indicated that cell proliferation and cell cycle–related gene expression levels of Ki67, cyclin B1 (Ccnb1), and cyclin D2 (Ccnd2) in Gck+/−db/db mice, but not cyclin D1 (Ccnd1) nor cyclin D3 (Ccnd3), were significantly increased compared with those in Gck+/+db/db mice (Fig. 3C).
In agreement with the findings of the microarray analysis, mRNA expression levels of β-cell–associated transcription factors, including v-maf musculoaponeurotic fibrosarcoma oncogene family, protein A (Mafa), Nkx6.1, and Pdx1, were significantly increased in Gck+/−db/db mice (Fig. 3D). Immunohistochemical analysis revealed an increase in the rate of MafA- and Nkx6.1-positive β-cells in Gck+/−db/db mice in comparison with Gck+/+db/db mice (Fig. 3E–H), whereas there was no difference in the rate of positive β-cells between the Gck+/+db/+ and Gck+/−db/+ mice, and the rate in the Gck+/+db/db mice was lower than those in the Gck+/+db/+ and Gck+/−db/+ mice (Supplementary Fig. 8). Furthermore, the MafA and Nkx6.1 protein expression levels were higher in Gck+/−db/db mice (Supplementary Fig. 9). These findings suggest that Gck+/−db/db mice displayed enhanced insulin content and β-cell proliferation by increasing expression levels of transcription factors associated with β-cell function, compared with Gck+/+db/db mice.
Effect of Glucokinase Haploinsufficiency in db/db Mice on Stress Markers and Glucose Metabolism
Because oxidative stress inactivates the β-cell function–associated transcription factors (29), we considered whether increasing expression levels of these transcription factors could be due to reduced metabolic stress in Gck+/−db/db mice. The microarray analysis revealed that the expression levels of metabolic stress-related genes were downregulated in Gck+/−db/db mice compared with Gck+/+db/db mice (Table 1 and Supplementary Table 3). Similarly, in addition to activating transcription factor 3 (Atf3), mRNA levels of cytochrome b-245 α polypeptide (Cyba) and neutrophil cytosolic factor 1 (Ncf1), which are parts of the NAD phosphate oxidase complex, were significantly decreased in Gck+/−db/db mice by quantitative real-time PCR (Fig. 4A). Subsequently, we examined mitochondrial morphology, which was commonly affected by metabolic stress, in β-cells of the two mouse groups. As shown in Fig. 4B, Tom20 staining of β-cells from Gck+/+db/db mice was heterogeneous, representing fragmentation of the mitochondrial network (30). By contrast, its staining of β-cells from Gck+/−db/db mice was less heterogeneous and extended throughout β-cells. Quantitatively, the mitochondrial area in β-cells of Gck+/−db/db mice was significantly larger than that of Gck+/+db/db mice (Fig. 4C).
Effect of glucokinase haploinsufficiency on stress markers, mitochondrial network, and metabolic factors in db/db mice aged 10 weeks. A: Gene expression levels of β-cell–associated oxidative stress. Expression of Atf3, Cyba, Cybb, Ncf1, Chop, and Casp3 in the Gck+/+db/db and Gck+/−db/db mice was determined by real-time quantitative PCR (n = 8–11). B: Representative Tom20- (red) and DAPI-stained nucleus (blue) immunofluorescent staining of pancreas sections from the Gck+/+db/db and Gck+/−db/db mice. Yellow scale bars represent 50 μm. C: Quantitation of the Tom20-positive area inside the insulin-positive area per islet divided by the number of β-cells in the Gck+/+db/db and Gck+/−db/db mice (n = 4–6). D: Gene expression levels of Pcx, Aldob, and Pfkfb3 in the Gck+/+db/db and Gck+/−db/db mice determined by real-time quantitative PCR (n = 5–11). Values are means ± SD. P values were determined using Student t test. *P < 0.05; **P < 0.01.
Effect of glucokinase haploinsufficiency on stress markers, mitochondrial network, and metabolic factors in db/db mice aged 10 weeks. A: Gene expression levels of β-cell–associated oxidative stress. Expression of Atf3, Cyba, Cybb, Ncf1, Chop, and Casp3 in the Gck+/+db/db and Gck+/−db/db mice was determined by real-time quantitative PCR (n = 8–11). B: Representative Tom20- (red) and DAPI-stained nucleus (blue) immunofluorescent staining of pancreas sections from the Gck+/+db/db and Gck+/−db/db mice. Yellow scale bars represent 50 μm. C: Quantitation of the Tom20-positive area inside the insulin-positive area per islet divided by the number of β-cells in the Gck+/+db/db and Gck+/−db/db mice (n = 4–6). D: Gene expression levels of Pcx, Aldob, and Pfkfb3 in the Gck+/+db/db and Gck+/−db/db mice determined by real-time quantitative PCR (n = 5–11). Values are means ± SD. P values were determined using Student t test. *P < 0.05; **P < 0.01.
Next, we compared glucose metabolism–associated mRNA expression levels, including glycolysis and the TCA cycle in islets between the two mouse groups, because significant changes in their expression levels were observed by microarray analysis (Table 1 and Supplementary Table 3). Not only Pcx upregulation and Aldob downregulation (not significantly) but also 6-phosphofructo-2-kinase/fructose-2,6- biphosphatase 3 (Pfkfb3) downregulation were observed in Gck+/−db/db mice compared with Gck+/+db/db mice (Fig. 4D). These results suggest glucokinase haploinsufficiency could ameliorate metabolic stress, leading to mitochondrial damage reduction, and that it could modify the pattern of gene expressions associated with glucose metabolism in db/db mice.
Effect of Glucokinase Haploinsufficiency in db/db Mice on Metabolites
The results we report here suggest some differences in intracellular metabolism in islets between the Gck+/+db/db and Gck+/−db/db mice. Therefore, we examined intracellular metabolites in islets of the two mouse groups using metabolomics analysis. Figure 5A is a hierarchical cluster analysis heat map of the detected 95 metabolites. This result indicated that alteration in metabolites occurred between the two groups. In detail, glycolytic intermediates and metabolites, including fructose 6-phosphate, pyruvic acid, and lactic acid, were decreased in Gck+/−db/db mice (Fig. 5B). In the TCA cycle, isocitric acid was increased in Gck+/−db/db mice when compared with those in Gck+/+db/db mice (Fig. 5B). Of note, a step up of the ratio of metabolites (Gck+/−db/db mice to Gck+/+db/db mice) from pyruvic acid or malic acid to citric acid (0.6 → 1.0 and 0.3 → 1.0, respectively) was observed. On the basis of the findings that the expression of genes encoding enzymes involved in the TCA cycle, including Pcx, was higher in Gck+/−db/db mice (Fig. 4D and Supplementary Table 4) and that the ATP level in Gck+/−db/db mice was double that of Gck+/+db/db mice (Fig. 5B), we suggest that the metabolism of the Gck+/−db/db mice is not dominated by glycolysis, as in diabetic islets (31), but is instead shifted toward the TCA cycle and oxidative phosphorylation.
Metabolome analysis in islets of the Gck+/+db/db and Gck+/−db/db mice aged 10 weeks. A: Total data for visualization of differentially expressed genes using hierarchical clustering in islets of the Gck+/+db/db and Gck+/−db/db mice (n = 4–5). B: Main metabolites in glycolysis and the tricarboxylic cycle and ATP contents in islets of the Gck+/+db/db and Gck+/−db/db mice (n = 4–5). Values are means ± SD. P values were determined using Student t test. *P < 0.05. G6P, glucose 6-phosphate; F6P, fructose 6-phosphate; F1,6P, fructose 1,6-diphosphate; 3-PG, 3-phosphoglyceric acid; 2-PG, 2-phosphoglyceric acid.
Metabolome analysis in islets of the Gck+/+db/db and Gck+/−db/db mice aged 10 weeks. A: Total data for visualization of differentially expressed genes using hierarchical clustering in islets of the Gck+/+db/db and Gck+/−db/db mice (n = 4–5). B: Main metabolites in glycolysis and the tricarboxylic cycle and ATP contents in islets of the Gck+/+db/db and Gck+/−db/db mice (n = 4–5). Values are means ± SD. P values were determined using Student t test. *P < 0.05. G6P, glucose 6-phosphate; F6P, fructose 6-phosphate; F1,6P, fructose 1,6-diphosphate; 3-PG, 3-phosphoglyceric acid; 2-PG, 2-phosphoglyceric acid.
Discussion
In this study, we demonstrated that glucokinase haploinsufficiency in pancreatic β-cells ameliorated glucose tolerance by insulin secretion augmentation associated with the increase in β-cell mass in db/db mice. Recently, it was reported that reducing glucokinase activity restored pulsatility and improved pulsatility correlated with enhanced insulin secretion in islets of db/db mice (32,33). These results are in agreement with our findings. The strength of our study is our findings about the impact of glucokinase haploinsufficiency on β-cell mass. In addition, we have shown the beneficial effect of glucokinase inactivation on the glucose tolerance of an in vivo model of diabetes.
A mouse model of β-cell–specific genetic activation of glucokinase initially showed greater insulin secretion and β-cell proliferation, but a subsequent progression of β-cell damage (24) and a similar phenotypic progression occurs in db/db mice (34). The β-cell mass of the Gck+/+db/db mice was increased at 10 weeks. However, no additional increase was observed at 24 weeks of age (Supplementary Fig. 10). Instead, β-cell damage was more evident in the form of greater fibrosis (Supplementary Fig. 7). These results imply that Gck transgenic mice and db/db mice have similar β-cell phenotypes, despite differences in obesity and hyperleptinemia. In the islets of db/db mice, there is evidence of higher expression of metabolic stress-related genes, lower expression of pancreatic β-cell–associated transcription factors, and mitochondrial damage (35,36). In contrast, we have shown that glucokinase haploinsufficiency in the islets of db/db mice reduces the expression of stress genes, increases the expression of the transcription factors, reduces mitochondrial damage, and improves the metabolic pattern. These effects of glucokinase haploinsufficiency could result in greater insulin synthesis and the preservation of β-cell mass under diabetic conditions.
We suggest that glucokinase haploinsufficiency could suppress metabolic stress by reducing the influx of excess glucose metabolites. Our data in the microarray analysis support this notion. Oxidative stress reduction by genetically induced antioxidant-enzyme overexpression or antioxidant therapy improved β-cell function and mass in db/db mice (37,38). In this condition, β-cell–associated transcription factors are important for the enhancement of β-cell function and mass. Guo et al. (29) reported that hyperglycemia and oxidative stress lead to β-cell dysfunction by targeting MafA and Nkx6.1 in db/db mice and that glutathione peroxidase-1 antioxidant enzyme overexpression in β-cells in db/db mice restored MafA, Nkx6.1, and β-cell function. These findings were consistent with our results, indicating that the numbers of MafA- and Nkx6.1-positive β-cells in Gck+/−db/db mice were significantly higher than those in Gck+/+db/db mice. The MafA-producing db/db mice had improved glycemic control and augmentation of β-cell function and mass (39), and Nkx6.1 overexpression led to improved glucose-stimulated insulin secretion and enhanced β-cell proliferation (40). Moreover, Taylor et al. (41) demonstrated that Nkx6.1 directly regulates the transcription of genes involved in glucose metabolism, including of Pcx, which plays an important role in metabolizing pyruvate into the TCA cycle. Because the islets of Gck+/−db/db mice had a higher Pcx expression level than those of Gck+/+db/db mice, and because the metabolic pattern shifted toward the TCA cycle and oxidative phosphorylation dominance in Gck+/−db/db mice, there is a possibility that the Nkx6.1-induced Pcx increase is involved in the augmentation of β-cell mass (42,43).
Another possible mechanism is that metabolic stress reduction limits mitochondrial damage. Our data demonstrated an improvement of mitochondrial morphology and a significant decrease of Pfkfb3 expression in Gck+/−db/db mice. Recently, it was reported that mitochondrial fragmentation followed by increased Pfkfb3 expression in β-cells enhanced the aerobic glycolysis that diverted pyruvate acid into lactic acid (30). That is, Pfkfb3 accelerates the metabolic pattern shift from the TCA cycle and oxidative phosphorylation dominance toward glycolytic pathway dominance, which is known and described as metabolic remodeling. Therefore, decreased Pfkfb3 expression could affect amelioration of the metabolic remodeling in Gck+/−db/db mice. Moreover, we need to consider the possibility that a reduction in glucokinase protects pancreatic β-cells from marked increases in intracellular Ca2+ in db/db mice. It has been reported that large increases in intracellular Ca2+ can induce mitochondrial damage and β-cell failure (44). Our gene expression analysis of Ca2+ binding and signaling revealed that the expression of the genes encoding S100 calcium-binding protein A6 (S100a6) and myocyte enhancer factor 2C (Mef2c) in Gck+/−db/db mice was significantly lower than in Gck+/+db/db mice, and that the expression of S100a13 and S100a4 in Gck+/−db/db mice tended to be lower than in Gck+/+db/db mice (Supplementary Table 5). Because S100a6 and S100a4 may represent markers of β-cell excitotoxicity (45), these results suggest glucokinase haploinsufficiency protects pancreatic β-cells against marked increases in intracellular Ca2+, thereby preventing mitochondrial damage and β-cell failure in db/db mice.
From a clinical point of view, glycemic control in patients with type 2 diabetes could be achieved either by promoting endogenous insulin secretion or by reducing both hyperglycemia and insulin secretory demand, which leads to providing β-cell rest (46). The latter strategies seem the most promising because they are expected to preserve β-cell mass. Indeed, β-cell proliferation is induced by genetically reducing insulin production in mouse islets and in a human induced pluripotent β-cell–like stem cell model (47,48). Additionally, in a genetic mouse model, less insulin secretion was associated with an extension of life span (49), which is consistent with the present findings (Fig. 1G). Similar results have been obtained in Drosophila and Caenorhabditis elegans (50–52). Furthermore, the results of a recent in vitro study indicated that relief from hyperexcitability and chronic, high intracellular Ca2+ concentration is important for the prevention of β-cell failure (53). Considering that a reduction in glucokinase expression could protect pancreatic β-cells from high intracellular Ca2+ in db/db mice, glucokinase inhibition could be expected to a new therapeutic strategy for β-cell rest in type 2 diabetes. The finding that glucokinase haploinsufficiency in β-cells of human islet amyloid polypeptide transgenic mice had a decreased incidence and quantity of islet amyloid (54) might demonstrate the usefulness of glucokinase inhibition.
Glucokinase inhibitors, including d-mannoheptulose, have been used as competitive antagonists to prevent glucose phosphorylation (33). However, there are challenges for clinical application. One is that the degree of glucokinase inhibition is important. Complete glucokinase inhibition would result in strong hyperglycemia, as shown in permanent neonatal diabetes mellitus with homozygous glucokinase-inactivating mutations (55). Another important point is the timing of glucokinase inhibition. Given that reduced numbers and/or function of pancreatic β-cells occurs even in the prediabetic state and are part of the natural history of type 2 diabetes progression (2–6), glucokinase inhibition could be more beneficial in individuals recently diagnosed with type 2 diabetes or prediabetes. Not all individuals with diabetes are eligible for this treatment strategy. More preclinical and clinical studies would determine the efficacy of glucokinase inhibition.
A limitation of the present study was that the mice were not pancreatic β-cell specific, although their hepatic expression of glucokinase was intact and their phenotype was very similar to that of β-cell–specific glucokinase knockout mice (9,56). β-Cell–specific knockout mice should be studied to confirm the present findings. On the other hand, it is unclear whether glucokinase inhibitors are pancreatic β-cell specific. Thus, it is necessary to explore the effect of glucokinase inhibition on the liver and brain.
In conclusion, glucokinase haploinsufficiency ameliorated glucose tolerance by insulin-secretion augmentation associated with the increase in β-cell mass in db/db mice. This finding proved that optimizing excess glucose signaling in β-cells by reducing glucokinase activity could prevent β-cell insufficiency due to excess glucose influx, leading to glucose tolerance improvement in diabetes status by preserving β-cell mass. Glucokinase inactivation in β-cells could be a potential strategy for the treatment of type 2 diabetes.
This article contains supplementary material online at https://doi.org/10.2337/figshare.13626674.
Article Information
Acknowledgments. The authors thank Dr. Takashi Kadowaki and Dr. Naoto Kubota (Department of Diabetes and Metabolic Diseases, Graduate School of Medicine, University of Tokyo, Tokyo, Japan) for kindly providing the Gck+/− mice. The authors also thank Marika Watanabe, Natsumi Fujimori, and Yuiko Takeda (Department of Rheumatology, Endocrinology, and Nephrology, Faculty of Medicine and Graduate School of Medicine, Hokkaido University, Sapporo, Japan) for their excellent technical assistance and animal care. The authors thank Dr. Robert Blakytny, from Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript.
Funding. This work was supported in part by Grants-in-Aid for Young Scientists (B) 26860683 and Scientific Research (C) 19K08992 from the Ministry of Education, Culture, Sports, Science and Technology of Japan; a Grant-in-Aid for the Translational Research program, Strategic PRomotion for practical application of INnovative medical Technology (TR-SPRINT), from the Japan Agency for Medical Research and Development; a Grant-in-Aid for Young Researchers from the Japan Association for Diabetes Education and Care; and Grants-in-Aid from the MSD Life Science Foundation, the Suzuken Memorial Foundation, the Akiyama Life Science Foundation, the Takeda Science Foundation, and the Suhara Memorial Foundation (to A.N.).
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. K.O., A.N., and Y.T. designed the studies. K.O., A.N., Y.Y., S.K., K.T., and N.K. performed the in vivo and ex vivo experiments. K.O., A.N., and H.N. analyzed data. K.O., A.N., H.M., H.K., K.Y.C., Y.T., and T.A. interpreted the data. K.O., A.N., and Y.T. drafted the manuscript. A.N. is a guarantor of this study and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Prior Presentation. Parts of this study were presented at the virtual 56th Annual Meeting of the European Association for the Study of Diabetes, 21–25 September 2020.