Hirotaka Watada

Glycation-dependent, reactive oxygen species-mediated suppression of the insulin gene promoter activity in HIT cells.

Prolonged poor glycemic control in non-insulin-dependent diabetes mellitus patients often leads to a decline in insulin secretion from pancreatic beta cells, accompanied by a decrease in the insulin content of the cells. As a step toward elucidating the pathophysiological background of the so-called glucose toxicity to pancreatic beta cells, we induced glycation in HIT-T15 cells using a sugar with strong deoxidizing activity, D-ribose, and examined the effects on insulin gene transcription. The results of reporter gene analyses revealed that the insulin gene promoter is more sensitive to glycation than the control beta-actin gene promoter; approximately 50 and 80% of the insulin gene promoter activity was lost when the cells were kept for 3 d in the presence of 40 and 60 mM D-ribose, respectively. In agreement with this, decrease in the insulin mRNA and insulin content was observed in the glycation-induced cells. Also, gel mobility shift analyses using specific antiserum revealed decrease in the DNA-binding activity of an insulin gene transcription factor, PDX-1/IPF1/STF-1. These effects of D-ribose seemed almost irreversible but could be prevented by addition of 1 mM aminoguanidine or 10 mM N-acetylcysteine, thus suggesting that glycation and reactive oxygen species, generated through the glycation reaction, serve as mediators of the phenomena. These observations suggest that protein glycation in pancreatic beta cells, which occurs in vivo under chronic hyperglycemia, suppresses insulin gene transcription and thus can explain part of the beta cell glucose toxicity.

The Human Glucokinase Gene β-Cell-Type Promoter: An Essential Role of Insulin Promoter Factor 1/PDX-1 in Its Activation in HIT-T15 Cells

The glycolytic enzyme glucokinase plays a primary role in the glucose-responsive secretion of insulin, and defects of this enzyme can cause NIDDM. As a step toward understanding the molecular basis of glucokinase (GK) gene regulation, we assessed the structure and regulation of the human GK gene β-cell-type promoter. The results of reporter gene analyses using HIT-T15 cells revealed that the gene promoter was comprised of multiple cis-acting elements, including two primarily important cis-motifs: a palindrome structure, hPal-1, and the insulin gene cis-motif A element-like hUPE3. While both elements were bound specifically by nuclear proteins, it was the homeodomain-containing transcription factor insulin promoter factor 1 (IPF1)/STF-1/PDX-1 that bound to the hUPE3 site: IPF1, when expressed in CHO-K1 cells, became bound to the hUPE3 site and activated transcription. An anti-IPF1 antiserum used in gel-mobility shift analysis supershifted the DNA protein complex formed with the hUPE3 probe and nuclear extracts from HIT-T15 cells, thus supporting the involvement of IPF1 in GK gene activation in HIT-T15 cells. In contrast to the insulin gene, however, neither the synergistic effect of the Pan1 expression on the IPF1-induced promoter activation nor the glucose responsiveness of the activity was observed for the GK gene promoter. These results revealed some conservative but unique features for the transcriptional regulation of the β-cell-specific genes in humans. Being implicated in insulin and GK gene regulations as a common transcription factor, IPF1/STF-1/PDX-1 is likely to play an essential role in maintaining normal β-cell functions.

PDX-1 Induces Insulin and Glucokinase Gene Expressions in αTC1 Clone 6 Cells in the Presence of Betacellulin

The pancreatic β- and α-cells are developmentally related to each other but reveal diverse gene expression patterns. Among the two important transcription factors for insulin gene expression, IEF1 is present both in α- and β-cells, but PDX-1/IPF1/STF-1/IDX-1, a homeodomain-containing transcription factor, is present in β-cells but not in α-cells. To elucidate the function of PDX-1 in the expression of β-cell-specific genes, we established stable αTC1 clone 6 (αTC1.6)-derived transfectants expressing PDX-1 and examined the changes in the gene expression patterns in them. The exogenous expression of PDX-1 in αTC1.6 cells alone could induce islet amyloid polypeptide (IAPP) mRNA expression in the cells but not the expression of insulin, glucokinase, or GLUT2 gene. However, when βcellulin was added to the medium, the PDX-1-expressing αTC1.6 cells, but not the control αTC1.6 cells, came to express insulin and glucokinase mRNAs. This did not occur with other growth factors such as epidermal growth factor, transforming growth factor α, and insulin-like growth factor I. GLUT2 mRNA remained undetectable in the PDX-1–expressing αTC1.6 cells. These observations demonstrate the potency of PDX-1 for the expression of the insulin, glucokinase, and IAPP genes and suggest that certain regulatory factors, which can partially be modified by βcellulin, also contribute to the β-cell specificity of gene expression.

Probucol preserves pancreatic β-cell function through reduction of oxidative stress in type 2 diabetes

Oxidative stress is induced under diabetic conditions and causes various forms of tissue damage in patients with diabetes. Recently, pancreatic beta-cells have emerged as a putative target of oxidative stress-induced tissue damage and this seems to explain in part the progressive deterioration of beta-cell function in type 2 diabetes. As a step toward clinical trial of antioxidant for type 2 diabetes, we investigated the possible anti-diabetic effects of probucol, an antioxidant widely used as an anti-hyperlipidemic agent, on preservation of beta-cell function in diabetic C57BL/KsJ-db/db mice. Probucol-containing diet was given to mice from 6 to 16 weeks of age. Immunostaining for oxidative stress markers such as 4-hydroxy-2-nonenal (HNE)-modified proteins and heme oxygenase-1 revealed that probucol treatment decreased reactive oxygen species (ROS) in pancreatic islets of diabetic animals. Oxidative stress is known to enhance apoptosis of beta-cells and to suppress insulin biosynthesis, but probucol treatment led to preservation of beta-cell mass and the insulin content. According to intraperitoneal glucose tolerance tests, the probucol treatment preserved glucose-stimulated insulin secretion and improved glucose tolerance at 10 and 16 weeks: insulin, 280+/-82 vs. 914+/-238 pmol/l (120 min, at 16 weeks; P<0.05); glucose, 44.6+/-2.4 vs. 35.2+/-2.6 mmol/l (120 min, at 16 weeks; P<0.05). Thus, our present observations demonstrate the potential usefulness of probucol for treatment of type 2 diabetes.

Demonstration of Two Different Processes of β-Cell Regeneration in a New Diabetic Mouse Model Induced by Selective Perfusion of Alloxan

To clarify the regeneration process of pancreatic β-cells, we established a new mouse model of diabetes induced by selective perfusion of alloxan after clamping the superior mesenteric artery. In this model, diabetes could be induced by the destruction of β-cells in alloxan-perfused segments, while β-cells in nonperfused segments were spared. Intraperitoneal glucose tolerance tests showed glucose intolerance, which gradually ameliorated and was completely normalized in 1 year with a concomitant increase of insulin content in the pancreas. Histological examination showed neoislet formation in the alloxan-perfused segment and the proliferation of spared β-cells in the nonperfused segment. In the alloxan-perfused segment, despite a marked reduction of islets in size and number at an early stage, both the number of islets, including islet-like cell clusters (ICCs), and the relative islet area significantly increased at a later stage. Increased single β-cells and ICCs were located in close contact with duct cell lining, suggesting that they differentiated from duct cells and that such extra-islet precursor cells may be important for β-cell regeneration in β-cell–depleted segment. In addition to β-cells, some nonhormone cells in ICCs were positive for nuclear insulin promoter factor 1, which indicated that most, if not all, nonhormone cells positive for this factor were β-cell precursors. In the nonperfused segment, the islet area increased significantly, and the highest 5-bromo-2-deoxyuridine–labeling index in β-cells was observed at day 5, while the number of islets did not increase significantly. This indicated that the regeneration of islet endocrine cells occurs mostly through the proliferation of preexisting intra-islet β-cells in the nonperfused segment. In conclusion, the regeneration process of β-cells varied by circumstance. Our mouse model is useful for studying the mechanism of regeneration, since differentiation and proliferation could be analyzed separately in one pancreas.

Expression of heparin-binding epidermal growth factor-like growth factor during pancreas development. A potential role of PDX-1 in transcriptional activation.

The development of the pancreas appears to be regulated by various growth factors. We report here the expression of heparin-binding epidermal growth factor (EGF)-like growth factor (HB-EGF) in the developing pancreas. Immunostaining of fetal and neonatal rat pancreata, in which endocrine cells are visible as cell clusters often associated with primitive ducts or ductular cells, revealed that most of the cluster-forming cells and primitive ducts or ductular cells express HB-EGF protein. In contrast, the exocrine pancreas lacked HB-EGF expression. Based on findings that the expression pattern was similar to that of the homeodomain-containing transcription factor PDX-1 (IDX-1/STF-1/IPF1) and that the regulatory region of the HB-EGF gene contained sequences similar to the PDX-1-binding A element, we examined whether PDX-1 could be a potential activator of HB-EGF gene expression. The results of reporter gene analyses suggested that the HB-EGF gene promoter is PDX-1-responsive and that the activity of the promoter in pancreatic beta cell-derived βTC1 cells depends on the PDX-1 binding site-like sequences. Gel-mobility shift analyses using an anti-PDX-1 antibody indicated that PDX-1 is a specific and dominant binding factor for an A element-like sequence in the HB-EGF gene. These observations suggest the possible involvement of HB-EGF in pancreas development. While PDX-1 is essential for pancreas development, HB-EGF may function as a mediator of PDX-1 and thus be involved in the development of the endocrine pancreas.

Induction of glycation suppresses glucokinase gene expression in HIT-T15 cells

Aims/hypothesis. Chronic hyperglycaemia in patients with Type II (non-insulin-dependent) diabetes mellitus often leads to a decline in glucose-responsive insulin secretion from pancreatic beta cells, a phenomenon called glucose toxicity. Upon hyperglycaemia, glycation reaction occurs in the beta cells and induces oxidative stress. To understand the molecular basis of the beta-cell glucose toxicity, we investigated the possible effects of glycation on the expression and enzymatic activity of glucokinase, which plays a crucial part in glucose-responsive insulin secretion.¶Methods. Glycation and reactive oxygen species were induced in HIT-T15 cells by treatment with d-ribose and effects on glucokinase gene transcription, glucokinase protein amount, glucose phosphorylation activity, and DNA-binding activities of putative glucokinase gene transcription factors were evaluated.¶Results. When glycation was induced in HIT-T15 cells, the activity of the human glucokinase gene beta-cell-type promoter was suppressed substantially (83 % reduction at 60 mmol/l d-ribose). Also, similar reductions in mRNA and protein amounts of glucokinase and in the Vmax of its enzymatic activity were observed. In agreement with the reduction in the promoter activity, the two major transcription factors of the glucokinase gene, the Pal-binding factor and PDX-1, reduced their binding to their target sequences in the glucokinase gene promoter in glycation-induced HIT cells. Because these effects of d-ribose were counteracted by aminoguanidine or N-acetylcysteine, reactive oxygen species, generated by the glycation reaction, appears to be involved in the phenomena.¶Conclusion/interpretation. The induction of the glycation reaction, which is known to occur in pancreatic beta cells in chronic hyperglycaemia, suppresses the glucokinase gene transcription and its enzymatic activity. Thus, hyperglycaemia-dependent inhibition of glucokinase activity could in part explain beta-cell glucose toxicity. [Diabetologia (1999) 42: 1417–1424]

Association of Methylenetetrahydrofolate Reductase Gene Polymorphism With Carotid Arterial Wall Thickening and Myocardial Infarction Risk in NIDDM

Hyperhomocysteinemia has been identified as an independent risk factor for arteriosclerotic diseases such as myocardial infarction (MI), peripheral arterial occlusive disease, and cerebrovascular disease (1^1). In contrast to the very low prevalence of homocysteinuria (a hereditary disease characterized by severe hyperhomocysteinemia, homocystinuria, and earlyonset atherosclerotic diseases), individuals with mild homocysteinemia are frequently observed. Recently, a significant portion of these individuals was shown to have thermolabile 5,10-methylenetetrahydrofolate reductase (MTHFR). MTHFR is an enzyme that catalyzes the reduction of 5,10-methylenetetrahydrofolate to 5-methyltetrahydrofolate and is thereby involved in the remethylation of homocysteine to methionine (5,6). The thermolability seems to be caused by a C-»T substitution at nucleotide 677 (Ala->Val substitution) of the MTHFR-encoding gene (7). As compared with individuals homozygous for the 677C allele, those homozygous for the 677C^T mutation, which account for > 10% of a French-Canadian population, revealed -70% decrease of the MTHFR activity (7,8). Those individuals also revealed a 30-50% increase of plasma homocysteine levels when the folate level was below the median value (8). As a step to understanding the pathophysiological role of the 677C-»T mutation in the MTHFR gene in the development of vascular disease in diabetes, we examined the effect of the mutation on carotid atherosclerosis assessed by high-resolution ultrasound B-mode imaging. The subjects in this study were 222 Japanese unrelated NIDDM subjects (163 male and 59 female, age 59.8 ± 8.4 years (40-75) (mean ± SD [range]). Those subjects with liver disease, renal dysfunction, and/or collagen disease were excluded. Fasting blood samples were drawn from the subjects, and serum total and HDL choles-

Incidence and time course of left ventricular dilation in the early convalescent stage of reperfused anterior wall acute myocardial infarction

The incidence and early process of left ventricular (LV) dilation in 52 patients with reperfused anterior wall acute myocardial infarction (AMI) were assessed. All patients achieved coronary reflow within 24 hours of the onset and had a patent infarct-related artery in the convalescent stage. Left ventriculography was performed at pre-reflow and 25 days (mean) later to determine LV end-diastolic volume (ml) with the area/length method. Short-axis echo images at the midpapillary muscle level were recorded at days 1, 7, 14, and 28 of the AMI. With use of the papillary muscles as the internal landmarkers, the LV wall was divided into the anterior and posterior segments, and length and thickness of each segment were determined. Among 52 patients, 10 (19%) had a > or = 20% increase in end-diastolic volume in the convalescent stage. Echocardiographic studies demonstrated that there were no significant changes in lengths and thicknesses of the anterior and posterior segments during follow-up study relative to his or her baseline value in 42 patients without LV dilation. In the patients with LV dilation, however, the anterior segment exhibited a mean increase of 25% in its length with a mean decrease of 21% in its thickness at day 7 relative to their baseline values, but no progressive expansion was observed after day 7. A mean increase of 7% in the posterior segment length without reduction in its thickness first became evident at day 28.(ABSTRACT TRUNCATED AT 250 WORDS)

Ubiquitous, but variable, expression of two alternatively spliced mRNAs encoding mouse homologues of transcription factors E47 and E12

Two basic helix-loop-helix (bHLH) transcription factors, E47 and E12, are involved in cell-specific gene expression as part of dimeric complexes which interact with the cis-acting motif E-box. Although both generated from a single gene (E2A) by means of alternative splicing, the structural difference in these bHLH regions between the two suggests that the two bHLH proteins may differ in some of their functions. As a step toward elucidating the individual implications of E47 and E12, we investigated the mRNA expression ratios of their homologues (A1 and kA1, respectively) in mouse tissues and cell lines. Both the A1 and kA1 mRNAs were ubiquitously expressed in all tissues examined. However, their ratios varied: e.g., skeletal muscle, 2.2 +/- 0.3 (mean +/- SE); spleen, 2.0 +/- 0.2; pancreatic islet cells, 1.2 +/- 0.2. The A1/kA1 ratios in the cell lines investigated were similar to those of their original tissues. In conclusion, the ubiquity in mRNA expression observed for both the E47 and E12 homologues in mouse provides support for their involvement in a broad range of transcriptional regulation. The variation in the A1/kA1 expression ratios, on the other hand, supports the idea that A1 (E47) and kA1 (E12) each have some unique roles in the functions of these E2A gene-encoded bHLH proteins.