Lessons learned from watching viruses assemble

Inside a cell, a single protein emerges from from the cell’s protein-making machinery, then another, and another. They move about—bouncing into one another, wriggling around, and drifting away again. Soon, two proteins link up. Sticky patches on the proteins come together, protecting each other from the cell’s watery contents. Positively charged proteins attract negatively charged nucleic acids. Hydrogen bonds form and break. An assembly process has begun.

But these proteins aren’t building a piece of cellular machinery. This cell has been hijacked, and now it is making more hijackers. It has been infected by a virus, which has transformed the cell into a factory for making more of its kind.

Scientists have spent their careers trying to understand how viruses, the ultimate nanomachines, build themselves from smaller components. Combining biochemical data, microscopy, and complex calculations, they are modeling the conditions and chemical properties that allow so many individual pieces to form a complex shape. The details of that self-assembly process could help researchers defeat the virus with antivirals or build protective shells for drug delivery.

Viruses are “incredibly simple yet incredibly complicated,” says Helena Maier, an expert in coronavirus replication at the Pirbright Institute. Their genomes might code for only a tiny fraction of the proteins made by plants or animals, but “they can take over entire biological systems to do what they need to do.”

Be they SARS-CoV-2, HIV, or phages that infect bacteria, all viruses comprise at least two components: their genetic material—DNA or RNA—and a protective shell called a capsid. Depending on the virus, that shell could be made from 100 to 10,000 individual protein subunits. The size, shape, and particular elements of the capsid are unique to each virus, but they are all built from smaller components, says Brandeis University’s Michael Hagan, who models how the protective protein cages of viruses form.

When it comes to the assembly of that cage, he says, “there are a lot of common threads.” As a capsid shuffles into its optimal shape, each interaction between subunits is relatively weak. Hagan says that might be the point: the process is self-correcting, and once the capsid is fully formed, it’s optimized for its job. The interactions that are crucial to viral replication would be a “perfect target for an antiviral,” Maier says.

A capsid can have one of two general shapes. Icosahedral capsids completely encase the viral genome in a round box, while helical capsids complex the genome creating a spring shape. According to Juan Perilla at the University of Delaware, whichever shape a virus’s capsid takes, the delicate balance of interactions that holds the capsid together can tell him how the virus enters the cell and what it does when it gets there. He describes a virus as a little machine with intimate knowledge of its host cell and the expertise to negotiate that cell’s defenses.

Take, for example, the Ebola virus. Like coronaviruses, filoviruses like those that cause Ebola have a genome made of RNA and feature a helical nucleocapsid. Ebola uses a complex entry mechanism to get into the cell, Perilla explains, but once inside, its nucleocapsid complex quickly falls apart to release the viral RNA. Individual capsid proteins interact with the infected cell’s RNA to disrupt healthy function.

By contrast, HIV has a different capsid structure—one that plays a different role, Perilla says. Inside the membrane of an HIV particle, an icosahedral shell of protein subunits arranged in hexagons and pentagons forms a conical box that protects a genome made of RNA. Once the virus infects the cell, this boxlike arrangement of proteins regulates many cellular processes while remaining undetected. It interacts with cellular proteins that help import the viral RNA into the nucleus, the very heart of the cell. Only when the virus is in the right place will the capsid crack open to deliver its cargo.

Understanding the minute details of these infection processes—the physical and mechanical properties that enable a capsid to disassemble and form again—could reveal ways to take down a virus. Researchers can design vaccines to stop viruses from getting into cells or develop antivirals to prevent different steps in a virus’s replication. For example, researchers can interfere with the production of the virus in cells or with the viral capsid’s properties.

Perilla runs vast calculations to understand capsid stability at an atomic level and explains that even a small disruption in the delicate balance of interactions can have catastrophic effects for a virus. He and his team use their calculations to find regions in the capsid that are important for structural stability. Perturbing those areas with a small molecule might mean the capsid opens at the wrong time or place, allowing the cell to digest the viral components. Conversely, those perturbations might instead lock the capsid shut so it never releases its nucleic acid cargo.

Using a drug to disrupt capsid assembly or stability isn’t just theoretical. In 2006, Gilead Sciences began a project to look for small molecules that would disrupt the formation of the HIV capsid. The team mixed small molecules with purified capsid proteins in biochemical screens to find compounds that either sped up or slowed down the spontaneous assembly of protein complexes. Then in 2009 and 2011, academic researchers published the crystal structures of the heptamers and pentamers that stitch together to form the capsid (Cell 2009, DOI: 10.1016/j.cell.2009.04.063; Nature 2011, DOI: 10.1038/nature09640). That information, and another 6 years of structure-guided design, resulted in a drug candidate, lenacapavir, that is still undergoing clinical trials.

In email correspondence with C&EN, Stephen Yant, director of HIV discovery virology at Gilead, explains that lenacapavir uses electrostatic interactions and hydrogen bonding to change how the HIV capsid assembles and interacts with the cell. For example, the drug promotes binding between the individual protein components during assembly and then binds to and stabilizes the multimers that join to form the capsid. Collectively, Yant says, these interactions accelerate capsid formation, “resulting in malformed capsids” that cannot be replicated.

During the development of lenacapavir, researchers also learned more about the role of the capsid protein, Yant says. They found that the drug not only disrupts capsid assembly but also interferes with other capsid-protein-dependent functions throughout the viral replication cycle. One of those disrupted functions, he says, is the process by which the HIV capsid helps import the virus's genome into an infected cell’s nucleus and then integrate it into the cell’s DNA.

And if HIV succeeds in infecting a cell, lenacapavir can slow disassembly and interrupt other capsid-protein-dependent functions to stop new copies of the virus from being made. Studies on HIV capsid biology “continue to uncover fascinating new roles for the capsid at nearly every stage of the viral replication cycle,” Yant says.

Capsid inhibitors could also address other viruses. For example, firms such as Arbutus Biopharma and Assembly Biosciences are betting that by inhibiting assembly of the hepatitis B virus (HBV) capsid, they can block replication and treat the disease more effectively. They aim to improve on the current HBV treatment, which requires a regimen of drugs that suppresses but does not eliminate the virus. Both Arbutus and Assembly have HBV capsid inhibitors in clinical trials.

The modeling and calculations that scientists like Perilla and Hagan perform can guide the design of capsid inhibitors by suggesting areas of the shell that are important for its stability or assembly. Perilla recently identified the key interactions for the Ebola nucleocapsid, and he and his team started investigating the SARS-CoV-2 nucleocapsid early in 2020. Those SARS-CoV-2 models, he says, are still evolving as more experimental and biochemical data become available.

While some researchers try to disrupt the intricate processes of infection, replication, and assembly, others rely on viruses to inspire self-assembling systems that can package drugs and even deliver them to where they need to go.

To design a self-assembling system that mimics a virus, researchers need to remember that viruses consist of molecules with certain affinities and specific interactions, says Daniela Wilson, head of systems chemistry at Radboud University. While descriptions of viruses can often ascribe agency to the little machines, she says, they are just a collection of interacting compounds. Their self-assembly relies on noncovalent interactions that are familiar to chemists, including hydrogen bonding, the hydrophobic effect, and electrostatic attraction.

Although viral assembly can involve host-cell proteins and other active processes, it is mostly random, Hagan says. Proteins just bump up against one another and then stick together as the capsid travels down a free-energy path. And it might seem counterintuitive, but entropy, the measure of disorder in a system, can also help the shell form. Instead of being able to freely rotate and move through space, each protein subunit becomes more fixed in space as it joins the assembling structure. But at the same time, counterions and water molecules are released as each new subunit joins the growing capsid, increasing the entropy overall. “It’s actually entropy driven at the end in some cases,” Hagan says.

Those same noncovalent interactions and entropic factors can be used to build artificial assemblies. For example, systems chemists like Wilson are trying to build small, bottom-up self-assembling structures that can deliver drugs. These structures form spontaneously, package up cargo, and then release the drug where it needs to go, functioning like a virus, but with a helpful rather than harmful payload.

But designing these systems isn’t trivial. Viruses have had millions of years to work out how to do it, Wilson says, and they are experts at building their protective structures from simple starting materials. She says chemists do have one design advantage: they can use materials and polymers that are different from the proteins, RNA, and DNA that viruses rely on.

University of Minnesota Twin Cities chemist Theresa Reineke agrees. “As a synthetic chemist, we have essentially infinite chemical space and infinite architectural space to work from,” she says. For example, Reineke’s lab develops delivery systems made from synthetic polymers; these systems are designed to deliver their therapeutic contents into cells similarly to the the way viruses deliver their genetic cargo.

Reineke’s team has built and characterized two architectures made of glycopolymers that, like a viral capsid, assemble using noncovalent interactions and entropy. In one architecture, a shell of polymers noncovalently associates around the payload. In the second, self-assembled polymer micelles attach to or wrap around a protein or nucleic acid payload but don’t fully contain it in a shell.

The second approach, Reineke says, might be more promising. The polymer seems to bind the payload tightly enough to get it into the cells but loosely enough that it can then be released (J. Am. Chem. Soc. 2019, DOI: 10.1021/jacs.9b06218). Just as with viral capsids, the binding needs to be balanced to find that sweet spot for release.

Many researchers, like Reineke, are using synthetic polymers to develop self-assembling containers. Others have built synthetic nucleocapsids out of nonviral, laboratory-designed proteins (Proc. Natl. Acad. Sci. U.S.A. 2018, DOI: 10.1073/pnas.1800527115) or created DNA-origami nanostructures that can deliver drugs and then release them when inside the right cells.

Hagan says that as the field of capsid biology and smart materials evolves, there are still “fundamental things to be learned about self-assembly from viruses because they do it so well.” Researchers expect that the growing understanding of how viruses assemble will inform the design and understanding of other self-assembling systems—and help researchers put together lots of small interactions to create something more significant than the constituent parts.

Viruses can cause all sorts of chaos, Hagan says. “We study them, and they’re fascinating. They make these beautiful structures, and one can learn so much from them,” he says. “But you can also see the havoc that they wreak.”