Insulin micro-secretion in Type 1 diabetes and related microRNA profiles

Abstract

The aim of this cross-sectional study was to compare plasma C-peptide presence and levels in people without diabetes (CON) and with Type 1 diabetes and relate C-peptide status to clinical factors. In a subset we evaluated 50 microRNAs (miRs) previously implicated in beta-cell death and associations with clinical status and C-peptide levels. Diabetes age of onset was stratified as adult (≥ 18 y.o) or childhood (< 18 y.o.), and diabetes duration was stratified as ≤ 10 years, 10-20 years and > 20 years. Plasma C-peptide was measured by ultrasensitive ELISA. Plasma miRs were quantified using TaqMan probe-primer mix on an OpenArray platform. C-peptide was detectable in 55.3% of (n = 349) people with diabetes, including 64.1% of adults and 34.0% of youth with diabetes, p < 0.0001 and in all (n = 253) participants without diabetes (CON). C-peptide levels, when detectable, were lower in the individuals with diabetes than in the CON group [median lower quartile (LQ)-upper quartile (UQ)] 5.0 (2.6-28.7) versus 650.9 (401.2-732.4) pmol/L respectively, p < 0.0001 and lower in childhood versus adult-onset diabetes [median (LQ-UQ) 4.2 (2.6-12.2) pmol/L vs. 8.0 (2.3-80.5) pmol/L, p = 0.02, respectively]. In the childhood-onset group more people with longer diabetes duration (> 20 years) had detectable C-peptide (60%) than in those with shorter diabetes duration (39%, p for trend < 0.05). Nine miRs significantly correlated with detectable C-peptide levels in people with diabetes and 16 miRs correlated with C-peptide levels in CON. Our cross-sectional study results are supportive of (a) greater beta-cell function loss in younger onset Type 1 diabetes; (b) persistent insulin secretion in adult-onset diabetes and possibly regenerative secretion in childhood-onset long diabetes duration; and (c) relationships of C-peptide levels with circulating miRs. Confirmatory clinical studies and related basic science studies are merited.

Figure 1

(A) Segmental linear regression of detectable C-peptide levels versus Type 1 diabetes duration. Overall r = 0.45, slope before threshold p = 0.0005, slope after threshold p = 0.66; diabetes duration threshold 8.8 (95% CI 5.7, 11.9) years (p < 0.0001). (B) Comparison of proportion of detectable C-peptide in participants with diabetes stratified by age of diabetes diagnosis and duration. Significant positive trend in percentage of detectable C-peptide with diabetes duration in diabetes diagnosed below 18 y.o. (p = 0.01) and borderline negative trend in diabetes diagnosed ≥ 18 y.o. (p = 0.07).

Figure 2

Venn diagram of miRs showing significant expression (Ct) differences in participants with diabetes with versus without detectable C-peptide, participants with and without diabetes with C-peptide levels below and above median. Arrows indicating the miRs abundance difference in comparison to the reference group (: compared with undetectable for C-peptide status, and : compared with below median C-peptide levels in participants with and without diabetes respectively).