Recombination in enteroviruses: the genetics, cell biology and biochemistry of a biphasic replicative mechanism of virus evolution

Many of the most important human and animal virus pathogens have positive strand RNA genomes. These include, for example, poliovirus, foot and mouth disease virus and deformed wing virus of honeybees. The majority of these RNA viruses evolve very rapidly, generating large populations of highly divergent progeny. This variation helps the virus evade the host immune system - including escaping immunity induced by vaccines - and may enable the virus to spread more efficiently to new hosts, including cross-species transfer.

The variation in the virus population is due to two things; the virus has an error-prone polymerase enzyme which, through mis-incorporation, results in imprecise copying during virus replication. This mis-incorporation can result in up to 0.1% divergence in newly synthesized virus genomes. Secondly, if two related viruses co-infect the same cell, the virus can recombine. Recombination facilitates very much more extensive changes of the virus genome - up to 60% in many cases. Recombinant viruses are essentially hybrids, with part of the virus genome derived from one parent, and part from the other parent. Clearly, by combining such extensive regions of two different viruses there are opportunities for very significant changes in the phenotype i.e. the host range, tissues tropism or pathogenic potential, of the resulting virus.

Our laboratory has demonstrated the evolution of a virulent recombinant form of deformed wing virus of honeybees which appears to be associated with global disease. Other studies have shown the evolution of neurovirulent poliovirus in a poorly vaccinated population following recombination with a related co-circulating virus.

The error-prone polymerases of positive strand RNA viruses are well characterised. In contrast, the process of recombination is only poorly understood. We have developed an assay that, for the first time, allows the recombination process to be divided into two parts - an initial strand transfer event and a secondary resolution event that markedly increases the fitness of the recombinant virus. We propose to use this assay to provide the first detailed analysis of the mechanism of recombination.

Our assay unequivocally demonstrates that the generation of a recombinant virus goes via an intermediate in which there are duplications of parts of the virus genome. "Evolution by duplication" is one of the fundamental processes in virus evolution, in which individual genes are duplicated and then can subsequently independently evolve through acquisition of point mutations. The assay we have developed allows this process to be studied experimentally.

We will use enteroviruses as a model system; these viruses are generally well-characterised, there are good laboratory systems for their analysis and they are representative of the types of viruses of humans and animals in which recombination is both observed and problematic. We have over 25 years experience studying this group of viruses, in humans and animals.

We will study how the sequence of the virus contributes to the frequency and site of recombination. We will analyse where this process occurs in cells and the contribution made to this event by defined cellular proteins. Finally we will investigate the biochemical mechanism involved in the primary recombination (strand transfer) and resolution events.

These studies will contribute to an understanding of the fundamental mechanism of the recombination process which is a major driving force in virus evolution. Additionally, by defining the viral and cellular proteins that are involved it will enable the future control, exploitation or inhibition, of recombination - for example, in vaccines of the future.

Our preliminary studies have demonstrated that recombination in positive strand RNA viruses involves a replicative biphasic process with a strand-transfer reaction and subsequent resolution event. The strand-transfer reaction appears to be promiscuous and results in genome duplications. These duplications are resolved during subsequent replication, generating viruses with full fitness. We have developed an assay that allows the two recombination events to be dissected and quantified. This proposal seeks to better understand the molecular mechanism of recombination, the biochemistry of the process, the contribution of cellular and viral proteins and the constraints that limit recombination to related parental viruses. We will use poliovirus and related human enteroviruses as a model system. These viruses are extremely well characterised, with excellent reagents available.
Using a combination of reverse genetics and NGS we will investigate the sequence specificity of the strand-transfer and resolution events. Using FISH we will study the co-localisation of recombining virus genomes (and, conversely, determine whether viruses that do not recombine never get the opportunity because they are located in different replication complexes). We will investigate the role of viral and cellular proteins in the two processes, using reverse genetics or shRNA suppression of expression. In each case, the assay we have developed will be used to characterise and quantify the influence on recombination. Finally, we will use in vitro biochemical assays to study the kinetics and specificity of the strand transfer reaction and to analyse why nucleotide analog mutagens increase the recombinant yield.
The copy-choice model for recombination in positive sense RNA viruses was established in 1986. The assay we have developed provides the first opportunity in almost three decades to dissect the process into constituent components to better understand this fundamental evolutionary mechanism.

Immediate beneficiaries of this research include academic and industrial researchers involved or interested in the mechanisms, modulation or consequences of virus recombination. This is an extensive and globally-distributed group. Obviously the research will be of benefit to those studying human enteroviruses. Recombination is well-characterised in these viruses but remains only poorly understood mechanistically, despite being first described nearly three decades ago. However, the impact will be much broader than just the groups studying recombination of these viruses. Our study involves aspects of cell biology and enzymology, so groups involved in analysis of the replication complex or fundamental studies of polymerase activity will also benefit. There are additional translational aspects to the likely impact from this research - specifically applied to enteroviruses (also see below). Non-recombinogenic live attenuated poliovaccines are currently being developed in Gates Foundation-funded studies and our results will be of interest to groups involved in these studies.

Since enteroviruses are acknowledged as an excellent paradigm for understanding positive strand RNA viruses the impact of our research will extend to those involved in fundamental or applied aspects of research on this widespread group of economically and evolutionarily important viruses. These include viruses of humans (e.g. flaviviruses), non-human animals (pestiviruses of cattle and iflaviruses of insects) and plants (tombusviruses) where the importance of recombination is well-established, but the experimental systems are either not available or their general application may not have been demonstrated. We already work on recombinant viruses of honeybees and have excellent links with groups globally involved and interested in this research.

The recombination assay we have developed has, for the first time, demonstrated that a generic antiviral (ribavirin) enhances recombination at sub-inhibitory levels. This has profound implications for the use and future development of polymerase-targeting antivirals (or, for similar reasons, for inclusion of function-modifying polymerase mutations which we also demonstrate influence recombination rates). These results will have impacts in academic and industrial research environments. Our assay may be developed for high-throughput quantification of recombination and resolution events, allowing the screening of potential antiviral compounds for recombination-enhancing (or inhibiting) activity.

Viruses generally evolve rapidly as a consequence of their short life cycles, high progeny yield and error-prone replication. Recombination plays a critical part in this rapid evolution, enabling very extensive change within the virus genome (far in excess of that achieved by the error-prone polymerase). We show in preliminary studies that this involves partial genome duplication. Aside from the impact this observation has on our understanding of protein function, this has important consequences for our understanding the "evolution by duplication" process of the virus genome. Our studies on the influences of this promiscuous strand-exchange event will therefore impact on evolutionary biologists studying how and why positive-strand viruses have historically evolve ... and how they might evolve in the future.

Immediate beneficiaries include researchers working on enterovirus recombination. Longer term - perhaps 3-5 years - there will be much wider impact on disparate academic and industrial studies covering fundamental aspects of virus evolution to screening antivirals for recombination-enhancing activity. Finally, recipients of novel drug therapies or non-recombinogenic vaccines (human and animal) might expect to benefit from this fundamental research within a decade.