First phase insulin release in the non-obese diabetic mouse: correlation with insulitis, beta cell number and autoantibodies

Abstract

The ontogenic variation of beta cell function and its relationship with the degree of islet damage and levels of autoantibodies have been studied in the non-obese diabetic (NOD) mouse model. We conducted in vivo first phase insulin release (FPIR) in response to intravenous glucose and studied its correlation with the degree of insulitis, islet cell antibody (ICA) and insulin autoantibody (IAA) levels in female NOD mice cross-sectionally at days 40 ( n = 19), 90 ( n = 21), 150–160 ( n = 21) and day 250 ( n = 20). The mean ± SEM FPIR values showed an age-related decline from day 40 (46.2 ± 5.3 μU/ml) to day 150–160 (17.8 ± 2.5 μU/ml) and then doubled at day 250 (34.5 ± 5 μU/ml), while the mean ± SEM insulitis scores increased progressively until day 150–160 (61.7 ± 6.1%) and then declined slightly at day 250 to 50.2 ± 6.2%. In female NOD mice with spontaneous diabetes ( n = 4) and streptozotocin-induced diabetic Swiss mice ( n = 5) FPIR was either absent or greatly attenuated. A statistically significant inverse correlation between FPIR and insulitis was found among NOD mice at days 90 ( P = 0.02; r = −0.52) and 150–160 ( P = 0.03; r = −0.48). However, no statistically significant correlation was observed at days 40 and 250. Morphometric techniques applied to day 150–160 pancreatic sections showed a statistically significant negative correlation between insulitis and beta cell number per unit area of islet tissue ( P = 0.0001; r = −0.75). At this age some islet beta cells showed different intensities of staining by immunofluorescence. ICA or IAA levels did not show any correlation with insulitis or FPIR at the various age groups. The present data showing good correlation between FPIR values and insulitis at some age groups and not at others may be due to the relative imprecision of the in vivo IVGTT test or may indicate true beta cell pathogenesis during autoimmune destruction. These results shed further insights into the physiological basis for the use of the FPIR test for predicting insulin-dependent diabetes mellitus in humans.