Pancreatic beta cell-specific expression of thioredoxin, an antioxidative and antiapoptotic protein, prevents autoimmune and streptozotocin-induced diabetes.

The cytotoxicity of reactive oxygen intermediates (ROIs) has been implicated in the destruction of pancreatic beta cells in insulin-dependent diabetes mellitus (IDDM). Thioredoxin (TRX), a redox (reduction/oxidation)-active protein, has recently been shown to protect cells from oxidative stress and apoptosis. To elucidate the roles of oxidative stress in the development of autoimmune diabetes in vivo, we produced nonobese diabetic transgenic mice that overexpress TRX in their pancreatic beta cells. In these transgenic mice, the incidence of diabetes was markedly reduced, whereas the development of insulitis was not prevented. Moreover, induction of diabetes by streptozotocin, an ROI-generating agent, was also attenuated by TRX overexpression in beta cells. This is the first direct demonstration that an antioxidative and antiapoptotic protein protects beta cells in vivo against both autoimmune and drug-induced diabetes. Our results strongly suggest that oxidative stress plays an essential role in the destruction of beta cells by infiltrating inflammatory cells in IDDM.

Malignant transformation in craniopharyngioma after radiation therapy: a case report and review of the literature.

OBJECTIVE: Craniopharyngioma is a benign epithelial tumor that is thought to arise from the remnant of the Rathke pouch. Malignant transformation in craniopharyngioma is extremely rare. Herein, we report a case of malignant transformation in craniopharyngioma after radiation therapy. MATERIALS AND METHODS: Histopathological and immunohistochemical analyses were carried out for specimens of the suprasellar tumor (from three resections, with the third surgery performed after radiation therapy). RESULTS: The resected tumors from the first and second surgeries comprised islands of loosely cohesive aggregates of epithelial cells, so-called stellate reticulum. At the periphery of the nests, palisaded columnar epithelium was observed. Wet keratins were scattered, and few mitotic figures were seen. The third surgical specimen was composed of irregular large nests of basaloid cells that had large, round to oval nuclei with prominent nucleoli, and mitotic figures were frequently seen (21/10 high power fields). In the center of the nests, eosinophilic ghost cells, resembling wet keratin, were observed. Accordingly, the diagnosis of malignant transformation in craniopharyngioma was made. Immunohistochemical studies revealed that the p53 protein was over-expressed in the malignant component, whereas its expression was much lower in the benign component. CONCLUSIONS: Similar to the ten previously reported cases of malignant transformation in craniopharyngioma, the present case occurred after radiation therapy. p53 protein overexpression was also observed in the earlier cases of malignant craniopharyngioma as well as in the present case (6/6 cases). We concluded that radiation therapy and p53 mutations could be involved in malignant transformation in craniopharyngioma.

Roughness analysis of the skin as a secondary evaluation criterion in addition to visual scoring is sufficient to evaluate ethnic differences in wrinkles.

Visual scoring has been used to evaluate ethnic differences in skin wrinkling, but it is not sufficient to fully evaluate those differences in wrinkles. We examined whether both the roughness analysis of the skin and visual scoring are sufficient to characterize ethnic differences in wrinkles in Japanese, Chinese and German women. One hundred and five Japanese, 96 Chinese and 90 German age-matched women participated in this study. The severity of their wrinkles in the skin at two sites at the periphery of the eye was evaluated by visual scoring using a photoscale and by roughness values obtained from three-dimensional analysis of skin replicas. Slight but significant differences were scarcely observed between Japanese and Chinese women as well as between Japanese and German women at the same age group using the visual scoring method. However, significant and clearer differences among those ethnic groups were observed using the roughness analysis of skin replicas. Below the eye, significant differences among those ethnic groups were observed using both visual wrinkle scoring and roughness analysis. However, the extent of increased roughness values with age was relatively small compared with the increased wrinkle scores. These results show that roughness analysis is more sensitive than the visual scoring method when comparing ethnic differences in wrinkles. We conclude that roughness analysis of the skin is an important secondary evaluation criterion to visual scoring necessary to evaluate ethnic differences of wrinkles.

Global and systematic demonstration for the practical usage of a direct in vivo measurement system to evaluate wrinkles.

The global and systematic demonstration for the practical usage of a direct three-dimensional in vivo measurement system (PRIMOS) to evaluate wrinkles was investigated. Ten repetitive measurements of the corner of the eye of a subject showed that the coefficient of variation (CV)% value was 7.0% in a typical line-length roughness parameter R(a) (the arithmetic mean of roughness), and that the CV% value in a typical surface area roughness parameter S(a) was 2.4%. The relationships between the roughness values obtained from the corners of the eye and the age or wrinkle scores of Japanese women aged 10-70 years was examined. The values of several roughness parameters within the evaluation line length or surface area increased with age and showed a good correlation coefficient (r > 0.743). Similar relationships between the wrinkle scores and the values of roughness parameters were observed (r > 0.699). The roughness values were widely distributed even in the same wrinkle score because the measurement areas were limited and the values of skin roughness, including the microreliefs and/or small warts, were included in the calculation. However, changes in roughness values are considerable following treatment with potent active ingredients such as retinoic acid, so that this in vivo evaluation method is sufficient to objectively evaluate wrinkles. We conclude that the direct three-dimensional analysis of wrinkles in vivo should become a popular method to objectively evaluate wrinkles in clinical tests of wrinkle-smoothing ingredients or following cosmetic surgery to provide evidence of quantitative results.

Suppression of hypertriglyceridemia of Ehrlich carcinoma-bearing mice by an antibiotic, ascofuranone.

Ehrlich ascites carcinoma-bearing mice exhibit hypertriglyceridemia. An antitumor antibiotic, ascofuranone, suppressed tumor-induced hypertriglyceridemia when administered i.p. even when no evident antitumor activity was observed without affecting the levels of free fatty acids, phospholipids, cholesterol, glucose, and total protein in plasma. Ascofuranone did not reduce plasma triglycerides of normal mice. Insulin and clofibrate, known modifiers of lipid metabolism, showed no significant suppression. Ascofuranone is also effective on solid tumor-induced hypertriglyceridemia. Another notable change of metabolism affected by tumor-bearing in the early stage where hypertriglyceridemia has not yet fully progressed is hypoglycemia. Although ascofuranone did not affect hypoglycemia, the suppressive effect on hypertriglyceridemia was more evident when ascofuranone was administered in the early stage than in the later stage. These results suggest that ascofuranone suppresses hypertriglyceridemia by specifically affecting the changes of host metabolism which is induced in the early stage of tumor bearing.

Emergency obstetric care in Pakistan: potential for reduced maternal mortality through improved basic EmOC facilities, services, and access.

OBJECTIVE: To ascertain and compare compliance with UN emergency obstetric care (EmOC) recommendations by public health care centers in Pakistan's Punjab and Northwest Frontier Province (NWFP) provinces. METHOD: Cross-sectional data were collected from July through September 2003 using UN process indicators. From each province, 30% of districts (n=19); were randomly selected; all public health facilities providing EmOC services (n=170) were included. RESULTS: The study found that out of 170 facilities only 22 were providing basic and 37 comprehensive EmOC services in the areas studied. Only 5.7% of births occurred in EmOC health facilities. Met need was 9% and 0.5% of women gave birth by cesarean section. The case fatality rate was a low 0.7%, probably due to poor record keeping. Access and several indicators were better in NWFP than in Punjab. CONCLUSION: Almost all indicators were below UN recommendations. Health policy makers and planners must take immediate, appropriate measures at district and hospital levels to reduce maternal mortality.

The inhibition of low density lipoprotein metabolism by transforming growth factor-beta mediates its effects on steroidogenesis in bovine adrenocortical cells in vitro.

Transforming growth factor-beta (TGF beta) has a differential effect on the growth and function of bovine adrenocortical cells in vitro. TGF beta inhibits basal as well as ACTH- or angiotensin II-stimulated steroid formation, with no evidence of change in cell growth. The major inhibitory effect of TGF beta occurs at a step before cholesterol formation, since treatment of adrenocortical cells with TGF beta decreased not only delta 4-steroid levels but also delta 5-steroid levels. The addition of cholesterol reverses the suppression of steroidogenesis induced by TGF beta. To determine the mechanism of this inhibition, the effect of TGF beta on low density lipoprotein (LDL) metabolism was investigated. Cells treated with TGF beta showed a significant suppression of [125I]iodohuman LDL ([125I]LDL) binding to the cell surface, followed by decreases in internalization and proteolytic degradation of [125I]LDL. Maximal inhibition of LDL metabolism was observed at a concentration of 1 ng/ml (4 X 10(-11) M) TGF beta. The stimulation of LDL metabolism by ACTH was also inhibited by TGF beta, and the inhibition observed correlated well with the inhibition of steroidogenesis. The inhibitory effect of TGF beta on [125I]LDL binding results from the decrease in the maximal LDL-binding capacity. The stimulation of LDL uptake induced by Bu2cAMP, cholera toxin, forskolin, and Ang II was also decreased by treatment with 1 ng/ml TGF beta. The specificity of this effect is quite high, since the inhibitory effects of TGF beta on LDL metabolism were not observed with either inhibin A or activin, two molecules that have considerable structural homology to TGF beta. We conclude that TGF beta specifically suppresses LDL metabolism in bovine adrenocortical cell cultures and that this step may mediate, at least in part, its role as a potent inhibitor of steroidogenesis.

Early induction of c-fos precedes increased expression of corticotropin-releasing factor messenger ribonucleic acid in the paraventricular nucleus after immobilization stress.

CRF plays a role in coordinating endocrine, physiological, and behavioral responses to stressful stimuli. Several kinds of stressors have been reported to induce an increase in CRF mRNA expression in the paraventricular nucleus of the hypothalamus (PVN). Recently, the expression of c-fos mRNA has shown promise as a useful tool for metabolic mapping at the cellular level, because various types of stimulation induce c-fos mRNA expression in specific neuron populations in various brain regions. The aim of the present study is to clarify a possible anatomical-temporal correlation between the early induction of c-fos and the enhanced expression of CRF mRNA after stress. Wistar male rats were exposed to immobilization stress for 60 min and killed before and 15, 30, 60, 90, 120, and 180 min after the beginning of immobilization. In situ hybridization was performed by hybridizing sections with 35S-labeled prepro-CRF and c-fos cRNA probes. Relative levels of CRF and c-fos mRNA were compared by estimating the number of grains over the PVN in emulsion-dipped autoradiograms. Rapid induction (within 15 min) of c-fos mRNA was noted in the parvocellular division of the PVN after immobilization stress. The level of c-fos mRNA peaked at 30 min, then gradually declined to the control level within 90 min after the beginning of stress [the number of grains over the PVN: control, 326 +/- 180; 15 min, 2091 +/- 680 (P less than 0.05 vs. control); 30 min, 3385 +/- 239 (P less than 0.05 vs. control)]. The distribution of c-fos mRNA was almost identical to that of CRF mRNA in the PVN. On the other hand, the time course of CRF mRNA induction was delayed to the c-fos mRNA expression. A significant increase in CRF mRNA levels was noted only 120 and 180 min after stress [the number of grains over PVN: control, 3868 +/- 221; 120 min, 5957 +/- 677 (P less than 0.05 vs. control); 180 min, 6600 +/- 450 (P less than 0.05 vs. control)]. The results demonstrate that increased expression of CRF mRNA is preceded by c-fos mRNA induction in the PVN after stress suggesting a role of c-fos in the activation of CRF gene expression.

The effect of glucose and free fatty acids on growth hormone (GH)-releasing factor-mediated GH secretion in rats.

We have examined the effect of glucose and FFA on GH-releasing factor (GHRF)-mediated GH secretion in rats under pentobarbital anesthesia. Hyperglycemia did not affect GH secretion induced by administration of 20, 100, and 200 ng GHRF/100 g body weight. In contrast, GH response to 50 ng GHRF/100 g body weight in lipid heparin-treated rats, which showed high plasma FFA levels, was significantly suppressed compared with the control group (plasma peak GH: control, 1526 +/- 263 ng/ml; lipid-heparin group, 377 +/- 69 ng/ml P less than 0.05, mean +/- SEM). This suppressive effect of FFA on GH secretion was abolished by pretreatment with antisomatostatin serum (ASS) (GH level at 4 min after GHRF administration: ASS-saline group, 1606 +/- 210 ng/ml; ASS-lipid-heparin group, 1531 +/- 174 ng/ml; mean +/- SEM). These results suggest that hyperglycemia does not change the GH response to GHRF and that elevation of plasma FFA suppresses GHRF-induced GH secretion by the stimulation of somatostatin secretion in rats.

The responses of plasma adrenocorticotropin and cortisol to corticotropin-releasing hormone (CRH) and cerebrospinal fluid immunoreactive CRH in anorexia nervosa patients.

Pituitary-adrenocortical responses to the iv injection of 100 micrograms synthetic ovine corticotropin-releasing hormone (CRH) were studied in 13 patients with anorexia nervosa, and the concentrations of immunoreactive CRH in cerebrospinal fluid were measured in 7 of them. Mean basal levels of plasma ACTH and cortisol were 32 +/- 5 pg/ml (+/- SEM) and 21.1 +/- 1.5 micrograms/dl, respectively. The latter value was significantly higher than that in age-matched normal women (P less than 0.005). The mean increments of plasma ACTH and cortisol in response to CRH injection in those 13 patients were 21 +/- 5 pg/ml and 5.3 +/- 1.7 micrograms/dl, respectively, significantly lower than those in normal women (58 +/- 6 pg/ml and 15.3 +/- 7.7 micrograms/dl, respectively; P less than 0.005). When 4 patients were reexamined after weight gains of between 3 and 22 kg, their responses to the CRH injection increased. The mean concentration of immunoreactive CRH in the cerebrospinal fluid of seven patients was 30.8 +/- 3.9 pg/ml (+/- SEM), which was higher than the value of 18.4 +/- 1.1 pg/ml (P less than 0.005) in control subjects with cervical spondylosis. These findings suggest the possibility that hypersecretion of CRH may occur in patients with anorexia nervosa.

The relationship between bone turnover and body weight, serum insulin-like growth factor (IGF) I, and serum IGF-binding protein levels in patients with anorexia nervosa.

Malnutrition is one of the risk factors for bone loss in patients with anorexia nervosa (AN). To clarify the effects of nutritional status on bone metabolism, we examined the relationship between serum levels of nutritional indicators [insulin-like growth factor I (IGF-I), IGF-binding protein-2 (IGFBP-2), and IGFBP-3] and markers for bone metabolism [serum osteocalcin and urinary excretion of C-terminal telopeptide of collagen type I (CrossLaps)] in 45 AN out-patients, including 8 severely malnourished patients who required hospitalization and iv hyperalimentation (IVH). Compared to healthy subjects, serum IGF-I and IGFBP-3 were lower, whereas IGFBP-2 was higher in out-patients who had a body mass index (BMI) less than 16.5 kg/m2. In these patients, urinary excretion of CrossLaps, a marker of bone resorption, was higher, whereas serum osteocalcin, a marker of bone formation, was lower than those in control subjects. All of these parameters were normal in patients whose BMI ranged from 16.5-18.5 kg/m2. Serum levels of osteocalcin correlated positively with BMI (r = 0.512; P<0.0001), IGF-I (r = 0.558; P<0.0001), and IGFBP-3 (r = 0.369; P<0.001) in AN out-patients. In the 8 severely malnourished AN patients, serum levels of IGF-I and osteocalcin significantly increased 3 and 7 days, respectively, after the start of a 5-week IVH therapy regimen and reached normal levels within 5 weeks, accompanied by still elevated urinary excretion of CrossLaps. The present study demonstrates that an improvement in nutritional status in AN patients during IVH therapy rapidly increases the serum IGF-I levels, followed by a progressive increase in osteocalcin, suggesting immediate start of bone formation. However, increased bone resorption appears to continue for at least 5 weeks.

Plasma GH responses to GHRH and insulin-induced hypoglycemia in man.

Changes in plasma GH levels in response to an intravenous bolus injection of 200 micrograms GHRH-44 or 0.1 U/kg body weight regular insulin were examined in normal men who were pre-treated with 200 micrograms GHRH-44 or 0.1 U/kg body weight regular insulin 120 min in advance. The prior bolus injection of GHRH-44 inhibited the plasma GH response to the subsequent administration of GHRH-44 whereas the plasma GH response to the subsequent injection of insulin was not influenced by the prior administration of GHRH-44. The prior administration of insulin attenuated the plasma GH response to the subsequently given GHRH-44. These results suggest that desensitization of GHRH receptors in somatotrophs and/or somatostatin hypersecretion induced by increase in plasma GH levels following the prior GHRH-44 administration may be involved in the mechanism by which the prior GHRH-44 administration or insulin-induced hypoglycemia suppressed plasma GH responses to the following GHRH-44 administration. Sudden suppression of somatostatin secretion, which causes rebound of GH secretion, may occur in insulin-induced hypoglycemia.

The somatostatin analog octreotide inhibits the secretion of growth hormone (GH)-releasing hormone, thyrotropin, and GH in man.

The effects of the somatostatin analog octreotide on plasma GH, TSH, and immunoreactive GH-releasing hormone (IR-GHRH) were studied in 10 normal men. After morning sc administration of 50 or 100 micrograms octreotide or placebo, plasma GH, TSH and GHRH were measured frequently for 6 h. Plasma GH or IR-GHRH concentrations did not change after placebo injection, but plasma TSH levels gradually decreased, in conformity with a circadian rhythm during the morning. The mean plasma GH levels after sc injection of 50 or 100 micrograms octreotide declined, and no spontaneous GH pulses occurred for 5 h. Plasma TSH decreased rapidly after both doses of octreotide and was significantly lower than the level after placebo treatment from 90-315 min (P less than 0.05) and 60-360 min (P less than 0.05 or P less than 0.01), respectively. Plasma IR-GHRH levels also were significantly lower from 30-360 min (P less than 0.05) in the group given 100 micrograms octreotide compared with the value in the placebo group. We conclude that octreotide inhibits not only GH and TSH secretion from the pituitary, but also GHRH release from the hypothalamus and/or peripheral tissues. These findings suggest that somatostatin controls GH secretion not only by suppressing pituitary secretion of GH but also by suppressing GHRH release from the hypothalamus.

Studies on the response of growth hormone (GH) secretion to GH-releasing hormone, thyrotropin-releasing hormone, gonadotropin-releasing hormone, and somatostatin in acromegaly.

The plasma GH response to GH-releasing hormone (GHRH), TRH, or GnRH administration was examined in 25 acromegalic patients. Plasma GH levels increased in 21 patients after GHRH, in 19 after TRH, and in 4 after GnRH. The four GHRH nonresponders had had acromegaly longer than had the GHRH responders. No specific combination of GH responsiveness to these 3 releasing hormones was found among the patients. Infusion of 1 mg GHRH for 150 min gradually increased plasma GH levels, with some fluctuations, from the beginning to the end of infusion in normal subjects and in 7 patients who were GHRH responders, but a bolus injection of 100 micrograms GHRH at the end of the infusion did not further elevate plasma GH levels. These results suggest that desensitization to GHRH occurred in the normal subjects and acromegalic patients. However, in 5 acromegalic patients who responded to both GHRH and TRH, a bolus injection of 500 micrograms TRH given at the end of the 150-min infusion of 1 mg GHRH evoked a further plasma GH rise. In 5 normal subjects and 2 patients who were responders to GHRH but not TRH, a bolus injection of 500 micrograms TRH did not cause plasma GH elevation at the end of 150-min infusion of 1 mg GHRH. These results imply that TRH and GnRH stimulate GH secretion from the adenoma cells in vivo through receptors different from those for GHRH. In vitro studies using cultured pituitary adenoma cells from 2 patients revealed that the responses of GH secretion to GHRH were similar to those in vivo. These data, therefore, suggest that the responsiveness of GH secretion to stimuli is determined by the specificity of the receptors on adenoma cells. The action of somatostatin-28 was more potent than that of somatostatin-14 in the suppression of GH secretion from adenoma cells.

Decreased serum levels of acid-labile subunit in patients with anorexia nervosa.

One of the observations in malnutrition is that serum insulin-like growth factor (IGF)-I levels are decreased, and this decrease is associated with an altered profile of IGF binding proteins (IGFBPs). In human circulation, IGFs are mostly present as an approximately 150-kDa ternary protein complex consisting of IGFs, IGFBP-3, and acid-labile subunit (ALS). In the present study, to clarify the effect of nutrition on serum ALS levels, we investigated 33 patients with anorexia nervosa. Serum levels of ALS were measured by RIA. Furthermore, we measured serum IGF-I, IGF-II, IGFBP-2, and IGFBP-3 levels in the patients. From these data, we investigated which was the best predictor of body mass index (BMI) as a nutritional status marker. In the patients with anorexia nervosa, the serum ALS levels ranged from 0.7-16.9, with a mean of 10.6 +/- 0.7 mg/L, and the levels were significantly lower than those of normal subjects (13.8 +/- 0.8 mg/L, P < 0.05). Serum ALS levels positively correlated with BMI (r = 0.41, P < 0.05), and the levels increased during treatment. The serum IGFBP-2 levels in the patients were increased (871 +/- 91 microg/L), and the levels inversely correlated with BMI (r = -0.52, P < 0.01). The serum IGF-I and IGFBP-3 levels were low (152 +/- 14 microg/L and 2.56 +/- 0.12 mg/L, respectively), and the levels positively correlated with BMI (r = 0.46, P < 0.01; and r = 0.39, P < 0.05, respectively). The serum IGFBP-2, IGF-I, and IGFBP-3 levels returned toward normal ranges as BMI in the patients improved during treatment. Serum IGF-II levels did not correlate with BMI (r = 0.24, P = 0.17). Stepwise regression analysis revealed that serum IGFBP-2 was the best marker of BMI among these variables. The present study suggested that ALS was regulated by nutritional status, the same as IGF-I, IGFBP-2 and IGFBP-3; but the serum IGFBP-2 was the best predictor of BMI as nutritional status marker among the parameters in patients with anorexia nervosa.

The effect of free fatty acids on growth hormone (GH)-releasing hormone-mediated GH secretion in man.

The effect of FFA on GH-releasing hormone (GHRH)-mediated secretion of GH was examined in six normal young men. Three of the men were infused with 250 ml of a lipid-heparin solution at 1.67 ml/min for 150 min, and the other three were given an equivalent volume of saline in the same manner. Thirty minutes after the start of infusion, 100 micrograms GHRH (the 44-amino acid form) were injected iv, and plasma GH and FFA were measured. One week later, the same men participated in an identical experiment, but the ones who had received lipid-heparin previously were given saline and vice versa. In both experiments, plasma FFA increased to 2.25 +/- 0.16 meq/liter (mean +/- SEM) 60 min after the start of lipid-heparin infusion, whereas FFA levels did not change significantly in the saline-treated group. Mean plasma GH levels reached peak concentrations in both groups 30 min after GHRH treatment. However, the peak GH response when lipid-heparin was given was significantly diminished (8.4 +/- 1.7 ng/ml), compared with the peak response when saline was given (28.9 +/- 7.1 ng/ml). These data suggest that plasma FFA elevations induced by lipid-heparin infusion inhibit GH secretion induced by GHRH.

Structure and hair follicle-specific expression of genes encoding the rat high sulfur protein B2 family.

High sulfur proteins are cysteine-rich proteins synthesized during the differentiation of hair matrix cells, and form hair fibers in association with hair keratin intermediate filaments. Rat high sulfur protein B2 genes were isolated after screening of a rat genomic library using the cDNA as a probe. Sequence analysis of a 4 kb fragment revealed two high sulfur protein genes, B2E and B2F. Both genes lacked introns, with B2F being located at 2 kb downstream of B2E. The 5' flanking regions of both genes had TATA and CAAT boxes, and consensus sequences of B2 genes. The upstream region of B2F had possible AP-1 and Sp-1 binding elements. The high sulfur protein B2E and B2F, which have putative 188 and 122 amino acids, respectively, comprised four distinct domains with a characteristic repetitive sequence. In situ hybridization indicated that the mRNA of high sulfur protein B2 was specifically localized in the cortex of the hair shaft, and northern blot analysis indicated that the expression of B2 increased in anagen and decreased in telogen, suggesting that high sulfur protein B2 synthesized in cortical cells during anagen contributes to the production of hair fibers.

Age-related association of MHC class I chain-related gene A (MICA) with type 1 (insulin-dependent) diabetes mellitus.

To assess the contribution of the HLA class I region to susceptibility to and heterogeneity of type 1 diabetes, we investigated the association of polymorphism of MHC class I chain-related gene A (MICA) with age-at-onset as well as susceptibility to type 1 diabetes. One hundred one Japanese patients and 110 healthy control subjects were studied. The frequency of A4 allele was significantly higher and that of A6 allele was significantly lower in patients than in control subjects. The frequency of A5.1 allele was highest in early-onset patients (23.0%), intermediate in intermediate-onset patients (9.2%) and lowest in late-onset patients (7.7%) (trend chi-squared test, p = 0.0098). A5. 1 allele was strongly associated with HLA-B7 and Cw7, suggesting that MICA*A5.1-B7-Cw7 haplotype contains a gene responsible for age-at-onset. A4 allele was associated with a susceptible haplotype, DR4-DQB1*0401, and A6 allele was associated with a protective haplotype, DR2-DQB1*0601, suggesting that the association of MICA with type 1 diabetes susceptibility may be due to linkage disequilibrium with class II haplotypes. These data suggest that MICA gene is associated with age-at-onset and that a gene (or genes) responsible for age-at-onset of type 1 diabetes is located in the HLA class I region, probably near the region of MICA-B-C.