The oncogenic Epstein-Barr virus (EBV) evades the immune system but has an Achilles heel: its genome maintenance protein EBNA1, which is essential for viral genome maintenance but highly antigenic. EBV has seemingly evolved a system in which the mRNA sequence encoding the glycine-alanine repeats (GAr) of the EBNA1 protein limits its expression to the minimal level necessary for function while minimizing immune recognition. Here, we identify nucleolin (NCL) as a host factor required for this process via a direct interaction with G-quadruplexes formed in GAr-encoding mRNA sequence. Overexpression of NCL enhances GAr-based inhibition of EBNA1 protein expression, whereas its downregulation relieves the suppression of both expression and antigen presentation. Moreover, the G-quadruplex ligand PhenDC3 prevents NCL binding to EBNA1 mRNA and reverses GAr-mediated repression of EBNA1 expression and antigen presentation. Hence the NCL-EBNA1 mRNA interaction is a relevant therapeutic target to trigger an immune response against EBV-carrying cancers.

Epstein-Barr virus (EBV) is tightly associated with certain human cancers, but there is as yet no specific treatment against EBV-related diseases. The EBV-encoded EBNA1 protein is essential to maintain viral episomes and for viral persistence. As such, EBNA1 is expressed in all EBV-infected cells, and is highly antigenic. All infected individuals, including individuals with cancer, have CD8+ T cells directed towards EBNA1 epitopes, yet the immune system fails to detect and destroy cells harboring the virus. EBV immune evasion depends on the capacity of the Gly-Ala repeat (GAr) domain of EBNA1 to inhibit the translation of its own mRNA in cis, thereby limiting the production of EBNA1-derived antigenic peptides presented by the major histocompatibility complex (MHC) class I pathway. Here we establish a yeast-based assay for monitoring GAr-dependent inhibition of translation. Using this assay we identify doxorubicin (DXR) as a compound that specifically interferes with the GAr effect on translation in yeast. DXR targets the topoisomerase-II–DNA complexes and thereby causes genomic damage. We show, however, that the genotoxic effect of DXR and various analogs thereof is uncoupled from the effect on GAr-mediated translation control. This is further supported by the observation that etoposide and teniposide, representing another class of topoisomerase-II–DNA targeting drugs, have no effect on GAr-mediated translation control. DXR and active analogs stimulate, in a GAr-dependent manner, EBNA1 expression in mammalian cells and overcome GAr-dependent restriction of MHC class I antigen presentation. These results validate our approach as an effective high-throughput screening assay to identify drugs that interfere with EBV immune evasion and, thus, constitute candidates for treating EBV-related diseases, in particular EBV-associated cancers.