Altered microRNA expression in individuals at high risk of type 1 diabetes

Type 1 diabetes (T1D) is an autoimmune disease characterized by the destruction of pancreatic insulin-producing ß cells. CD4+ T cells are integral to the pathogenesis of T1D, but biomarkers that define their pathogenic status in T1D are lacking. miRNAs have essential functions in a wide range of tissues/organs, including the immune system. We reasoned that CD4+ T cells from individuals at high risk for T1D (pre-T1D) might be distinguished by an miRNA signature. We sorted CD4+ T cells from 9 healthy and 7 pre-T1D individuals into 6 subsets, namely naïve, resting regulatory (rTreg), activated regulatory (aTreg), transitional memory (Ttm), central memory (Tcm) and effector memory (Tem) cells, and then compared miRNA profiles between these subsets and between pre-T1D and healthy individuals by deep sequencing. Differential expression of miRNAs was detected in each of the CD4+ T cell subsets. For example, expression of miRNAs that induce apoptosis (miR-15a) or FOXP3 instability (miR-31) was increased in rTreg and aTreg cells, respectively, in pre-T1D individuals, whereas miR-150 was increased in Tem cells of pre-T1D individuals. Importantly, increased miR-150 expression could be detected by qRT-PCR in total CD4+ T and PBMCs of pre-T1D individuals. Consistent with it being a marker of pathogenic CD4+ T cells, we showed that miR-150 regulates IFN-? production in mouse CD4+ T cells. Thus, comprehensive profiling identifies miRNA profiles that not only distinguish CD4+ T cell subsets but also discriminate individuals with preclinical T1D. The ability to detect differentially expressed miRNAs in total CD4+ T cells or PBMCs should facilitate clinical application of miRNAs as biomarkers. Overall design: CD4+T cells from healthy and individuals at high risk for autoimmune type 1 diabetes were sorted into 6 subsets, which resulted in 80 samples, 38 for healthy and 42 for high risk individuals. Each sample was barcoded and miRNA libraries were constructed and subsequently subjected to deep-sequencing on the Illumina GAII or HiSeq platform. The Fastq files are have deconvoluted and stripped of the barcode adaptor sequences.

80

Zhang Y, Feng ZP, Naselli G, Bell F, Wettenhall J, Auyeung P, Ellis JA, Ponsonby AL, Speed TP, Chong MM, Harrison LC