Enteroviral infection as a cause for beta-cell failure and diabetes

Abstract

Autoreactive T-cell mediated destruction of insulin-containing beta-cells is the hallmark of Type 1 Diabetes (T1D). Although the disease is known for centuries, neither is the cause of onset clearly identified, nor has the worldwide raise of incidences been stopped. To provide a cure, it is mandatory to uncover the causes triggering the disease such as environmental and/ or genetic factors that induce the activation of the immune system. Based on epidemiological studies, previous virus infections, especially by Coxsackieviruses B (CVB) of the Picornaviridae family, are highly associated with the onset of T1D. However, the proof of causality between CVB infection and T1D onset is still missing in humans and evidences of beta-cell infection in human pancreata remain controversial.In this work, I aimed to identify the initial pathways to enterovirus-mediated beta-cell destruction and to develop a highly sensitive and specific method to characterize expression and localization of viral RNA in pancreatic tissue sections. Cultured human islets and beta-cell lines were used to analyze the first cellular mechanism during viral infection. From the tested viruses, only the enteroviruses (CVB3 and CVB4) induced apoptosis in infected beta-cells. Both CVBs also infiltrated, but with different ratios, alpha-cells that remained unaffected by the virus. The Pathogen-recognition receptors (PRRs) TLR3 and PKR were the first receptors that recognized the virus in beta-cells and TLR7 in alpha-cells, whereas MDA5 and RIG-I bound the genome at a much later time point. TLR3 and PKR had a dual role in beta-cells, as they triggered the first immune response and cytokine release, but also activated the survival pathways AKT and ERK, upon virus recognition. Both activated kinases promoted viral replication and counter-balanced the absence of each other. While through this work, it became clear, that enterovirus infection is causative for beta-cell destruction and dysfunction in human islets in vitro. But the question, whether virus infection indeed coincide with diabetes in vivo remained elusive. With a newly developed method I could detect enteroviral genomes in human T1D donor tissues. I adapted single molecular fluorescence in situ hybridization (smFISH) to screen for viral RNA molecules in formalin-fixed paraffin-embedded (FFPE) pancreatic tissue sections. Specifically designed oligonucleotide probes to detect a wide range of Coxsackieviruses were successfully tested in an array of cell lines and pathological tissues. Oligonucleotide probes show a specificity and sensitivity that exceed classical immunohistochemistry (IHC) assays and was comparable in its sensitivity to PCR. With the combination of smFISH and IHC I could detect and locate viral RNA in T1D donor tissues in a blinded concordance study. I confirmed former VP1 staining and showed that enteroviral genomes are located in the pancreas of patients with T1D, but not exclusively within insulin-containing cells. The results of this thesis strongly point towards viral induced T1D onset in humans. They represent a central step for diabetes care and provide groundwork for future preventive strategies.