We analysed an miRNA profiling panel in plasma of age-matched (same week of age) normoglycaemic and recently diabetic NOD mice (n = 5 each) and validated the findings in matched LCM islet endocrine cells and immune infiltrates (key resources are shown in electronic supplementary materials [ESM] Table 1). Differentially expressed miRNAs were further investigated: (1) for in situ expression according to insulitis severity; (2) in the cellular subsets particularly enriched for the miRNA(s) of interest; (3) in plasma samples from an additional group of recently diabetic (n = 12), age-matched normoglycaemic (n = 25), 4-week-old (n = 14) and 8-week-old (n = 14) prediabetic NOD mice; and (4) for changes in expression after an in situ targeted therapeutic intervention. Finally, we analysed plasma miR-409-3p expression in two independent human cohorts: a first cohort (‘Cohort 1’) of non-diabetic individuals (control group) (n = 25), autoantibody-negative relatives of individuals with type 1 diabetes (n = 9), and individuals recently diagnosed with type 1 diabetes (n = 18), whose blood samples were collected and processed following a strict standard operating procedure (SOP); and a second, historical, cohort group (‘Cohort 2’) of individuals without diabetes (control group) (n = 17), individuals recently diagnosed with type 1 diabetes (n = 23) and individuals with longstanding type 1 diabetes (n = 13), as well as individuals with rheumatoid arthritis (n = 18).
NOD mice were housed and inbred in the KU Leuven animal facility (Leuven, Belgium) since 1989 and kept under semi-barrier conditions, as described [17, 18]. Female mice were screened for diabetes by glycosuria (Diastix Reagent Strips; Bayer, Leverkusen, Germany) and venous blood glucose levels (Accu-Chek; Roche Diagnostics, Vilvoorde, Belgium). Mice were diagnosed as diabetic when positive for glycosuria and for two consecutive blood glucose measurements > 11.1 mmol/l. Recently diabetic NOD mice were 12 to 22 weeks old; age-matched normoglycaemic animals were used as controls. Non-obese resistant (NOR) mice were obtained from the Jackson Laboratory (Bar Harbor, ME, USA) and killed at around 21 weeks of age. NOD mice of 4 and 8 weeks of age were used to evaluate miR-409-3p expression during the prediabetic phase.
For anti-CD3 (aCD3) monoclonal antibody (mAb) therapy, recently diabetic NOD mice were treated for 7 consecutive days with increasing i.v. doses (day 1, 0.42 μg; day 2, 1.67 μg; days 3–7, 4.46 μg per mouse) of anti-mouse CD3 mAb (clone 145-2C11; BioXCell, West Lebanon, NH, USA) (key resources are shown in ESM Table 2). Blood was collected by heart puncture or submandibular bleeding and centrifuged at 2000 g for 10 min at 4°C to obtain plasma. Collected plasma was immediately stored at –80°C. Mice were bred and housed according to protocols approved by the Katholieke Universiteit Leuven Animal Care and Use Committee (Leuven, Belgium; project number 116/2015) and experiments complied with EU Directive 2010/63/EU for animal experiments.
In Cohort 1, individuals with recently diagnosed type 1 diabetes (< 1 year of diagnosis; n = 18) and autoantibody-negative non-diabetic first-degree relatives of individuals with type 1 diabetes (n = 9) and without diabetes (n = 25) were recruited in the outpatient diabetes centre, University Hospital UZ Leuven (Leuven, Belgium), and at the diabetes outpatient unit, Siena University Hospital (Siena, Italy) (Table 1). Informed written consent was obtained from all individuals enrolled in the study, which was approved by the local ethics committee (Comitato Etico Regionale per la Sperimentazione Clinica della Toscana – Area Vasta Sud Est – cod. INNODIA01).
All non-diabetic individuals were negative for islet autoantibodies and for signs of ongoing endocrine dysfunction or of other autoimmune diseases. A SOP was followed to collect plasma samples (see ‘Human donors – blood collection procedure’ section in ESM Methods). Additional plasma samples were analysed; these were derived from Cohort 2 composed of non-diabetic individuals (control group) (n = 17), individuals recently diagnosed with type 1 diabetes (< 1 year of diagnosis, n = 23) and individuals with longstanding type 1 diabetes (> 2 years of diagnosis, n = 13), in addition to individuals with rheumatoid arthritis (ESM Table 3). All individuals with type 1 diabetes were positive for at least one autoantibody. Autoantibodies were analysed as described previously .
Plasma RNA extraction
Total RNA, including miRNAs, was extracted from 50 μl NOD mouse plasma or from 100 μl human plasma using miRNeasy kit (Qiagen, Hilden, Germany) (see ESM Methods).
Extracellular circulating miRNA profiling and data analysis
miRNA profiling was performed using TaqMan miRNA array rodent microfluidic cards (Panel A v2.1; Life Technologies, Carlsbad, CA, USA). Megaplex reverse transcriptase reaction and miRNA arrays were performed according to the manufacturer’s protocols (Life Technologies). Samples were checked for housekeeping stability and haemolysis rate (see ESM Methods; ESM Fig. 1a,b).
LCM and insulitis grading
Islet endocrine cells and juxtaposed infiltrating immune cells were microdissected separately as described previously . LCM islet endocrine cells and immune infiltrates were pooled separately according to insulitis score: score 0 (no infiltration) and 1 (peri-insulitis) were pooled together; tissues with score 2 (infiltration in < 50% of the islet area) and 3 (infiltration in ≥ 50% of the islet area) were pooled separately (see ESM Methods; ESM Fig. 2).
Cell staining, flow cytometry and cell sorting
Single cell suspensions were prepared from pancreases in digestion medium (RPMI medium + 5% FCS, 2 mmol/l l-glutamine, 0.05 mmol/l β-mercaptoethanol, 100 U/ml penicillin, 100 mg/ml streptomycin, 1 mg/ml collagenase-VIII and 0.02 mg/ml DNase-I; Life Technologies) for 15 min in a shaking incubator at 150 rpm at 37°C. The phenotype analysis was performed with flow cytometry by staining the cells with anti-CD4, -CD8, -CD44 and -CD62L mAbs, all from eBioscience (San Diego, CA, USA) (ESM Fig. 3). Cells were sorted directly into Trizol-LS (Life Technologies). RNA was extracted, quantified and quality controlled as above.
miRNA and mRNA single-assay real-time PCR
To validate miRNA expression in the reverse transcription (RT)-PCR single-assay reactions, 1 ng of RNA extracted from LCM islet immune infiltrates or flow-sorted immune cells, RNA extracted from 50 μl NOD mouse plasma or 100 μl human plasma was reverse transcribed using the Megaplex protocol with pre-amplification, and analysed using real-time PCR through specific TaqMan miRNA assays in 96 well plates (all from Life Technologies). To analyse mRNA expression from LCM islet endocrine cells and immune infiltrates, reverse transcription reactions were performed using ImProm-II RT (Promega, Madison, WI, USA) and TaqMan gene expression assays (Life Technologies), adopting a pre-amplification step following the manufacturer’s protocol (see ESM Methods).
miR-409-3p absolute quantification
Tenfold serial dilutions from 10−2 to 10−8 nmol/l of the synthetic miR-409-3p (mirVana miRNA mimic, Life Technologies) were reverse transcribed and assayed in parallel with RNA extracted from human plasma samples (ESM Fig. 4). Single-assay real-time PCR reactions were carried out as reported above (see ESM Methods).
Prediction and pathway enrichment analysis of miRNA targets
Putative targets of miR-409-3p, as well as pathway enrichment analysis, were investigated employing miRWalk v.3.0 (http://mirwalk.umm.uni-heidelberg.de). The results were visualised in Cytoscape v.3.6.1. using the ClueGo plugin [20, 21].
Statistical analyses were performed using GraphPad Prism Version 7.00 (GraphPad Software, La Jolla, CA, USA). Two-tailed Mann–Whitney U tests were used for comparisons between groups; two-tailed Wilcoxon signed-rank tests were used for comparisons between paired samples. Data are reported as mean ± SEM. For all analyses, a p value of ≤0.05 was considered significant. Randomisation and blinding were not carried out.