Identification and characterisation of innate antiviral proteins with activity against Ebolaviruses

Student thesis: Doctoral ThesisDoctor of Philosophy


Ebola virus (EBOV), a negative sense RNA virus, encodes seven structural proteins of which two, VP35 and VP24 have developed mechanisms to evade and antagonise the interferon system. VP35 antagonises the RNA sensing pathway, whereas VP24 antagonises downstream interferon signalling by inhibition of STAT1 nuclear translocation. Previous work had shown inhibition of EBOV by interferon treatment and further IFN-β treatment had been utilised in a clinical trial during the West African outbreak. Several interferon stimulated genes (ISGs) have shown to be antiviral against EBOV including tetherin, IFITMs, ZAP and RBBP6. However, until recently a large scale ISG screen against EBOV had not been conducted.
In this study, we conducted an ISG overexpression screen containing 398 ISGs against EBOV utilising an EBOV trVLP (transcription and replication-competent virus-like particle) system. We then characterised one of the genes, TRIM25, an E3 ligase linked with the RNA sensing pathway. We found that the antiviral phenotype of TRIM25 overexpression was MAVS dependent although independent of RIG-I and TBK1 indicating a novel antiviral mechanism distinct from classical RNA sensing and downstream signalling. Importantly the antiviral activity of IFN-I treatment against EBOV was attenuated in both TRIM25 and MAVS knockout cells but not in RIG-I or TBK1 knockout cells. Furthermore, we showed that the antiviral phenotype was also dependent on ZAP, a TRIM25 associated factor. Overexpression of ZAP-L (the long isoform of ZAP) was also antiviral against EBOV and was TRIM25 dependent. Previously EBOV L polymerase has been shown to be the target of ZAP however we showed that this was not responsible for the observed antiviral activity but instead was linked to CG content of the minigenome, indicating a novel mechanism.
The antiviral activity of TRIM25 was further characterised by the association and ubiquitination with EBOV NP. Ubiquitinated NP did not appear to be targeted for proteasomal or lysosomal degradation and was linked to K27, K33 and K63 linked polyubiquitin chains. However, the antiviral activity of TRIM25, and ubiquitination of NP, was dependent on RING domain dimerisation as well as protein and RNA binding domains. TRIM25 or ZAP overexpression was also associated with the dissociation of incoming NP from viral genomic RNA, measured by RNA-IP. However, NP/ RNA dissociation of TRIM25 was independent of ZAP but effects of ZAP overexpression were TRIM25 dependent. Most importantly IFN-I treatment resulted in NP and RNA dissociation and this was in a TRIM25 and ZAP dependent manner. Further we showed that ZAP-L is capable of binding EBOV genomic RNA in a TRIM25 and CG content dependent manner.
We investigated the antiviral effects of interferon in TRIM25 and ZAP knockout cell lines utilising a full-length EBOV model however results were inconclusive. To investigate the apparent differences between full-length EBOV and the EBOV trVLP system we utilised an infection based trVLP model and showed that both TRIM25 overexpression was antiviral and the antiviral phenotype of IFN-I was attenuated in both TRIM25 and ZAP knockout cells. As there may be differences in VP24 expression between full length EBOV and EBOV trVLPs and as VP24 has been implicated in virion compaction we investigated the impact of VP24 on EBOV trVLPs. However, no differences in TRIM25 antiviral phenotype were observed between WT and ∆VP24 EBOV trVLPs.
Finally, we investigated the roles of two mutations that had arisen in EBOV NP and VP24 in the creation of a lethal murine model. The reason for the requirement of NP and VP24 mutations is unclear but as lethality models are also created in STAT1-/-, IFNAR-/- or MAVS-/- knockout mice previously it has been hypothesised that IFN plays a role. However similar to previous work our results indicated no differences between WT and mouse adapted (MA-EBOV) VP24 with regards to IFN antagonism in both human and murine cells. Furthermore, we investigated differences between propagation of WT and MA-EBOV as well as Makona strain EBOV trVLPs in human and murine cells and showed that attenuation was not alleviated in TRIM25 knockout cells.
Taken together this piece of work potentially uncovers a novel antiviral mechanism of TRIM25 against EBOV separate from the classical RLR signalling pathway, which appears to be critical to the antiviral activity of interferon. Furthermore, the hypothesised mechanism of action put forward by this study is the dissociation of EBOV NP from viral genome by TRIM25, potentially by the ubiquitination of NP, allowing ZAP to recognise CG rich regions and bind.
Date of Award1 Oct 2021
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorStuart Neil (Supervisor) & Katherine Doores (Supervisor)

Cite this