Keratin Intermediate Filaments as Regulators of β-cell Structural and Functional Integrity

Epithelial keratins (K) are the largest family of cytoskeletal intermediate filaments (IF) and participate in a variety of cell and tissue functions including epithelial polarity, targeting proteins and organelles, cell signalling, growth, proliferation and differentiation. Keratins, constituting type I and II, are also involved in cell stress and apoptosis pathways, and their defects are shown to cause or predispose to 80 different human diseases. Simple epithelial K8 and K18 play important roles in organs responsible for glycaemic control, as K8 null mice show a disrupted systemic blood glucose regulation, altered pancreatic β-cell ultrastructure and increase in glycogen synthesis in liver. Moreover, pancreatic islet K8/K18 are upregulated in experimental diabetes and reduced K8 levels in K8 heterozygote mice led to a higher type 2 diabetes incidence. Although these findings provide insights into the importance of keratins in β-cells, the specific functions of β-cell keratins remain ambiguous. This thesis aims to characterize the β-cell keratin network and uncover the β-cell specific functions of these proteins in basal and cell stress processes.

In study I, we found that K7, a less known type II keratin, is a major constituent of IF network in mouse β-cells. The keratin bundles, at least expressing K7/K18, are enriched near β-cell vertices and lateral edges, with fine filaments extending throughout the cytoplasm. In the absence of K8 in vivo, the bundles localizing to the lateral edges and cytoplasmic filaments become scarce. We showed that K18 is the only type I keratin in β-cells and subsequently, drives the expression of K7 and K8, with K7 more closely following the K18 levels. These findings were evident by disruptions of K7 and K8 filaments in MIN6 insulinoma cell overexpressing the filament assembly mutant hK18 R90C, and an unbalanced increase of K7 compared to K8 in islets of mice overexpressing human K18. We also demonstrate the involvement of K7 in diabetes cell stress response, as K7/K18 are upregulated and more broadly distributed throughout the β-cell cytoplasm in pre- and diabetic mice compared to control. This thesis further provides evidence on a relatively similar islet keratin upregulation in type 2 diabetic patients.

In study II, using a novel β-cell specific K8 deficient mouse (K8flox/flox; Ins-Cre), we uncovered the β-cell autonomous role of K8 in cell structural and functional integrity, and showed that K8 is a driver of STZ sensitivity in β-cells. The specific K8 deletion in K8flox/flox; Ins-Cre β-cells lead to a significant decrease in K18, while no major change is detected in the overall islet K7 levels. The K8 deletion and reduced K18 does not lead to histo-morphological changes in islet tissue, but cause the islets to be more fragile. Further analysis showed that K8 deletion cause disruptions in membrane robustness and a decrease in adhesion and polarity protein E-cadherin. This further renders the β-cells mechanically compromised, evident by activation of mechanosensory yes-associated protein. On ultrastructural level, K8flox/flox; Ins-Cre mice show alterations in β-cell mitochondrial morphology, and disruptions in glucose stimulated insulin response in vivo, while the systemic blood glucose regulation is unimpaired. Using K8flox/flox; Ins-Cre mice and in vitro models, we demonstrated that β-cell K8 is the regulator of plasma membrane glucose transporter 2 (GLUT2) targeting and intact filaments are necessary for this process. Subsequently, since GLUT2 is the STZ transporter in β-cells, a lower diabetes incidence and less islet damage in K8flox/flox; Ins-Cre mice after acute STZ exposure, showed K8 as the driver of STZ sensitivity in these cells.

In summary, the β-cell keratin network is a regulator of cell structural and functional integrity and is involved in β-cell response to diabetes stress.