Ebola virus replication stands out

Abstract

A recent study (Deflubé et al. PNAS, 2019) reports that Ebola virus genomes have variable 3’ terminal nucleotides. This finding is a departure from other nonsegmented, negative strand RNA viruses (NNSVs) that have been studied to date and has broad implications for filoviral replication and innate immune evasion.

Rare filovirus disease (FVD) outbreaks are often associated with high case fatality rates, sometimes up to 90% [1]. Although factors determining patient survival are poorly understood, clinical observations suggest a high probability of death is associated with high viral load due to uncontrolled viral replication [2]. While progress has been made towards understanding filoviral innate immune evasion mechanisms, much work remains to be done to better define filoviral RNA synthesis [1]. These gaps limit the development of therapeutics, such as direct-acting antivirals (DAAs) targeting the viral RNA synthesis complex.

Much of what is assumed for Ebola virus, a member of the Filoviridae family, is largely borrowed from studies of other NNSVs, including respiratory syncytial virus (RSV) [3] and vesicular stomatitis virus (VSV) [4]. The Ebola virus RNA genome is encapsidated by nucleoprotein (NP), which controls access to the genomic RNA by the RNA synthesis complex. The viral genome is transcribed to generate subgenomic capped and polyadenylated mRNAs for viral protein translation. Subsequently, the viral genome is replicated to generate antigenomic and genomic RNAs, which are then bound by NP [1].

These two major processes, transcription and replication, are carried out by the RNA synthesis complex, which includes the viral RNA dependent RNA polymerase (RdRp or L protein) and viral protein 35 (VP35), an essential cofactor that bridges the interaction with NP [1]. However, the exact roles that cis-regulatory elements in the genome and viral proteins play in viral replication are not well understood, in part due to the complexity and multifunctional nature of these components.

A new study by Deflubé and colleagues [5] sheds light on a key part of the puzzle by discovering that the 3’ ends of genomic and antigenomic RNAs are variable. This is intriguing, as the polymerase needs to processively replicate the whole template, approximately 19 kb long, and recognize the 3’ ends of both genome and antigenome for initiation. Variable 3’ terminal sequences could possibly implicate different initiation mechanisms if all of these 3’ ends serve as initiation templates. To gain mechanistic insights of initiation, the authors used a modified minigenome assay that abolishes transcription and that is restricted to the antigenome synthesis step of replication. The results are striking. The polymerase always begins synthesis opposite the first C of the 3’ CCUGUG template, regardless of substitutions, deletions, or additional nucleotides added to the 3’ end. In addition, RNA synthesis is abolished if the first C is mutated, suggesting that the polymerase precisely locates its binding site. This observation is in contrast to what is currently known about RSV and VSV replication initiation, which involve de novo initiation from the exact 3’ end of templates [3, 4]. In the case of RSV, both replication and transcription begin at the leader (le) promoter region at the 3’ end of the viral genome. However, initiation occurs independently at different sites within the le (genome replication from position 1U and mRNA transcription from position 3C), with the same L protein capable of selecting both sites. Although the promoter sequences of RSV do not align with other viruses within the same family, their sequences all begin with 3’ UG, suggesting that the replication product is initiated 5’ AC. While the overall genomic structures and polymerase active sites in Ebola virus are similar to those of other NNSVs, the Ebola virus genomic ends are unconventional [6].

How do Ebolaviruses generate these variable 3’ ends with additional nucleotides? Two possible mechanisms are proposed by Deflubé, et al. First, sequencing results of some extended 3’ ends support a back-priming model where delayed encapsidation of RNA by NP at the termini results in formation of RNA hairpins that allow for templated addition of 3’ ends. The complex kinetics involved in RNA encapsidation, RNA secondary structure formation, and enzymatic reactions offer ample space to explain the variable 3’ ends generated. Second, terminal transferase activity has been demonstrated in some RdRps in positive sense RNA viruses [7]. Ebola virus may utilize similar mechanisms [8], but additional studies are needed.

Why do Ebola viruses have these variable 3’ ends with additional nucleotides? It appears that additional nucleotides may offer a replication advantage. Given that a major hurdle in replication is stabilization of the initiation complex, these additional nucleotides may stabilize incoming NTPs for the first few reactions to happen. In addition, a variable 3’ end may facilitate innate immune evasion by impairing recognition by cytosolic RIG-I like receptors (RLRs) that are sensitive to 3’ and 5’ overhangs [9]. Previous studies revealed that dsRNA binding by VP35 is important for Ebola virus IFN antagonism. Structural and functional analyses show that the VP35 C-terminal interferon inhibitory domain (IID) can recognize RNA, including those with overhangs. Mimicry of cellular RIG-I like receptors is part of the arsenal of Ebola virus immune evasion tactics [10].

Unique replication products defined by Deflubé, et al. are the latest findings that shed light on the unusual replication initiation mechanism utilized by Ebola viruses, which not too long ago was considered to be similar to other NNSVs.