Computationally-designed synthetic protein assemblies that evolve like viruses

Like Comment

By Marc Lajoie and Gabriel Butterfield

Baker lab, University of Washington 

Behind the paper: Butterfield GL*, Lajoie MJ*, Gustafson HH, Sellers DL, Nattermann U, Ellis D, Bale JB, Ke S, Lenz GH, Yehdego A, Ravichandran R, Pun SH, King NP, and Baker D (2017) Evolution of a Designed Protein Assembly Encapsulating its Own RNA Genome. Nature. doi:10.1038/nature25157.

The paper in Nature is here:


In the Baker and King labs at the University of Washington, we have been computationally designing protein materials that self-assemble in various pre-defined symmetric geometries1–4. Some of these materials resemble icosahedral viral capsids even though their building blocks are derived from unrelated natural enzymes and have no relation to proteins of viral origin. This begged the question: could we functionally recapitulate a viral nucleocapsid from first principles without using any viral components? By installing positively charged residues on the interior surface of the icosahedral assemblies, we found that they could encapsulate their own full-length genomes to form a nucleocapsid just like viruses do. This was really exciting: it was the first time that a capsid of non-viral origin had packaged its own genome. We had taken two otherwise unrelated proteins, and created a synthetic system that could undergo evolution.

So if nucleocapsid genesis just requires an icosahedral capsid and electrostatic RNA packaging, why doesn’t this happen more often? Viruses have complex lifecycles: they must replicate in their host cell, assemble infectious particles that encapsulate their own genomes, escape from their host cell, travel through harsh environments to reach a new host cell, enter the new cell, and finally uncoat to release their genome for the next round of replication. This is where we could help our synthetic nucleocapsids cheat: as long as they could assemble and encapsulate their own full-length genome inside of the producer Escherichia coli cells, we could help them through the rest of their complicated lifecycle by hand. We cracked open the producer cells, purified intact nucleocapsids, selected functional variants against various biochemical challenges (e.g., RNase A, mouse blood, circulation in living mice), harvested the intact nucleocapsid RNA by phenol-chloroform extraction, replicated full-length genomes in vitro by reverse transcription PCR, and delivered their genetic information to new host cells by cloning into expression plasmids and electrotransforming into fresh E. coli cells. Through this process, one step of design and three steps of evolution created nucleocapsids that can package their own genomes, resist degradation in blood, and circulate in living mice as well as the best recombinant adeno-associated viruses5,6

Genome encapsulation is a fundamental feature of life; evolution of complex behaviors is made possible by linking genotype to phenotype. Surprisingly, starting from a blank slate (a synthetic nucleocapsid that had never been exposed to RNases, blood, or mouse circulation), we found that there were many distinct solutions that could increase fitness. This apparent ease of evolving in complex environments suggests that several of the early steps in viral evolution could be accessible to a surprisingly broad sequence space. Additional viral mechanisms such as cell entry and cargo release may be attainable by combining rational design with evolution, and these properties could prove useful for targeted non-viral drug delivery.


1.        King, N. P. et al. Computational design of self-assembling protein nanomaterials with atomic level accuracy. Science 336, 1171–4 (2012).

2.        King, N. P. et al. Accurate design of co-assembling multi-component protein nanomaterials. Nature 510, 103–8 (2014).

3.        Hsia, Y. et al. Design of a hyperstable 60-subunit protein icosahedron. Nature 535, 136–139 (2016).

4.        Bale, J. B. et al. Accurate design of megadalton-scale two-component icosahedral protein complexes. Science 353, (389-394).  (2016).

5.        Sommer, J. M. et al. Quantification of adeno-associated virus particles and empty capsids by optical density measurement. Mol. Ther. 7, 122–8 (2003).

6.        Drouin, L. M. et al. Cryo-electron Microscopy Reconstruction and Stability Studies of the Wild Type and the R432A Variant of Adeno-associated Virus Type 2 Reveal that Capsid Structural Stability Is a Major Factor in Genome Packaging. J. Virol. 90, 8542–51 (2016).

Gabriel Butterfield

PhD Student, University of Washington