Laser capture-microdissected islet transcriptomics

Type 1 Diabetes (T1D) is an autoimmune disease characterized by T cell-mediated destruction of insulin-producing pancreatic β-cells. The pathogenesis of T1D is not fully understood but involves development of autoantibodies (AAbs) followed by a progressive decline in first phase insulin response. Live imaging of T1D pancreatic slices has revealed β cell dysfunction irrespective of the acute presence or absence of CD3+ T cells. However, the mechanisms that drive this dysfunction in the prediabetic period remain unclear. In-situ longitudinal studies of human islet cell biology in the context of T1D are essentially impossible. Hence, we leveraged the availability of pancreas tissues from the Network for Pancreatic Organ donors with Diabetes (nPOD) program to phenotypically and transcriptionally characterize laser capture-microdissected islets across the natural history of T1D. Fresh frozen pancreas sections were obtained from donors with no diabetes (ND, n=16), single autoantibody positive (sAAb, n=6), multiple autoantibody positive (mAAb, n=5) and patients with Type 1 Diabetes (T1D, n=16). Blocks were selected based on immunohistochemistry staining and contained INS+ islets and/or T cell infiltration as defined by six or more CD3+ cells immediately touching the islet endocrine cells. Blocks were cut into thick (10 μm) serial (3-4) sections and placed onto PEN-slides (Leica) along with thin (4 μm) sections placed onto Superfrost slides (Thermofisher) before and after the thick sections. Thin pancreas sections were immunostained for insulin (INS), CD3, HLA class I and ATP5B to delineate islets with residual beta cells, T cell infiltration, antigen presentation and mitochondrial function. Thick sections were used to laser-microdissect single islets from dehydrated slides. Laser microdissections yielded 3-4 sections per islet captured into 0.6 mL RNAse-free tubes. Sets of 7-25 islets were laser-captured per donor. A total of 260 islets were obtained across the four different clinical phenotypes (ND=73, sAAb=40, mAAb=37, T1D=110). Microarray gene expression profiling of single islets was performed using the Affymetrix Human Gene 2.0 ST arrays.