Reversion to wildtype of a mutated and nonfunctional coxsackievirus B3CRE(2C)

•Mutated coxsackievirus B3CRE(2C) is viable and reverts to wildtype in replication.•5′ terminal deletions occur in mutated CRE(2C) virus before CRE(2C) reversion.•Reversion occurs in a 5′ terminally deleted genome lacking 5′ UU.•These suggests the CRE(2C) has a function in terminally deleted genome replication. The cis-acting replication element (CRE) in the 2C protein coding region [CRE(2C)] of enteroviruses (EV) facilitates the addition of two uridine residues (uridylylation) onto the virus-encoded protein VPg in order for it to serve as the RNA replication primer. We demonstrated that coxsackievirus B3 (CVB3) is replication competent in the absence of a native (uridylylating) CRE(2C) and also demonstrated that lack of a functional CRE(2C) led to generation of 5′ terminal genomic deletions in the CVB3 CRE-knock-out (CVB3-CKO) population. We asked whether reversion of the mutated CRE(2C) occurred, thus permitting sustained replication, and when were 5′ terminal deletions generated during replication. Virions were isolated from HeLa cells previously electroporated with infectious CVB3-CKO T7 transcribed RNA or from hearts and spleens of mice after transfection with CVB3-CKO RNA. Viral RNA was isolated in order to amplify the CRE(2C) coding region and the genomic 5′ terminal sequences. Sequence analysis revealed reversion of the CVB3-CKO sequence to wildtype occurs by 8 days post-electroporation of HeLa cells and by 20days post-transfection in mice. However, 5′ terminal deletions evolve prior to these times. Reversion of the CRE(2C) mutations to wildtype despite loss of the genomic 5′ termini is consistent with the hypothesis that an intact CRE(2C) is inherently vital to EV replication even when it is not enabling efficient positive strand initiation.