Endothelial cell functions impaired by interferon in vitro: Insights into the molecular mechanism of thrombotic microangiopathy associated with interferon therapy.

INTRODUCTION: Interferon (IFN)-α and IFN-β approved for treatment of chronic hepatitis C viral infection and multiple sclerosis respectively have been linked to thrombotic microangiopathy (TMA) affecting renal function. Since the molecular mechanisms underlying this severe complication remain largely unclear, we aimed to investigate whether IFN affects directly in vitro endothelial cell functions associated with angiogenesis and blood haemostasis, as well as endothelial cell-derived vasodilators of nitric oxide (NO) and prostacyclin.

METHODS: Proliferation and survival of human umbilical vein endothelial cells (HUVECs) were measured by BrdU incorporation and alamarBlue assays. Angiogenesis was evaluated in co-cultures of HUVECs and human dermal fibroblasts. Fibrinolysis molecules were measured with ELISA. NO and prostacyclin were measured using a fluorescent NO-specific probe and a competitive enzyme immunoassay, respectively.

RESULTS: HUVEC proliferation was dose-dependently inhibited by IFN-β1a and IFN-β1b, but not by IFN-α2a and IFN-α2b. Consistently, IFN-β1a and IFN-β1b also reduced survival of HUVECs, but this again was not observed with IFN-α. However, both IFN subtypes inhibited VEGF-induced development of capillary-like structures, but the effect of IFN-α was less potent than IFN-β. In addition, both IFN subtypes upregulated interferon inducible protein 10 production from treated co-cultures while suppressing angiogenesis. Furthermore, intracellular NO generation was reduced by IFN-α2a and IFN-β1a, whereas prostacyclin release from HUVECs was not affected by IFN. Importantly, both IFN-β1a- and IFN-β1b-treated HUVECs showed a marked reduction in urokinase-type plasminogen activator release and a much greater secretion of plasminogen activator inhibitor-1 than tissue-type plasminogen activator compared with untreated cells, suggesting decreased fibrinolytic activity. IFN-α, however was less effective in modulating the fibrinolysis system.

CONCLUSIONS: We demonstrate the detrimental effects of IFN on endothelial cell functions mediated with angiogenesis and fibrinolysis, which could potentially cause the loss of physiological endothelium thromboresistance and facilitate the development of vascular complications in a clinical setting. Mechanistically, our findings have implications for understanding how IFN therapy can foster the development of TMA.