Dietary Selenium Requirement for the Prevention of Glucose Intolerance and Insulin Resistance in Middle-Aged Mice.

ABSTRACT

BACKGROUND: Although dietary selenium (Se) deficiency or excess induces type 2 diabetes-like symptoms in mice, suboptimal body Se status usually causes no symptoms but may promote age-related decline in overall health. OBJECTIVES: We sought to determine the dietary Se requirement for protection against type 2 diabetes-like symptoms in mice. METHODS: Thirty mature (aged 4 mo) male C57BL/6J mice were fed a Se-deficient torula yeast AIN-93M diet supplemented with Na2SeO4 in graded concentrations totaling 0.01 (basal), 0.04, 0.07, 0.10, and 0.13 (control) mg Se/kg for 4 mo (n = 6) until they were middle-aged (8 mo). Droplets of whole blood were used to determine glucose tolerance and insulin sensitivity in the mice from ages 5 to 8 mo. Postmortem serum, liver, and skeletal muscle were collected to assay for selenoprotein expression and markers of glucose metabolism. Data were analyzed by 1-way ANCOVA with or without random effects for time-repeated measurements using live mice or postmortem samples, respectively. RESULTS: Compared with control, the consumption of basal diet increased (P < 0.05) fasting serum insulin (95% CI 52%, 182%) and leptin (95% CI 103%, 118%) concentrations in middle-aged mice. Dietary Se insufficiency decreased (P < 0.05) 1) glucose tolerance (13-79%) and insulin sensitivity (15-65%) at ≤0.10 mg Se/kg; 2) baseline thymoma viral proto-oncogene phosphorylation on S473 (27-54%) and T308 (22-46%) at ≤0.10 and ≤0.07 mg Se/kg, respectively, in the muscle but not the liver; and 3) serum glutathione peroxidase 3 (51-83%), liver and muscle glutathione peroxidase 1 (32-84%), serum and liver selenoprotein P (28-42%), and liver and muscle selenoprotein H (39-48%) and selenoprotein W (16-73%) protein concentrations at ≤0.04, ≤0.10, ≤0.07, and ≤0.10 mg Se/kg, respectively. CONCLUSIONS: Mice fed diets containing ≤0.10 mg Se/kg display impaired glucose tolerance and insulin sensitivity, suggesting increased susceptibility to type 2 diabetes by suboptimal Se status at levels ≤23% of nutritional needs.