RNA Helicase A/DHX9 Forms Unique Cytoplasmic Antiviral Granules That Restrict Oncolytic Myxoma Virus Replication in Human Cancer Cells.

Abstract

ABSTRACTRNA helicase A/DHX9 is required for diverse RNA-related essential cellular functions and antiviral responses and is hijacked by RNA viruses to support their replication. Here, we show that during the late replication stage in human cancer cells of myxoma virus (MYXV), a member of the double-stranded DNA (dsDNA) poxvirus family that is being developed as an oncolytic virus, DHX9, forms unique granular cytoplasmic structures, which we named “DHX9 antiviral granules.” These DHX9 antiviral granules are not formed if MYXV DNA replication and/or late protein synthesis is blocked. When formed, DHX9 antiviral granules significantly reduced nascent protein synthesis in the MYXV-infected cancer cells. MYXV late gene transcription and translation were also significantly compromised, particularly in nonpermissive or semipermissive human cancer cells where MYXV replication is partly or completely restricted. Directed knockdown of DHX9 significantly enhanced viral late protein synthesis and progeny virus formation in normally restrictive cancer cells. We further demonstrate that DHX9 is not a component of the canonical cellular stress granules. DHX9 antiviral granules are induced by MYXV, and other poxviruses, in human cells and are associated with other known cellular components of stress granules, dsRNA and virus encoded dsRNA-binding protein M029, a known interactor with DHX9. Thus, DHX9 antiviral granules function by hijacking poxviral elements needed for the cytoplasmic viral replication factories. These results demonstrate a novel antiviral function for DHX9 that is recruited from the nucleus into the cytoplasm, and this step can be exploited to enhance oncolytic virotherapy against the subset of human cancer cells that normally restrict MYXV.IMPORTANCE The cellular DHX9 has both proviral and antiviral roles against diverse RNA and DNA viruses. In this article, we demonstrate that DHX9 can form unique antiviral granules in the cytoplasm during myxoma virus (MYXV) replication in human cancer cells. These antiviral granules sequester viral proteins and reduce viral late protein synthesis and thus regulate MYXV, and other poxviruses, that replicate in the cytoplasm. In addition, we show that in the absence of DHX9, the formation of DHX9 antiviral granules can be inhibited, which significantly enhanced oncolytic MYXV replication in human cancer cell lines where the virus is normally restricted. Our results also show that DHX9 antiviral granules are formed after viral infection but not by common nonviral cellular stress inducers. Thus, our study suggests that DHX9 has antiviral activity in human cancer cells, and this pathway can be targeted for enhanced activity of oncolytic poxviruses against even restrictive cancer cells.