Uncovering the natural history of cystic fibrosis progression within human pancreas and the potential role of the stellate cell as a mediator of β-cell dysfunction in cystic fibrosis-related diabetes

Abstract

Cystic fibrosis (CF) is the commonest life-shortening autosomal recessive inherited disease. It is caused by mutations in the CF transmembrane conductance regulator gene (CFTR). This encodes the CFTR protein, which forms an ion channel mediating the secretion of chloride ions, sodium absorption and water transport across the cell membrane in epithelial cells. Loss of CFTR function leads to thickened fluid secretions particularly affecting the lungs and pancreas. Cystic fibrosis-related diabetes (CFRD) is a common non-pulmonary co-morbidity in CF affecting 2% of children increasing to 20% in adolescents and 50% in adults. Yet, the underlying pathogenesis is not fully understood. To understand how pancreatic changes during CF may impact on the islet compartment leading to CFRD development, this project aimed to investigate the morphological changes within CF pancreata as the disease progresses. Also, it explored whether a semi-quantitative scoring of histopathological changes in CF, including duct dilatation, fibrosis, and fat replacement could meaningfully elucidate the course of the disease. AI quantitative analysis (HALO) was employed to further quantify the changes in exocrine and endocrine compartments of CF pancreata. Following on, this thesis aimed to establish an in vitro primary human islet stellate cell (ISC) model, which could mimic the cell-mediated ‘stress’ environment which islets are exposed to in CF and or CFRD pancreata. Subsequently, transcriptomic analysis of CF pancreatic tissue and EndoC-βH1 cells following treatment with ISC secretome was conducted. A total number of 29 human pancreata biopsied at post-mortem from individuals with CF with age range 0–27-year-old were provided from Exeter Archival Diabetes Biobank (EADB) and a pathology bank collected by Prof. Günter Klöppel, Technical University of Munich, Germany. Fifty-eight control human pancreata were obtained from EADB and Quality in Organ Donation (QUOD) donors with age range of 0-29 year-old. Pancreatic tissue was stained with haematoxylin and eosin (H&E). In addition, a sub-cohort was stained with Sirius Red / Fast Green (SRFG), and immunohistochemistry staining for four endocrine hormones, chromogranin A, inflammatory cell markers CD45 and CD68, endothelial cell marker CD31 and α-smooth muscle actin (α-SMA). CF pancreata were classified into: fibroatrophic pattern alone (CF Pattern 1); fibroatrophic and lipoatrophic pattern (CF Pattern 2); and lipoatrophic pattern alone (CF Pattern 3). Semi quantitative scoring revealed an increase in ductal dilation in later-stage CF in fibrotic regions, severe ductal loss in adipocyte-rich regions and an increase islet remodelling as the disease progresses from Pattern 1 to Pattern 3. AI-driven image analysis revealed islets to be increasingly disorganised with disease progression with significantly decreased insulin+ area [%]. In CF, peri-islet fibrosis was significant in CF Pattern 2 with associated α-SMA+ stellate cells. NanoString nCounter analysis in CF and control post-mortem tissue confirmed an increase in fibrosis-related gene expression including TGF-β1 in CF. As a potential mediator of pancreatic fibrosis, it was hypothesised that pancreatic stellate cells may secrete cytokines driving β-cell dysfunction in CF. In vitro primary human ISC model was established, characterised and ISC-secretome was shown to decrease EndoC-βH1 insulin content and secretion in the presence of high glucose. Deeper molecular phenotyping with bulk RNA sequencing revealed downregulation of β-cell transcription factors including MafA, NKX 6.1 and PDX1 in EndoC-βH1 treated with ISC secretome. Collectively, these studies offered meaningful insights on pancreatic pathology natural history in CF, supporting a role of pancreatic stellate cells as potential mediators of endocrine dysfunction.