Plasma small RNA profiling reveals a three-miRNA signature associated with early beta cell dysfunction across glucose tolerance stages.

aims/hypothesisType 2 diabetes mellitus is a multifactorial disease marked by progressive beta cell dysfunction; however, no reliable, easily measurable circulating biomarkers are currently available to track this decline. As microRNAs (miRNAs) tightly regulate beta cell physiology and are secreted and found in blood, we investigated whether specific circulating miRNAs reflect in vivo beta cell function in living donors.

methodsWe conducted a cross-sectional study on two independent cohorts: a discovery cohort comprising 78 individuals with different levels of glucose tolerance [23 with normal glucose tolerance (NGT), 22 with impaired glucose tolerance (IGT) and 33 with type 2 diabetes] and a validation cohort comprising 158 individuals (71 non-diabetic individuals and 87 with type 2 diabetes) who were administered an oral glucose tolerance test (OGTT) and/or a mixed meal test (MMT) with measurement of glucose, insulin and C-peptide. We profiled plasma RNA by small RNA sequencing, and analysed differences between glucose tolerance groups using DESeq2. We used linear regression analyses to determine the association between miRNA abundance and clinical and metabolic measures. We validated selected miRNAs using droplet digital PCR (ddPCR). Finally, we specified a set of features that were used as an input for a LASSO model to select a biomarker signature that was able to estimate beta cell rate sensitivity (RS) values as a measure of beta cell function.

resultsWe identified 11 miRNAs that were differentially expressed across the three glucose tolerance groups of the discovery cohort. Integrated analyses pinpointed a three-miRNA signature (miR-34a-5p, miR-1306-5p and miR-335-5p) that correlated with beta cell RS, which is an early indicator of insulin secretory impairment. We incorporated this panel of selected miRNAs with age and 1 h post-load glucose into a LASSO regression model that enabled discrimination of RS values (Spearman's ρ=0.43, p<0.05). We also confirmed model performance in an additional cohort of 71 non-diabetic individuals and 87 individuals with type 2 diabetes (Spearman's ρ=0.23, p<0.05). To provide proof of concept for potential clinical use, we substituted basal glucose for 1 h post-load glucose while retaining miRNA levels and age; this adaptation preserved the significance of the model in the discovery cohort (Spearman's ρ=0.46, p<0.05), and showed an interesting, but not significant, association in the validation cohort (Spearman's ρ=0.15, p=0.061). CONCLUSIONS/

interpretationThis three-miRNA signature combined with simple clinical parameters constitutes a promising biomarker for early beta cell dysfunction and type 2 diabetes progression that deserves further validation in additional cohorts.