A world of virus structures: understanding how non-icosahedral capsids are built

Viruses are probably the most successful pathogens on earth. They are everywhere, and they infect every other type of organism, including plants, animals (including humans), fungi, bacteria of all types, and even other viruses. Wherever we have looked for them, they have been found. As a result, they are of huge societal importance, impacting directly on our lives because of their effects on human and animal health, agriculture, and thus on food security on a truly global scale.

All viruses face a common challenge, in that they must package their genomic information (either DNA or RNA) within a protective container called a capsid, that shields the genome from the external environment, and delivers it intact to a new cell to start a new round of infection. One way in which these capsids are made is to build a highly symmetric container from a single type of protein. However, to do this the single type of protein has to adopt multiple shapes to build a container of the right size - much like the hexagons and pentagons that are needed to make a football. Building the right capsid, and building it perfectly, is a fundamental part of the viral replication cycle, but our understanding of how this 'conformational switching' happens is very poor. One way that it could occur is for the protein to bind to defined sequences within the genomic DNA or RNA; this binding would drive the conformational change. However, this process is poorly understood. In part, this is because for the vast majority of viruses the capsids have the very high symmetry described above, which means that when we solve their structures, symmetry averaging washes out details of any specific interactions between the protein (which is the same in each position) and the DNA or RNA (which is not, because it has to have a unique sequence that encodes the virus' genes). The high symmetry many viruses rely on is therefore tremendously unhelpful when we try to study the molecular mechanisms involved in assembly.

In this proposal we want to exploit two hugely exciting recent discoveries in our laboratories, that will allow us to overcome this barrier and discover, for the first time, the cryptic rules that allow these viruses to efficiently self-assemble. We have been working on two different families of virus that are each important pathogens of food and textile crops globally, and thus are major threats to food security and agricultural economies across the developed and developing world; Geminiviruses and Umbraviruses. In each, the virus has evolved a (different) novel innovation that means the capsid has a "non-standard" structure which is no longer quite as symmetric as is normally the case. Remarkably, in the preliminary structure of each which we have solved with 5-fold symmetry (rather than the 60-fold symmetry for an icosahedral virus), we can now see details of DNA (for Geminiviruses) and RNA (for Umbraviruses) bound to the viral coat proteins. This grant application will allow us to solve high resolution structures of these non-standard virus capsids without any symmetry averaging at all. Together with biochemical and bioinformatics experiments, we will uncover the details of genome binding, how this changes protein conformation, and where these features lie within the viral genome. This will (a) provide fascinating new fundamental biological insights that are important in understanding how viruses work, (b) provide a mechanistic understanding that could lead to new ways to prevent them working, and (c) make clear the rules for virus assembly that could allow us to change the way viruses assemble, to make capsids of, for example, different sizes for biotechnology applications.

Many isometric viruses have capsids where a single protein adopts multiple conformations, but how these conformations are specified remains poorly understood. Increasing evidence suggests that for many viruses, degenerate sequences distributed through the genome act as packaging signals that both direct capsid assembly and specific encapsidation of the genome. However, when the structures of such viruses are solved, typically very little RNA or DNA is observed. In large part this is because the high symmetry averaging implicit in icosahedral structures obscures information for bound nucleic acid. We wish to exploit our two recent discoveries that will help overcome this problem. Firstly, we recently solved the first high-resolution structure of the unique, twinned capsid of a geminivirus. This particle has ~28% of its ssDNA genome ordered at high resolution, and has unique sequence-specific protein:DNA contacts at the interface between the two halves of the virus. Secondly, we have determined preliminary structures of the particles generated by the pea enation disease complex (a Luteovirus and an Umbravirus), which include a canonical T=3 capsid, and a novel oblate particle in which there are ordered segments of ssRNA. Both the geminivirus and oblate particles are 5-fold, rather than 60-fold symmetric, and both have ordered nucleic acid density exactly where the capsids differ from icosahedral capsids. This strongly suggests that DNA/RNA recognition is occurring with sequences/structures that direct particle assembly. This grant is intended to (a) solve asymmetric structures that can resolve details of DNA/RNA recognition without the confusing effects of symmetry, and (b) use molecular virology, biochemistry and bioinformatics to complement structural biology and understand the sequence/distribution of sequences that dictate assembly, and (c) having defined the rules for assembly, design synthetic genomes that assemble novel particles with bespoke geometries/size

This project will address an outstanding problem in virology - namely, how are multiple different conformations of a single capsid protein controlled in order to build an exquisitely intricate protein assembly? Specifically, how does a virus control which conformation occurs at each position? This is absolutely critical if sealed, stable capsids are to be built. We will exploit recent BBSRC-funded progress in two systems: geminiviruses and Luteovirus/Umbravirus disease complexes, that will give us genuinely novel insights into this fundamental process. Furthermore, we will do this using viral systems that are major threats to food security and agricultural economies across the developed and developing world.

We therefore expect this research to have impact in several areas. Firstly, the assembly of virus particles is vital for the replication cycles of all viruses, as it enables the labile genetic material to be spread intact through the harsh external environment. Understanding the structure and assembly of virus capsids will provide insights into the diversity of viral architectures and the mechanism of encapsidation of genetic material. This is especially true in the case of the oblate particle we tentatively ascribe to PEMV-2, which is entirely novel in virology. Additionally, for the systems studied here, the capsids built are not standard icosahedra, but have specific shapes that require additional conformational switching, and/or precise spatial control of that conformational switching. This research will therefore give new information on the RNA and DNA binding that drives such fine-tuned conformational control. Thus, the proposed research will have substantial impact in virology in general. Understanding the mechanism of virus assembly will greatly assist those wishing to develop anti-viral strategies by interfering with the assembly process. Since viral diseases limit both animal and plant productivity, this project is directly relevant to the BBSRC Research priority of Food Security.

In addition to being highly relevant to those working on virus structure and function, the research proposed will have an impact on those wishing to develop bespoke virus-like particles as delivery vehicles, a rapidly expanding field. VLPs offer a potential means of delivering therapeutic molecules (e.g. siRNAs or genes for heterologous expression) into target cells. Furthermore, the results will elucidate the basic principles underlying assembly and structure which will relevant to many viruses of both plants and animals, especially those with ssDNA and ssRNA genomes.

As well as the scientific outputs of the project, a significant impact will be the training of staff in state-of-the-art techniques in plant-based expression systems (expertise currently centred at JIC) and advanced training in cryoEM (expertise currently centred at UoL). Critically, the general increase in the skill-base of the other institution (i.e. plant expression in the Ranson lab at UoL and CryoEM for staff at JIC), will have major, long-term impact on the science carried out at both institutions, and increasing the scientific capability of the UK. We request funding in the justification of resources to enable reciprocal workshops to facilitate the delivery of this impact.

Furthermore, although this is essentially an academic research project, both UoL and JIC will consider intellectual property (IP) issues at an early stage in any discoveries. In particular, we will consider commercial opportunities that may arise from the creation of bespoke capsids, and the potential use of protein binding reagents for VLPs in both diagnostics and biotechnology.