Abstract
Flaviviruses, including Zika virus (ZIKV), are a significant global health concern, yet no licensed antivirals exist to treat disease. The small membrane (M) protein plays well-defined roles during viral egress and remains within virion membranes following release and maturation. However, it is unclear whether M plays a functional role in this setting. Here, we show that M forms oligomeric membrane-permeabilising channels in vitro, with increased activity at acidic pH and sensitivity to the prototypic channel-blocker, rimantadine. Accordingly, rimantadine blocked an early stage of ZIKV cell culture infection. Structure-based channel models, comprising hexameric arrangements of two trans-membrane domain protomers were shown to comprise more stable assemblages than other oligomers using molecular dynamics simulations. Models contained a predicted lumenal rimantadine-binding site, as well as a second druggable target region on the membrane-exposed periphery. In silico screening enriched for repurposed drugs/compounds predicted to bind to either one site or the other. Hits displayed superior potency in vitro and in cell culture compared with rimantadine, with efficacy demonstrably linked to virion-resident channels. Finally, rimantadine effectively blocked ZIKV viraemia in preclinical models, supporting that M constitutes a physiologically relevant target. This could be explored by repurposing rimantadine, or development of new M-targeted therapies.