Researchers Show How New Viruses Evolve, and in Some Cases, Become Deadly
[Via NSF News]
Researchers at Michigan State University (MSU) have demonstrated how a new virus evolves, shedding light on how easy it can be for diseases to gain dangerous mutations. The findings appear in the current issue of the journal Science.
The scientists showed for the first time how the virus called “Lambda” evolved to find a new way to attack host cells, an innovation that took four mutations to accomplish. This virus infects bacteria, in particular the common E. coli bacterium. Lambda isn’t dangerous to humans, but this research demonstrated how viruses evolve complex and potentially deadly new traits, noted Justin Meyer, MSU graduate student, who co-authored the paper with Richard Lenski, MSU Hannah Distinguished Professor of Microbiology and Molecular Genetics.
[More]
This is a fascinating study. The lambda virus infects E. coli through a surface protein called LamB – the protein was even named because of this property. So, in the experiment, some E. coli evolved resistant forms dealing with LamB to prevent infection. Then lambda developed a new form that found another way to infect the cell by using another protein, OmpF.
These new viral forms needed at least 4 separate mutations to accomplish this. And all 4 changes were needed to become fully infective.
The key aspect to get from this is not that the 4 mutations needed to occur all at once for them to appear. These changes were co-evolving with the changes in the LamB protein, so the necessary mutations became more prominent in the viral population as the E. coli also changed.
Essentially, as changes in LamB appeared – to reduce infectivity of thevirus – changes in the lambda virus were selected for – changes that brought back some of that infectivity. As LamB made further changes, lambda changed more until the fortuitous step when all 4 mutations created a virus that no longer needed LamB to enter the cell. This new form could use OmpF.
And only the ones with all four changes could use OmpF.
One of the criticisms lofted at natural selection is that intermediary forms are not useful. They say. “Lambda can only use OmpF to get into cells if all 4 mutations are present, right? So how do the 2 or 3 mutation forms even get any purchase in the population? It would seem that only viruses with all 4 mutations present simultaneously – a very rare occurrence – would be selected for.”
But this work shows how the intermediate changes appeared, were selected for and finally resulted in an entirely new activity.
Now, think if we had only two forms of bacteria – one with only LamB and one with only OmpF. The viruses that infected each would have 4 different changes between them but we would see no direct way for the virus to get from one form to the other with any intermediary stages. Each viral form only infects the one population, not the other.
But there were intermediate forms at one time.
This is the sort of scaffolding model used against the above criticism. Nothing evolves in isolation. These sorts of co-evolution events happen all the time. Then if the intermediate forms disappear, there does not appear to be any linkage, even though there once was.

This nice diagram was in the issue of Science. The co-evolving populations of E. coli and phage create a ridge – or scaffold – between changes in the gene frequencies of the two populations. Continuing selection can remove the ridge, resulting in what appears to be two different ‘species’ of E. coli and virus.
The other interesting aspect of this work was that the effect was contingent but repeatable. That is, they started the selective culture of the bacteria and virus in multiple flasks. Not every flask developed the new mutant viruses but those that did had the same sorts of mutations in the viral genome.
This demonstrates the random nature of natural selection – not every population will have the same success dealing with a changing environment – while also showing, however, the repeatability of the process.
In addition, they showed that the selective events of the bacteria also drove the evolution of the viruses. Bacterial populations with slightly different starting genomes produced different results. The chance of the viruss to develop the new ability depended on the bacterial populations in which they found themselves.
And they end their paper stating that the bacterial population in several cases had developed resistance to the OmpF mediated viral infection, thus showing that the co-evolution was not at an end and suggesting that the virus would have to find a new method to infect the cell.
This sort of evolution was accomplished in weeks with the right selective pressures. What would happen with centuries or eons?
by 

