Machine learning-based reproducible prediction of type 2 diabetes subtypes.

AIMS/HYPOTHESIS: Clustering-based subclassification of type 2 diabetes, which reflects pathophysiology and genetic predisposition, is a promising approach for providing personalised and effective therapeutic strategies. Ahlqvist's classification is currently the most vigorously validated method because of its superior ability to predict diabetes complications but it does not have strong consistency over time and requires HOMA2 indices, which are not routinely available in clinical practice and standard cohort studies. We developed a machine learning (ML) model to classify individuals with type 2 diabetes into Ahlqvist's subtypes consistently over time. METHODS: Cohort 1 dataset comprised 619 Japanese individuals with type 2 diabetes who were divided into training and test sets for ML models in a 7:3 ratio. Cohort 2 dataset, comprising 597 individuals with type 2 diabetes, was used for external validation. Participants were pre-labelled (T2Dkmeans) by unsupervised k-means clustering based on Ahlqvist's variables (age at diagnosis, BMI, HbA1c, HOMA2-B and HOMA2-IR) to four subtypes: severe insulin-deficient diabetes (SIDD), severe insulin-resistant diabetes (SIRD), mild obesity-related diabetes (MOD) and mild age-related diabetes (MARD). We adopted 15 variables for a multiclass classification random forest (RF) algorithm to predict type 2 diabetes subtypes (T2DRF15). The proximity matrix computed by RF was visualised using a uniform manifold approximation and projection. Finally, we used a putative subset with missing insulin-related variables to test the predictive performance of the validation cohort, consistency of subtypes over time and prediction ability of diabetes complications. RESULTS: T2DRF15 demonstrated a 94% accuracy for predicting T2Dkmeans type 2 diabetes subtypes (AUCs ≥0.99 and F1 score [an indicator calculated by harmonic mean from precision and recall] ≥0.9) and retained the predictive performance in the external validation cohort (86.3%). T2DRF15 showed an accuracy of 82.9% for detecting T2Dkmeans, also in a putative subset with missing insulin-related variables, when used with an imputation algorithm. In Kaplan-Meier analysis, the diabetes clusters of T2DRF15 demonstrated distinct accumulation risks of diabetic retinopathy in SIDD and that of chronic kidney disease in SIRD during a median observation period of 11.6 (4.5-18.3) years, similarly to the subtypes using T2Dkmeans. The predictive accuracy was improved after excluding individuals with low predictive probability, who were categorised as an 'undecidable' cluster. T2DRF15, after excluding undecidable individuals, showed higher consistency (100% for SIDD, 68.6% for SIRD, 94.4% for MOD and 97.9% for MARD) than T2Dkmeans. CONCLUSIONS/INTERPRETATION: The new ML model for predicting Ahlqvist's subtypes of type 2 diabetes has great potential for application in clinical practice and cohort studies because it can classify individuals with missing HOMA2 indices and predict glycaemic control, diabetic complications and treatment outcomes with long-term consistency by using readily available variables. Future studies are needed to assess whether our approach is applicable to research and/or clinical practice in multiethnic populations.

Protective role of human insulin against the cytotoxicity associated with human mutant S20G islet amyloid polypeptide.

AIMS/INTRODUCTION: Islet amyloid polypeptide (IAPP) is a main component of islet amyloid in type 2 diabetes and cosecreted from ß-cell with insulin. Clinical evidence from the patients with S20G mutation of the IAPP gene, as well as experimental evidence that insulin could inhibit amyloid formation of IAPP, suggests that a gradual reduction of insulin could be related to the cytotoxicity associated with S20G-IAPP through long-term deterioration of ß-cells in type 2 diabetes. Our objective was to show an effect of human insulin on S20G-IAPP associated cytotoxicity. MATERIALS AND METHODS: We analyzed the cytotoxicity associated with S20G-IAPP by controlling human insulin expression using adenovirus vectors with micro ribonucleic acid specifically against human insulin in endocrine AtT-20ins cells, which express human insulin permanently. Additionally, we carried out a follow-up study of circulating IAPP and insulin in type 2 diabetic patients. RESULTS: S20G-IAPP expression was associated with a decrease in viability and an increase in terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate-biotin nick end labeling-positive cells in AtT-20ins cells. Furthermore, downregulation of human insulin enhanced the cytotoxicity associated with S20G-IAPP, and induced the cytotoxicity associated with wild-type (WT)-IAPP. Reduction of ubiquitin carboxy-terminal hydrolase L1 activity enhanced cytotoxicity under the downregulation of human insulin expression in both S20G- and WT-IAPP transduced cells. A 5-year follow up of type 2 diabetic patients showed a disproportionate increase of serum fasting IAPP-to-insulin ratio from baseline. CONCLUSIONS: Human insulin plays a protective role against the cytotoxicity associated with S20G-IAPP, as well as WT-IAPP. The findings could suggest long-term deterioration of insulin secretion associates with IAPP linked cytotoxicity in type 2 diabetes.

Kir6.2 E23K polymorphism is related to secondary failure of sulfonylureas in non-obese patients with type 2 diabetes.

AIMS/INTRODUCTION: The Kir6.2 E23K polymorphism was studied with a special reference to secondary sulfonylurea (SU) failure in non-obese patients with type 2 diabetes. MATERIALS AND METHODS: We recruited 278 non-obese (body mass index ≤30.0 kg/m(2)) Japanese patients with type 2 diabetes who had a history of SU treatment (for 11.2 ± 6.3 years) and compared the frequency of the secondary SU failure among the genotypes of the polymorphism. Genotyping of the Kir6.2 E23K was carried out by polymerase chain reaction-restriction fragment length polymorphism. RESULTS: The genotype frequencies of the polymorphism were similar to those previously reported in Japanese patients with type 2 diabetes. The frequency with which patients deteriorated into secondary SU failure was significantly higher in those with the KK genotype than those with EE or EK genotypes. Among 214 patients who eventually received insulin therapy because of secondary SU failure, the period of SU treatment in those with the KK genotype was significantly shorter than those with the EE or EK genotype, although the period from diagnosis to the start of SU treatment was not significantly different. CONCLUSIONS: These data suggest that the Kir6.2 E23K polymorphism is related to the acceleration of secondary SU failure in non-obese Japanese patients with type 2 diabetes.

[Clinical evaluation of insulin measuring kit not cross-react with insulin analogue preparations].

Clinical evaluation of insulin assay system reacting with only human insulin molecule (kit B) was performed by comparing it with conventional insulin assay system (kit A) cross-reacting with insulin analogue as well as human insulin preparation. In vitro, the kit B was confirmed to cross-react with only human insulin, not with insulin analogue preparations such as insulin aspart, lyspro and glargine. In non-insulin treated diabetic patients, postprandial and post-insulin injected serum immunoreactive insulin (IRI) concentrations measured by kit B were almost the same as those measured by the kit A. On the other hand, in diabetic patients treated with insulin analogue preparations, postprandial and post-insulin injected serum IRI levels measured by kit B were obviously low compared with those by kit A. After intravenous injection of insulin analogue preparations (0.1 unit/kg), insulin lyspro or insulin aspart, serum IRI levels measured by the kit B were not increased but gradually decreased in contrast to the obviously increased serum IRI level measured by the kit A. From these results, the kit B was confirmed not to measure the insulin analogue preparations in vitro and in vivo.

A mouse model for monitoring islet cell genesis and developing therapies for diabetes.

Transient expression of the transcription factor neurogenin-3 marks progenitor cells in the pancreas as they differentiate into islet cells. We developed a transgenic mouse line in which the surrogate markers secreted alkaline phosphatase (SeAP) and enhanced green florescent protein (EGFP) can be used to monitor neurogenin-3 expression, and thus islet cell genesis. In transgenic embryos, cells expressing EGFP lined the pancreatic ducts. SeAP was readily detectable in embryos, in the media of cultured embryonic pancreases and in the serum of adult animals. Treatment with the γ-secretase inhibitor DAPT, which blocks Notch signaling, enhanced SeAP secretion rates and increased the number of EGFP-expressing cells as assayed by fluorescence-activated cell sorting (FACS) and immunohistochemistry in cultured pancreases from embryos at embryonic day 11.5, but not in pancreases harvested 1 day later. By contrast, treatment with growth differentiation factor 11 (GDF11) reduced SeAP secretion rates. In adult mice, partial pancreatectomy decreased, whereas duct ligation increased, circulating SeAP levels. This model will be useful for studying signals involved in islet cell genesis in vivo and developing therapies that induce this process.

Autophagy protects against human islet amyloid polypeptide-associated apoptosis.

UNLABELLED: Aims/Introduction: Islets in type 2 diabetes are characterized by deposition of islet amyloid polypeptide (IAPP) as well as ß-cell dysfunction. The unique amyloidogenic character of human (h)IAPP is associated with cytotoxicity. Autophagy is a ubiquitous system of cellular recycling that contributes to cell survival. Thus, we examined whether autophagy could ameliorate hIAPP-associated cytotoxicity. MATERIALS AND METHODS: First, we used a COS-1 cell model, lacking endogenous IAPP that might affect cytotoxicity related to exogenous hIAPP. Next, we used the mouse ß-cell line, MIN-6 cells. Both cells were transfected with hIAPP or rat (r)IAPP expression constructs, or transfected with bicistronic vectors expressing green fluorescent protein (GFP) and either hIAPP or rIAPP for flow cytometry analysis. Cell viability and apoptosis markers were studied in relation to chemical or genetic modulation of autophagy. RESULTS: The viability of cells expressing hIAPP was significantly decreased as compared with those expressing rIAPP and the cleavage of pro-caspase-3 was elevated in hIAPP-transfected cells. The formation of autophagosomes and the conversion of microtubule-associated protein light chain 3B I to II were elevated in hIAPP-expressing cells. The viability of hIAPP-expressing cells was increased after treatment with rapamycin, an inducer of autophagy, and decreased after treatment with 3-methyladenine, an inhibitor of autophagy. In MIN-6 cells, annexin positive cells were increased by 3-methyladenine and decreased by rapamycin using flow cytometry. Knocking down of the autophagy protein 5 gene decreased hIAPP-transfected cell viability. CONCLUSIONS: Autophagy is co-localized with hIAPP expression and it plays a protective role in hIAPP-associated apoptosis. (J Diabetes Invest, doi: 10.1111/j.2040-1124.2010.00065.x, 2010).

Progressive deterioration of insulin secretion in Japanese type 2 diabetic patients in comparison with those who carry the S20G mutation of the islet amyloid polypeptide gene: A long-term follow-up study.

UNLABELLED: Aims/Introduction: In order to clarify the enhanced ß-cell dysfunction in type 2 diabetic patients carrying the S20G mutation of the islet amyloid polypeptide gene (S20G-patients), we first estimated the decline of insulin secretion in Japanese type 2 diabetic patients without the S20G mutation (non-S20G-T2D-patients) by long-term observation, and then compared it with that of the S20G-patients. MATERIALS AND METHODS: We followed 70 non-S20G-T2D-patients (body mass index <30 kg/m(2)) for more than 10 years and six S20G-patients for more than 5 years. We measured fasting C-peptide (F-CP) every 1-2 years and carried out a glucagon test at least once during the follow-up period. F-CP and a 5-min value of C-peptide after glucagon injection (5'-CP) were used as the indices of insulin secretion. We excluded patients who had renal dysfunction and/or anti-insulin antibodies in the insulin-treated patients. The individual annual declines were calculated from the slopes of the regression lines between C-peptide levels and duration (years after diagnosis). RESULTS: The mean individual annual declines of both F-CP and 5'-CP were significantly greater in the S20G-patients than the non-S20G-T2D-patients (F-CP; 0.047 ± 0.026 vs 0.011 ± 0.037 nmol/L/year, P = 0.025, 5'-CP; 0.139 ± 0.055 vs 0.022 ± 0.012 nmol/L/year, P = 0.008). CONCLUSIONS: We established the annual decline of insulin secretion in the Japanese type 2 diabetic patients by the long-term observation. The results show that the decline of insulin secretion is more rapid in the S20G-patients than the non-S20G-T2D-patients. (J Diabetes Invest, doi: 10.1111/j.2040-1124.2011.00102.x, 2011).

Chronic kidney disease has a more powerful impact on peripheral arterial disease than metabolic syndrome in Japanese type 2 diabetic patients.

BACKGROUND: Chronic kidney disease (CKD) and metabolic syndrome have been recognized as risk factors for cardiovascular disease. However, there is no information comparing their impact on macroangiopathy in diabetic patients. Thus, we studied the prevalence of CKD and metabolic syndrome in Japanese type 2 diabetic patients and then compared their impact on peripheral arterial disease (PAD) in type 2 diabetic patients. METHODS: This study focused on Japanese type 2 diabetic patients without hemodialysis (n = 1014). Patients with albuminuria, including microalbuminuria and/or an estimated glomerular filtration rate less than 60 mL/min/1.73(2), were diagnosed as having CKD. PAD was defined as ankle-brachial blood pressure index less than 0.9. RESULTS: The prevalence of CKD and metabolic syndrome was 47.1% and 39.6%, respectively. In four age- and duration-matched groups classified by the presence or absence of CKD and metabolic syndrome, the prevalence of PAD was significantly higher in groups with CKD alone than those with metabolic syndrome alone, and the high prevalence in the groups with CKD was not influenced by the coexistence with metabolic syndrome. CONCLUSIONS: This study indicates that CKD has more powerful impact on PAD than metabolic syndrome in type 2 diabetic patients.

The angiotensin II receptor blocker telmisartan improves insulin resistance and has beneficial effects in hypertensive patients with type 2 diabetes and poor glycemic control.

The angiotensin II receptor blocker (ARB) telmisartan has a molecular structure that confers it partial agonist properties similar to those of peroxisome proliferators-activated receptor-gamma molecule, which is thought to modulate tissue response to insulin. In order to investigate the effects of telmisartan on insulin sensitivity and glucose metabolism, we enrolled 14 hypertensive patients under treatment with ARB other than telmisartan who had insulin resistance [homeostasis model for insulin resistance (HOMA-IR)>2.0] but no severe glucose tolerance (HbA1c<6.5%), and HOMA-IR was compared before and after the displacement by telmisartan. We also enrolled 27 obese (body mass index>25kg/m(2)) and hypertensive patients with type 2 diabetes under treatment with ARB other than telmisartan, and HbA1c was assessed before and after the displacement by telmisartan. The telmisartan significantly improved HOMA-IR in hypertensive patients and also significantly decreased HbA1c in type 2 diabetic patients especially in the patients with poor glycemic control (HbA1c>==8.0%). These results indicate that telmisartan improves insulin resistance and gives beneficial effects in hypertensive patients with type 2 diabetes and a poor glycemic control.

[Clinical usefulness of measurement of urine type IV collagen for detection of early phase of nephropathy in type 2 diabetic patients].

Urine type IV collagen concentrations in type 2 diabetic patients were measured by enzyme immunoassay which has crossreactivity with only intact type IV collagen, and the clinical usefulness for estimating the early phase of diabetic nephropathy was evaluated. Precision of the measurement system was satisfactory for clinical use and the value did not influenced by the presence of sediments in urine. In whole type 2 diabetic patients (N=132), urine type IV collagen concentration (microg/g of creatinine) increased with development of nephropathy and showed significantly increase even in normoalbuminuria when compared with that in normal control subjects (N=117). In type 2 diabetic patients (N=100) with mild microalbuminuria (less than 100 mg/g of creatinine), multiple regression analysis revealed that HbA1C was extracted as a significant valuable for urine type IV collagen, while body mass index was extracted as a significant valuable for urine albumin. In these subjects, urine type IV collagen was significantly lower in the patients with good metabolic control (HbA1C<8.0%) than those with poor control (HbA1> or =8.0%), while the urine albumin was not significantly different between those two groups. These results suggest that measurement of urine type IV collagen in type 2 diabetic patients is useful for detection of early phase of diabetic nephropathy.

Prevalence of metabolic syndrome in Japanese type 2 diabetic patients and its significance for chronic vascular complications.

Prevalence of metabolic syndrome (MetS) in type 2 diabetes and its association with vascular complications were studied in 637 Japanese type 2 diabetic patients. MetS was diagnosed using criteria proposed by the Japanese study group for the definition of MetS in 2005. The prevalence of MetS in patients studied was higher in males (45.9%) than females (28.0%). The prevalence of MetS was 53.0% in males and 35.4% in females in patients with duration of less than 10 years, and decreased with an increase in duration. Upon comparing patients groups complicated with and without MetS, we determined the MetS group had significantly higher levels of fasting serum C-peptide and high-sensitivity C-reactive protein, and a significantly lower level of serum adiponectin. However, the prevalence of coronary heart disease, brain infarction, or peripheral arterial disease was not significantly different between these groups. On the other hand, the prevalence of microangiopathy in the group with MetS was significantly higher than in that without MetS, and became significantly higher along with an increase in duration. This study clarifies the prevalence of MetS in Japanese type 2 diabetic patients, and suggests that MetS is associated with microangiopathy rather than macroangiopathy in Japanese type 2 diabetic patients.

[Clinical usefulness for measurement of plasma brain natriuretic peptide in diabetic patients].

Plasma brain natriutetic peptide (BNP) concentrations in type 2 diabetic patients were measured by newly developed enzyme immunoassay, and their clinical application was evaluated. Precision of the measurement system was satisfactory for clinical use, and the value obtained by this system had good correlation to that by radioimmunoassay. Tubes containing NaF in addition to EDTA, usually used for measurement of plasma glucose and HbA1c in diabetic patients, could be used for the collection of plasma sample. In 133 type 2 diabetic patients who had no symptom for heart failure, plasma BNP was elevated in those with ischemic heart disease and it was also significantly elevated even in the patients who had no ischemic change on double Master two-step exercise testing than that in control subjects. In 52 patients receiving the examination by cardiosonography, plasma BNP levels significantly correlated to the left ventricular mass index, and also had a significant correlation to peak flow velocity in early diastole/peak flow velocity in late diastole (E/A) ratio, one of a simple index for an asymptomatic diastolic heart failure. Multiple logistic regression analysis revealed that age and coronary heart disease were extracted as a significant valuable for plasma BNP level in type 2 diabetic patients. These results suggest that measurement of plasma BNP in type 2 diabetic patients was useful as a screening method for evaluating the latent deterioration of heart function such as asymptomatic ischemic heart disease.

Molecular scanning of the betacellulin gene for mutations in type 2 diabetic patients.

Betacellulin (BTC), a member of the epidermal growth factor (EGF) family, is an important factor in the growth and/or differentiation of pancreatic beta cells. In this point of view, we determined the transcriptional start site of the human BTC gene and screened the protein-coding region for mutations. The transcriptional start site was located 347 bp upstream from the translational initiation codon. After screening the protein coding exons (exons 1-5), we identified two novel missense mutations, Cys (TGC) to Gly (GGC) at codon 7 (C7G) and Leu (TTG) to Met (ATG) at codon 124 (L124M), and a single nucleotide substitution (-31c/t) in the intron 2. The C7G was located in the signal peptide and the L124M in the transmembrane domain and this Leu at codon 124 was conserved among human, bovine, rat, and mouse. The frequencies of these variants, however, were similar between type 2 diabetic patients (n = 228) and non-diabetic control subjects (n = 170). These data suggest that genetic variations in the protein-coding region of the human BTC gene are unlikely to be a major contributor to development of type 2 diabetes.

Cilostazol, a phosphodiesterase inhibitor, reduces microalbuminuria in the insulin-resistant Otsuka Long-Evans Tokushima Fatty rat.

We evaluated association between hyperinsulinemia/insulin resistance and microalbuminuria in the insulin-resistant Otsuka Long-Evans Tokushima Fatty (OLETF) rat. OLETF rats showed glomerular hyperfiltration (an increase in creatinine clearance and a decrease in fractional excretion of Na) and microalbuminuria at the insulin-resistant prediabetic stage, and both were related to expression of transforming growth factor (TGF)-beta(1) and extracellular matrix protein such as fibronectin and collagen (a(1)) IV. Cilostazol, a selective type III cyclic nucleotide phosphodiesterase (PDE) inhibitor, normalized glomerular hyperfiltration and microalbuminuria with a parallel decline of TGF-beta(1) and extracellular matrix protein mRNA expression. Cilostazol may be beneficial to lessen early glomerular nephropathy in a state of hyperinsulinemia/insulin resistance.

PAX4 mutation (R121W) as a prodiabetic variant in Okinawans.

We previously reported that a missense mutation at codon 121 (CGG(Arg) to TGG(Trp), R121W) of PAX4 may be associated with the onset of type 2 diabetes in Japanese. In this study, we determined the frequency of the R121W mutation of PAX4 and characterized the prodiabetic phenotype in a population-based study. Healthy 372 residents participated in annual health check-ups in Nishihara (Okinawa, Japan) and unrelated 193 type 2 diabetic patients from the outpatient clinic of Ryukyu University Hospital were enrolled. Diagnosis of diabetes was based on the 1997 American Diabetes Association criteria. The R121W mutation in PAX4 was genotyped by PCR-RFLP analysis. In healthy residents, R121W mutation was detected in 12 of 372 residents (3.1%). The prevalence of newly diagnosed type 3 diabetes (25% vs. 5%, p=0.004) and HbA(1c) (5.6+/-1.9 vs. 5.1+/-0.7, p=0.026) was higher in the variants than in the wild-types. The odds ratio of diabetes in the R121W variants was 5.98 with 95% confidence interval from 1.50 to 23.9. The R121W mutation was observed in 12 of the 193 type 2 diabetic patients (6.2%). Onset-ages of diabetes were earlier (37+/-10 vs. 47+/-13 years, p=0.010) and the rate of insulin user was two times higher (83% vs. 41%, p=0.005) in the variants. The R121W mutation in PAX4 is a predisposing factor for the development of type 2 diabetes in Okinawans.

Cytotoxic T-lymphocyte antigen-4 gene polymorphisms and human T-cell lymphotrophic virus-1 infection: their associations with Hashimoto's thyroiditis in Japanese patients.

Cytotoxic T-lymphocyte antigen-4 (CTLA-4) decreases the immune response of T cells by inactivating the signal that occurs with interaction between CD28 on T cells and B7 on antigen-presenting cells. Gene polymorphisms involving CTLA-4 promoter (-318 C/T), exon 1 (49 A/G), and exon 4 (microsatellite (AT)n) have been linked to Hashimoto's thyroiditis (HT) and other autoimmune diseases. HT also has a reported association with human T-cell lymphotrophic virus-1 (HTLV-1) infection. We investigated the occurrence of CTLA-4 polymorphisms in Japanese patients with HT with and without anti-HTLV-1 antibodies (HTLV-1 Ab). DNA samples from 143 patients with HT and 199 controls were subjected to polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis using the restriction enzymes, Bbv 1, Tse 1, and Mse 1. In the HTLV-1 Ab-positive group the exon 1 G allele was more frequent in patients with HT than in controls (67% vs. 53%, p = 0.0377), and in HTLV-1 Ab-negative group it was also frequent in patients with HT than in controls (68% vs. 53%, p = 0.0041). Frequency of the G allele in HT with HTLV-1 Ab was comparable to those without HTLV-1 Ab. Frequency of polymorphism in the promoter did not differ between patients with HT and controls, nor between controls with and without HTLV-1 Ab. HTLV-1 infection is not associated with CTLA-4 polymorphisms in either HT or controls. HTLV-1 infection is not regulated by genetic factor such as CTLA-4, and may affect occurrence of HT as an independent purely environmental factor.

Evidence for a deficient pancreatic beta-cell response in a rat model of hyperthyroidism.

To clarify mechanism behind the abnormal glucose tolerance, observed in hyperthyroidism, we studied genomic and nongenomic effects of thyroid hormone on insulin secretion using a rat model of hyperthyroidism. Male Sprague-Dawley rats were intraperitoneally injected with vehicle, low (100 microg/kg) or high dose (600 microg/kg) of thyroxin (T(4)) for 2 weeks. Rats treated with high dose, but not low dose, of T(4), showed an increase in serum T(3) levels, and a decrease in body weight as compared to control rats. In rats treated with either dose of T(4), fasting blood glucose levels were increased, but serum insulin levels were similar to those of controls. After an oral glucose load, blood glucose levels were increased in rats treated with high dose, but not low dose, of T(4). Serum insulin levels after the oral glucose load were decreased in rats treated with either dose of T(4). After an intravenous glucose load, blood glucose levels were comparable among groups, but serum insulin levels tended to be low in T(4)-treated rats. Steady-state blood glucose levels were comparable among groups. The insulin secretory responses to high glucose (20mM) or arginine (10mM) of the isolated pancreas was decreased in rats treated with high dose, but not low dose, of T(4). Mean insulin secretory response to glucose and arginine were decreased by 40.1% and by 60.4% in high-dose-T(4)-treated rats. Addition of T(3) in the perfusion medium decreased glucose-induced insulin release. Ratios of proinsulin mRNA levels to beta-actin mRNA were decreased in the islets of T(4)-treated rats (0.45 +/- 0.07 vs control 0.61 +/- 0.03, p < 0.05). Levels of TR (thyroid hormone nuclear receptor) alpha1 + cErb Aalpha2 mRNA, but not TRbeta1, were decreased in the pancreatic islets of T(4)-treated rats. Calculated islet area was increased, but the number of beta-cells determined immunohistochemically was not increased in T(4)-treated rats, nor the volume density of insulin positive islets. We concluded that a deficient pancreatic beta-cell response to glucose, rather than insulin resistance, was responsible for abnormal glucose tolerance in this model of hyperthyroidism. Thyroid hormone causes a decrease in glucose-induced insulin secretion. We observed nongenomic and genomic effects of thyroid hormone on glucose-induced insulin secretion.

Remission of Graves' hyperthyroidism and A/G polymorphism at position 49 in exon 1 of cytotoxic T lymphocyte-associated molecule-4 gene.

We studied whether a patient with Graves' disease will go into remission during antithyroid drug (ATD) treatment. Remission of Graves' hyperthyroidism is predicted by a smooth decrease in TSH receptor antibody (TRAb) during ATD treatment. Cytotoxic T cell lymphocyte-associated molecule-4 (CTLA-4) may play an important role in the development of Graves' hyperthyroidism and in its remission. We studied A/G polymorphism at position 49 in exon 1 of the CTLA-4 gene in 144 Japanese Graves' patients. We intended to reveal the possible association of CTLA-4 gene polymorphism with the remission of Graves' hyperthyroidism. All patients with Graves' disease were treated with ATD. Thyroid-stimulating antibody and TSH binding inhibitory Ig were measured as TRAb. We analyzed CTLA-4 genotypes and alleles with PCR. We calculated the frequencies of CTLA-4 genotypes and alleles. A significant increase in the frequency of the G allele was seen in Graves' patients compared with controls (P = 0.0095). Graves' patients were divided into three groups (A, B, and C) according to time of TRAb disappearance after the start of ATD treatment. In group A patients TRAb had disappeared within 1 yr after the start of ATD treatment, in group B TRAb had disappeared between the beginning of the second year and the end of the fifth year of treatment, and in group C TRAb continued to be positive after 5 yr of ATD treatment. The frequencies of the GG genotype and the G allele were significantly higher in group C patients with persistently positive TRAb over 5 yr of ATD treatment than in the other groups (P < 0.0001). Group C patients did not have the AA genotype. The periods of time until remission were significantly shorter in the AA genotype. Graves' patients with the G allele need to continue ATD treatment for longer periods.