Search

The effects of genetic variation on molecular functions predisposing to type 2 diabetes are still largely unknown. Here, in a specifically designed diabetes model, we couple separate gene loci to mechanisms of beta-cell pathology. Niddm1i is a major glucose-controlling 16-Mb region in the diabetic GK rat that causes defective insulin secretion and corresponds to loci in humans and mice associated with type 2 diabetes. Generation of a series of congenic rat strains harboring different parts of GK-derived Niddm1i enabled fine mapping of this locus. Congenic strains carrying the GK genotype distally in Niddm1i displayed reduced insulin secretion in response to both glucose and high potassium, as well as decreased single-cell exocytosis. By contrast, a strain carrying the GK genotype proximally in Niddm1i exhibited both intact insulin release in response to high potassium and intact single-cell exocytosis, but insulin secretion was suppressed when stimulated by glucose. Islets from this strain also failed to respond to glucose by increasing the cellular ATP-to-ADP ratio. Changes in beta-cell mass did not contribute to the secretory defects. We conclude that the failure of insulin secretion in type 2 diabetes includes distinct functional defects in glucose metabolism and insulin exocytosis of the beta-cell and that their genetic fundaments are encoded by different loci within Niddm1i. [ABSTRACT FROM AUTHOR]

Type 2 diabetic patients often die because of end-stage renal failure, but no definitive reliable factor predicting long-term renal outcome has been identified. We tested whether a renal arterial resistance index (R/I) > or =80, using Doppler ultrasound technique, was predictive of worsening renal function. The primary end points of the study were 1) the course of glomerular filtration rate (GFR) and 2) the albumin excretion rate in 157 microalbuminuric, hypertensive, type 2 diabetic patients after a 7.8-year follow-up period (range 7.1-9.2). Kaplan-Meier curves for the primary end point (decrease of GFR > or = -3.0 ml/min per 1.73 m(2) per year) was two to three times more frequently observed in patients with R/I > or =80. Four- to fivefold fewer patients showed a regression to normoalbuminuria during the follow-up period from baseline microalbuminuria in the cohort with R/I > or =80. Overt proteinuria did develop in 24% of patients with R/I > or =80 and in 5% of patients with R/I <80 (P < 0.01). In conclusion, intrarenal arterial resistance appears to play a nontrivial role in deteriorating renal function in type 2 diabetic patients. R/I is a noninvasive diagnostic procedure, which strongly predicts the outcome of renal function in type 2 diabetic patients, even when GFR patterns are still normal.

We studied the following in normo- and microalbuminuric hypertensive type 2 diabetic patients: 1) transcapillary escape rate of albumin (TERalb) and 2) expression of mRNA slit diaphragm and podocyte proteins in renal biopsies. Normoalbuminuric subjects had renal cancer, and kidney biopsy was performed during surgery. TERalb was evaluated by clearance of (125)I-albumin. Real-time PCR of mRNA slit diaphragm was measured in kidney specimens. Albumin excretion rate (AER) was by definition lower in normoalbuminuric subjects than in microalbuminuric subjects with typical diabetic glomerulopathy (group 1), in microalbuminuric subjects with normal or near-normal glomerular structure (group 2), and in microalbuminuric subjects with atypical diabetic nephropathy (group 3). This classification was based on light microscopy analysis of renal tissue. TERalb (%/h) was similar in normoalbuminuric and microalbuminuric group 1, 2, and 3 diabetic patients (medians: 14.1 vs. 14.4 vs. 15.7 vs. 14.9, respectively) (ANOVA, NS). mRNA expression of slit diaphragm proteins CD2AP, FAT, Actn 4, NPHS1, and NPHS2 was higher in normoalbuminuric patients than in microalbuminuric patients (groups 1, 2, and 3) (ANOVA, P < 0.001). All diabetic patients had greater carotid artery intimal thickness than normal control subjects using ultrasound technique (ANOVA, P < 0.01). In conclusion, the present study suggests that microalbuminuria identifies a subgroup of hypertensive type 2 diabetic patients who have altered mRNA expression of slit diaphragm and podocyte proteins, even before glomerular structure shows abnormalities using light microscopy analysis. On the contrary, altered TERalb and increased carotid artery intimal thickness are shown by all hypertensive type 2 diabetic patients, both with normal and altered patterns of AER.

We have previously reported suggestive linkage to chromosome 5p13-q13 in type 1 diabetic families. ISL1, a transcription factor involved in pancreas development, maps to this region. Sequencing of the ISL1 gene in patients and control subjects identified seven single nucleotide polymorphisms (SNPs) and one microsatellite in noncoding regions. Four haplotypes formed by six of these SNPs and one microsatellite were associated with type 1 diabetes in Swedish families (P < 0.04). To identify possible interactions with the 5q11-q13 region, we applied pathway-restricted linkage analysis by analyzing for effects from regions encoding other transcription factors that are active during pancreas development and maintenance of insulin production. Linkage analysis allowing for interaction between 5q11-q13 and 7q32 resulted in an increase of logarithm of odds from 2.2 to 5.3. This increase was estimated to correspond to a P value <0.0016 using permutation. The transcription factor PAX4 is located at 7q32 and participates downstream of ISL1 in the transcription factor cascade critical to beta-cell development. Association with type 1 diabetes was also observed using the transmission disequilibrium test for two haplotypes at the PAX4 locus (P < 0.05). We conclude that pathway-restricted linkage analysis assists in the identification of possible gene-gene interactions and that 5q11-q13 and 7q32 together constitute a significant susceptibility factor for type 1 diabetes.

The major diabetes quantitative trait locus (Niddm1), which segregates in crosses between GK rats affected with spontaneous type 2-like diabetes and normoglycemic F344 rats, encodes at least two different diabetes susceptibility genes. Congenic strains for the two subloci (Niddm1f and Niddm1i) have been generated by transfer of GK alleles onto the genome of F344 rats. Whereas the Niddm1f phenotype implicated insulin resistance, the Niddm1i phenotype displayed diabetes related to insulin deficiency. Individual islets from 16-week-old congenic rats were characterized for insulin release and oxygen tension (pO(2)). In the presence of 3 mmol/l glucose, insulin release from Niddm1f and Niddm1i islets was approximately 5 pmol. g(-1). s(-1) and pO(2) was 120 mmHg. Similar recordings were obtained from GK and F344 islets. When glucose was raised to 11 mmol/l, insulin release increased significantly in Niddm1f and F344 islets but was essentially unchanged in islets from GK and Niddm1i. The high glucose concentration lowered pO(2) to the same extent in islets from all strains. Addition of 1 mmol/l tolbutamide to the perifusion medium further increased pulsatile insulin release threefold in all islets. The pulse frequency was approximately 0.4 min(-1). alpha-Ketoisocaproate (11 mmol/l) alone increased pulsatile insulin release eightfold in islets from Niddm1f, Niddm1i, and control F344 rats but had no effect on insulin release from GK islets. These secretory patterns in response to alpha-ketoisocaproate were paralleled by reduction of pO(2) in Niddm1f, Niddm1i, and control F344 islets and no change of pO(2) in GK islets. The results demonstrate that Niddm1i carries alleles of gene(s) that reduce glucose-induced insulin release and that are amenable to molecular identification by genetic fine mapping.