Bill Gelbart, UCLA, Los Angeles, California, USA

Making Viruses, in Vivo and in Vitro
Viral particles are unique among biological disease agents because of their exceptional variety, simplicity and beauty. Their variety is due to the fact that they have been evolving as long as there has been life itself, and evolving much faster than their hosts. Their simplicity and beauty, on the other hand, arise from the fact that they are not alive. Rather, they are necessarily dependent on their hosts for all the biochemical machinery required for their replication.

In the first lecture the case of double-stranded (ds) DNA viruses is discussed. These are represented predominantly by viruses that infect bacteria, and this is argued to be a consequence of the fact that most bacterial viruses – “phages” – do not enter their host cells; rather, only their genome does. The physical driving force for delivering the viral genome is due to the dsDNA being a stiff polymer that strongly repels itself. We discuss theoretical and experimental work that has recently demonstrated that bacterial viruses are characterized by large pressures and forces associated with the packaging of dsDNA in their capsids. Pressures as high as 50-100 atm. and forces as high as 50-100 pN have now been calculated and measured for a variety of phages and shown to do the job of ejecting these genomes into their host cells. It has also been demonstrated that an infectious virus of this kind can be synthesized in vitro from its purified components (and consumption of ATP). The large majority of plant and animal viruses, on the other hand, involve single-stranded (ss) RNA genomes.

In the second lecture this fact is related to the very different physical properties of ssRNA vs. dsDNA. Specifically, dsDNA is a relatively inflexible linear polyelectrolyte, whereas ssRNA is a flexible, branched, polyelectrolyte. As a consequence, dsDNA genomes can only be packaged by active processes that involve a great deal of work being performed in order to build up the stored energy density – pressure – needed for ejection and initiation of infection. Many ssRNA genomes, on the other hand, are able to spontaneously co-self-assemble with capsid proteins to form infectious virions. This remarkable fact has been demonstrated for several different plant viruses, both spherical (icosahedral) and cylindrical (helical). We describe this work, as well as recent, on going, attempts to reconstitute – from purified components – an infectious enveloped virus, i.e., one (like many mammalian viruses) whose capsid is further protected by a phospholipid membrane.

The third lecture discusses the idea of a “speed limit” for viral mutation, in the context of a variety of molecular evolution issues underlying the physical aspects of viral infectivity discussed in the first two lectures. For example, how do the overall lengths of the dsDNA genomes of phages need to evolve in order for their fully-packaged pressures to be high enough for their efficient ejection, but low enough for their mechanical stability? Do ssRNA genomes have to evolve – not “just” to encode for the right proteins, but also – to have many-gene sequences that give rise to the right secondary/tertiary structure, and hence the right size and shape, for spontaneous packaging by their capsid proteins? Do these viruses evolve to have different codon usage than their hosts? In general, what are the constraints on molecular evolution of viral genomes that follow from their functioning as simple physical “devices”?



Alex Evilevitch, Laurence Lavelle, Charles M. Knobler, Eric Raspaud, and William M. Gelbart (2003). Osmotic pressure inhibition of DNA ejection from phage. Proc. National Academy of Sciences, USA, 100:9292-9295.

Roya Zandi, David Reguera, Robijn F. Bruinsma, William M. Gelbart, and Joseph Rudnick (2004). Origin of icosahedral symmetry in viruses. Proc. National Academy of Sciences, USA, 101:15556-15560.