Evaluation of islet hormone expression profiles and transitional endocrine phenotypes in post mortem pancreatic tissue from human donors with type 1 diabetes, cystic fibrosis and cystic fibrosis related diabetes

Abstract

All forms of diabetes are characterized by abnormalities in blood glucose regulation due to altered β-cell function of the pancreatic islets. Type 1 diabetes (T1DM) is associated with an autoimmune process targeting the islet β-cells. In cystic fibrosis related diabetes (CFRD), loss of β-cell function is associated with exocrine pancreas fibrosis secondary to duct obstruction. Until now, apoptosis due to autoimmunity has been considered the primary mechanism underlying insulin loss in T1DM. Prevalence of residual β-cells and thus, insulin-containing islets (ICIs), even in cases of long-standing diabetes suggests that loss of β-cell function, rather than death alone, may contribute to hyperglycaemia in T1DM. In type 2 diabetes, this β-cell dysfunction has increasingly been linked to transitional endocrine cell phenotypes due to loss of end-differentiated protein markers (de-differentiation) and/or expression of other, non-β-cell hormones (transdifferentiation). Evaluation of differences in islet hormone expression in pancreata with and without diabetes is central to elucidating such phenotypic shifts that may be underlying diabetes development. The studies within this thesis aimed to quantify differences in islet hormone expression profiles and determine transitional endocrine phenotypes in normal and diseased pancreas by immunofluorescence (IF) staining of pancreatic tissue sections derived from deceased donors with T1DM, cystic fibrosis (CF) and CFRD in comparison to control donors. In Chapter 3, the aim was to quantify changes in pancreatic islet hormone expression profiles and explore any evidence of transitional phenotypes in T1DM pancreas from two distinct cohorts showing differential insulitic (islet immune infiltration) patterns. The two cohorts of T1DM patients were identified based on age of disease onset: Cohort 1 (>13 years old) and Cohort 2 (<7 years old). A significant decrease in the number of ICIs in T1DM compared to their age-matched controls was observed but endocrine cell number in remaining T1DM islets was comparable to agematched control donors across both cohorts. Reduced β-cell number was mirrored by an increased number of cells expressing non-β-cell hormones in keeping with potential trans-differentiation events. Moreover, polyhormonal and ‘hormone-empty’ cells were iii identified in patients with T1DM potentially evidencing β-cell trans- and dedifferentiation events. To overcome the inevitable time restrictions and potential for subjective bias intrinsic to manual quantification of tissue immunostaining phenotypes, an automated method of image analysis for high throughput quantification of islet cell phenotypes was established and validated in Chapter 4. Automated assessment using Vectra slide scanner and analysis by inForm® software was carried out on each T1DM donor to compare with manual analysis. The two methods were shown to be comparable, but validation confirmed that a minimum of 50 islets are required for quantitative sampling to match manual quantification. Moving forward with this approach will enable timeefficient sampling of much larger numbers of islets ensuring that outcomes are representative of the whole organ even when underlying pathology is characterised by its heterogeneity. In Chapter 5, the aim was to evaluate CF transmembrane conductance regulator (CFTR) expression in normal human pancreata and assess islet hormone expression in CF and CFRD. Whilst it is established that β-cell dysfunction in CF can lead to diabetes, the mechanism by which the CFTR channel influences insulin secretion remains debated. Thus, determination of the localisation of CFTR RNA and protein in normal human pancreata using sophisticated techniques was carried out to determine if CFTR influences β-cell function through cell-intrinsic or extrinsic mechanisms. It was observed that CFTR is absent within β- or any other islet endocrine cell types strongly suggesting that CFTR impacts β-cell function through non-cell autonomous derived factors. On assessment of CF and CFRD pancreata, a decrease in β-cells compared to age-matched controls was observed. Moreover, in CF/CFRD pancreata, abnormal endocrine cell distribution was observed within ducts with apparent budding of islets from the ductal epithelium. Quantification of hormone expression within these ductal regions and ductuloinsular complexes revealed a very high number of non-β-hormone producing cells. Endocrine cells in the ducts were found to be mostly glucagon-positive and ‘hormone-empty’ cells, implying possible attempted β-cell regeneration through intermediate phenotypes. Together these studies have confirmed maintenance of significant numbers of endocrine cells in T1DM and CFRD and have provided clear evidence of transitional phenotypes supporting a role for cell plasticity as opposed to death alone in diabetes pathogenesis. This opens the exciting possibility that, by controlling different stressors central to each type of diabetes development, restoration and renewal of β-cells is not impossible but a goal requiring active pursuit towards curative therapies for this devastating disease.