The N6-methyladenosine RNA epigenetic modification modulates the amplification of coxsackievirus B1 in human pancreatic beta cells.

Type 1 diabetes (T1D) is characterized by a prolonged autoimmune attack resulting in the massive loss of insulin-producing beta cells. The initiation and progression of T1D depends on a complex interaction between genetic, immunological and environmental factors. Epidemiological, experimental and clinical evidence suggest a link between viral infections, particularly Coxsackievirus type B (CVB), and T1D development. Specifically, infections by the CVB serotype 1 (CVB1) contribute to the triggering of autoimmunity against beta cells in genetically predisposed individuals, and prolonged and probably non-lytic infections by CVB are associated with the development of T1D. However, the molecular mechanisms underlying CVB1 replication and establishing persistent infections in human pancreatic beta cells remain poorly understood. Here we show that the N6-methyladenosine (m6A) RNA epigenetic modification machinery regulates CVB1 amplification in the human beta cells. Using small interfering RNA (siRNA) targeting m6A writers and erasers, we observed that downregulation of m6A writers increases CVB1 amplification, while the downregulation of m6A erasers decreases it. Notably, the inhibition of Fat Mass and Obesity-associated protein (FTO), a key m6A eraser, reduced by 95% the production of infectious CVB1 in both human insulin-producing EndoC-βH1 cells and in induced pluripotent stem cell (iPSC)-derived islets. The FTO inhibitor reduced CVB1 expression within 6 h post-infection, suggesting a direct regulation of the CVB1 genome by m6A modification. Furthermore, in the absence of viral replication, FTO inhibition also decreased the translation of the incoming CVB1 genome, indicating that m6A plays a critical role in the initial stages of viral RNA translation. In addition, modulation of the m6A machinery affected the type I interferon response after poly-IC transfection, a mimic of RNA virus replication, but did not affect the cellular antiviral response in CVB1-infected cells. Altogether, these observations suggest that m6A directly affects CVB1 production. Our study provides the first evidence that the m6A epigenetic modification machinery controls CVB amplification in human pancreatic beta cells. This suggests that the m6A machinery is a potential target to control CVB infection in T1D and raises the possibility of an epigenetic control in the establishment of persistent CVB infections observed in the pancreas in individuals with type 1 diabetes.