Stress granules and RNA-binding proteins in Epstein-Barr virus infection and cancer

Biomolecular condensates are membrane-less compartments formed within a cell, that have a broad range of functions, depending on the type and localisation. One example of a biomolecular condensate is the stress granule (SG), which forms within the cytoplasm and contains translational machinery, mRNA and RNA binding proteins sequestered to foci following stress-induced translational stalling. Several stresses are known to induce SG formation, commonly via the phosphorylation of eukaryotic initiation factor 2 a (eIF2a) by four main kinases, haem-regulated inhibitor (HRI), double-stranded RNA activated protein kinase R (PKR), PKR-like endoplasmic reticulum kinase (PERK), and general control non-depressible 2 (GCN2). These kinases are commonly activated following viral infection, inhibiting translation, and preventing viral replication; therefore, many viruses have developed mechanisms to evade this process. Epstein-Barr virus (EBV) is a gamma herpesvirus that infects humans, developing a persistent latent infection within B cells. EBV displays a biphasic growth cycle, exhibiting both latent and lytic cycles and it is well understood that latent EBV infections contribute to the development of several cancers. Whilst many viral processes have been attributed to the formation of these cancers, many others remain to be determined. While several studies have investigated herpesviruses, eIF2a and SG formation processes, little is known regarding EBV and these mechanisms. We aimed to investigate whether lytic EBV altered SG formation similarly to other lytic viruses. Whilst also focusing on latent EBV, eIF2a and SG formation. Misregulation of SG formation has been associated with several human diseases, including cancer, therefore we aimed to determine whether any link existed between latency and altered SG assembly. We hypothesised that while latent EBV viral product expression is restricted, and may not activate the eIF2a pathway, it may be capable of altering the expression of several key SG-associated components. Lytic EBV was shown to inhibit SG formation following induction via eIF2a-dependent and independent pathways, providing the first evidence that lytic EBV can circumvent the SG response, albeit through currently unknown mechanisms. We found that latent EBV infection did induce activation of PKR, PERK or phosphorylation of eIF2a. This was reinforced by our finding that latent EBV infection did not promote SG formation, nor alter chemical-induced SG assembly. However, we provide evidence that latent EBV infection alters the transcription and splicing of a key SG protein, TIA-1. T-cell intracellular antigen-1 (TIA-1), an RNA-binding protein, is central to SG formation and displays both oncogenic and tumour suppressor roles achieved via alternative isoforms. We found that latent EBV infection reduced mRNA levels of TIA-1b, known to act as a tumour suppressor, in B cells, whilst maintaining the levels of TIA-1a, the oncogenic isoform. We suggest that this manipulation of TIA-1 isoform expression may contribute to the development of cancer during persistent latent EBV infection, and may provide a novel target for preventative measures. This study provides the first basis for the understanding of how the biphasic life cycle of EBV may affect eIF2a, SG formation and associated RNA-binding proteins, to benefit its growth and survival.