The natural hosts for the Ebola virus are thought to be bats. However, this RNA virus can also infect huma ns, and there have been numerous reported outbreaks of the viral infection originating in African countries over the past 40 years1. The largest such outbreak was between 2013 and 2016, and resulted in 28,616 suspected cases and 11,310 deaths, mainly in Guinea, Liberia and Sierra Leone (go.nature.com/2qtbj6i). The fatality rate can be high: for example, an outbreak that began in 2018 in the Democratic Republic of the Congo has so far resulted in 685 cases of infection and 419 deaths, a fatality rate of approximately 60% (go.nature.com/2qtdirv).
Ebola infection begins with fever, muscle pain and headache, followed by vomiting, diarrhoea, rash and symptoms of impaired kidney and liver function. Basic supportive care for those infected, such as treatment to combat dehydration, can help to prevent it being fatal2. However, in addition to managing symptoms, there is a need to develop other approaches that prevent or treat the disease, such as vaccines, antiviral therapies or antibody treatments3–6. Although there have been some clinical trials, no drugs or vaccines have yet been approved for clinical use. And because it can’t be predicted where the next Ebola outbreak will occur, it is difficult to identify those most at risk of infection, and so plan a vaccination strategy. Writing in Cell, Batra et al.7 report their investigation of natural host defences against the Ebola virus. Their identification of a human protein that can affect the success of viral replication might open new avenues of research into antiviral treatments.
Batra and colleagues expressed tagged versions of Ebola proteins individually in human cells grown in vitro, and used co-immunoprecipitation and mass spectrometry techniques to identify human proteins that interacted with viral proteins. They used this information to generate a map of the network of such interactions — termed an interactome map. The authors found 194 interactions between host and viral proteins, one of which was between the human protein RBBP6 — a type of enzyme called a ubiquitin ligase — and an Ebola protein called VP30. Various Ebola proteins, including VP30, function in the viral polymerase protein complex, which makes viral RNA, and Batra and colleagues transferred DNA sequences encoding these proteins into human cells grown in vitro. They found that the expression of VP30, at both the RNA and protein level, depended on the level of RBBP6, with high levels of RBBP6 being associated with low levels of VP30 expression and with low activity of the viral polymerase complex. When the authors used a technique called gene silencing to reduce the expression of RBBP6 in the human cells expressing Ebola proteins, the activity of the Ebola viral polymerase complex and viral replication was increased.
The authors carried out an X-ray structural analysis to model the interaction between VP30 and RBBP6. These structural results, and those from their other experiments investigating protein binding to VP30, revealed that a key feature of interest in RBBP6 is a sequence of 23 amino-acid residues (a peptide) that includes a motif called PPxPxY (in which P is the amino acid proline, Y is the amino acid tyrosine and x is any other amino-acid residue). This motif inserts into a cleft in VP30 that is known8 to bind to an Ebola protein called NP (Fig. 1). NP is part of the viral polymerase complex and serves as a scaffold that binds to viral RNA. VP30 binding to NP promotes viral RNA synthesis. The researchers noted that the PPxPxY motif is also present in NP in a region that binds to VP30. This suggests that NP and RBBP6 might compete for binding to VP30.
In in vitro experiments, Batra et al. showed that the binding between RBBP6 and VP30 was five times stronger than that between NP and VP30, indicating that RBBP6 should have the capacity to effectively sequester VP30 from NP. Their experiments suggest that this peptide motif alone has a key role in limiting the interaction between VP30 and NP. Furthermore, if, instead of RBBP6, an engineered chimaeric protein composed of a PPxPxY-containing peptide and a fluorescent marker called green fluorescent protein was expressed in human cells expressing Ebola proteins, the chimaeric protein decreased Ebola polymerase complex activity and viral propagation compared with the effect observed in cells that received only green fluorescent protein.
Because Batra and colleagues’ approach to identifying the interactions between viral and host proteins is based on the expression of single types of viral protein, it does not identify host-protein interactions that might occur only when viral proteins are part of multi-protein complexes. Further analyses will therefore be needed to identify any such interactions, and to verify that the interactome map is correct. Also, because these experiments were conducted using genetic-engineering techniques, rather than studying a natural process of viral infection of human cells, it is difficult to assess the extent to which these events ultimately affect the level of viral replication. It is possible that, like other host antiviral proteins, RBBP6 levels vary and are subject to regulation by unknown mechanisms. Further study is needed to understand exactly how RBBP6 affects Ebola virus replication.
During the 2013–16 Ebola outbreak9,10, no mutations in VP30 were reported in the region of the protein that corresponds to the RBBP6 binding site. This suggests that evolutionary selective pressure to evade host targeting by RBBP6 is limited, or that a viral mutation that drives resistance to RBBP6 is not selected for because it has a detrimental effect on the virus — possibly because such a mutation might also affect an interaction between VP30 and NP that is essential for viral replication. It would be interesting to test whether bats inhibit Ebola virus replication using RBBP6.
A key discovery of this study is that a peptide that includes the PPxPxY motif, when fused to green fluorescent protein, is sufficient to inhibit virus replication in vitro (the effect in vivo was not evaluated). The cleft on VP30 to which this peptide binds could therefore be a promising target for efforts to develop antivirals against Ebola. Although it is not known whether the peptide inhibits viral replication when it is not fused to green fluorescent protein, this discovery could provide a starting point for developing and optimizing other small compounds that inhibit viral replication. That might lead to the development of a class of antiviral that has high specificity for the Ebola virus.
Nature 566, 190-191 (2019)