Novel de novo mutation in sulfonylurea receptor 1 presenting as hyperinsulinism in infancy followed by overt diabetes in early adolescence

Abstract

Objective: Congenital hyperinsulinism, usually associated with severe neonatal hypoglycemia, may progress to diabetes, typically during the 4th decade of life in nonpancreatectomized patients. We aimed to genotype the ATP-sensitive K(+) channel in a 10.5-year-old girl presenting with overt diabetes following hyperinsulinism in infancy.

Research design and methods: A female aged 10.5 years presented with new-onset, antibody-negative diabetes (A1C 10.6%). She was born large for gestational age (5 kg) to a nondiabetic mother and developed frequent hypoglycemic episodes, which persisted until age 3 years and responded initially to intravenous glucose and later to oral sweets. Currently, she is fully pubertal and obese (BMI 30.2 kg/m(2)), with a partially controlled convulsive disorder (since age 1 year) and poor school performance. Glucose levels were >11.1 mmol/l throughout 72 h of continuous glucose monitoring, with low insulin secretion during intravenous glucose tolerance testing. KCNJ11 and ABCC8 mutation analysis was performed, and the mutation identified was characterized in COSm6 cells.

Results: A novel, de novo heterozygous ABCC8 sulfonylurea receptor (SUR)1 mutation (R370S) was identified in the patient's DNA but not in that of either parent. Cotransfection of Kir6.2 and mutant SUR1 demonstrate that the mutated protein is expressed efficiently at the cell surface but fails to respond to MgADP, resulting in minimal channel activity. Interestingly, the heterozygous channel (WT:R370S) responded well to glibenclamide, a finding that lead to the successful initiation of sulfonylurea therapy.

Conclusions: This new ABCC8 mutation is associated with neonatal hyperinsulinism progressing within 10 years to insulinopenic diabetes. Consistent with in vitro findings, the patient responded to sulfonylurea treatment. The mechanism causing the relatively rapid loss in beta-cell function is not clear, but it may involve mutation-induced increased beta-cell apoptosis related to increased metabolic demand.

FIG. 1.

Clinical studies and trials. A: Continuous subcutaneous glucose monitoring using the MiniMed CGMS Sensor. Measurements were made every 10 s and averaged over 5 min. During 48 h, hyperglycemia was constant, with no hypoglycemic events. B: IVGTT revealing a retained, although insufficient, insulin secretion capacity and failure to restore the euglycemia. C: A therapeutic trial with 1 mg repaglinide before meals. As an example of one out of two consecutive trials, the 2-h postprandial glucose levels following repaglinide administration decreased from 9.7–10.5 mmol/l to 5–8.3 mmol/l. D: A therapeutic trial with 5 mg glibenclamide twice daily lowered postprandial glucose levels from 8.3–11 mmol/l to 7.2–7.8 mmol/l.

FIG. 2.

The R370S mutation. A: Sequence analysis revealed a heterozygous G→C mutation in the SUR1 gene resulting in the substitution of arginine for serine at the 370th amino acid of SUR1 protein. B: Detection of the mutation in the proband, but not in either the parents or the healthy sibling, by primer mismatch Sdu1 restriction analysis of PCR-amplified genomic DNA. The mutant is digested, whereas the WT is not. Direct sequencing of DNA from both parents failed to reveal any evidence of mosaicism.

FIG. 3.

Expression studies. COS cells were untransfected (unt) or transiently transfected with Kir6.2 and SUR1 of WT, R370S, or WT and R370S at 1:1 equal molar cDNA ratio (WT:R370S) and subjected to various expression and functional analyses. A: Surface expression level of K_ATP channels assessed by chemiluminescence assays. The relative chemiluminescence units of untransfected, WT, and R370S samples are 234 ± 8 (n = 11), 4,788 ± 509 (n = 14), and 4,736 ± 485 (n = 15), respectively. The WT and R370S expression levels are not significantly different (P > 0.5). B: Channel response to metabolic inhibition assessed by ⁸⁶Rb⁺ efflux assays. Untransfected cells exhibited background efflux (<10% in 40 min), whereas cells expressing WT channels exhibited >90% efflux, which was inhibited by 1 μmol/l glibenclamide. Cells expressing the R370S mutant had near background flux levels, and cells expressing a mixture of WT and R370S showed intermediate flux level between WT and the mutant. C: Representative current traces of WT and R370S channels in response to MgADP stimulation obtained by inside-out patch clamp recording. In contrast to WT channels, which were activated by ADP, the mutant failed to be stimulated. D: Averaged response of channels to MgADP stimulation. The current in 0.1 mmol/l ATP or 0.1 mmol/l ATP/0.5 mmol/l ADP was expressed as percentage of those observed in K-INT solution. In 0.1 mmol/l ATP/0.5 mmol/l ADP, the percent current of WT, WT:R370S, and R370S was 64.3 ± 7.9 (n = 10), 23.4 ± 4.7 (n = 6), and 5.4 ± 2.3 (n = 7), respectively; the values of both WT:R370S and R370S are significantly lower than that of WT (P < 0.05 and 0.005, respectively). E and F: Same as C and D, except that channel response to diazoxide was examined. The percent current in 0.1 mmol/l ATP/0.3 mmol/l diazoxide for WT, WT:R370S, and R370S was 43.0 ± 5.2, 22.8 ± 3.0, and 3.9 ± 1.0, respectively. Again, both the simulated heterozygous and homozygous expression conditions gave significantly less response than WT (P < 0.001). G: Representative current traces showing channel response to 10 nmol/l glibenclamide in K-INT. Note: The inhibition was irreversible. H: Averaged channel response to 10 or 100 nmol/l glibenclamide. The percent current in 10 nmol/l glibenclamide was 53.3 ± 6.0 (n = 8), 70.6 ± 8.1 (n = 4), and 55.4 ± 4.8 (n = 5) for WT, WT:R370S, and R370S, respectively; these values are not significantly different (P > 0.1). In 100 nmol/l glibenclamide, the percent current was 34.0 ± 4.3 (n = 10), 28.4 ± 5.3 (n = 5), and 30.9 ± 4.3 (n = 5), respectively; the values are not significantly different (P > 0.5).