Poliovirus evolution: the strong silent type

Abstract

The high mutation rates of poliovirus and other RNA viruses promote viral diversity by rapid sampling of mutations. In this issue, Lauring et al., 2012 report that underlying synonymous codon variation in virus populations can have a considerable impact on adaptive potential with implications for viral fitness and virulence.

Pathogens often operate at the brink of elimination. This can be especially apparent for RNA viruses where high mutation rates increase the prospect of genetic escape from immune recognition, but also produce scores of mutations detrimental to viral fitness. Indeed, altering the fidelity of viral genome replication with drugs like ribavirin can nudge populations into ‘error catastrophe’ where deleterious mutations begin to outweigh any benefits gained by adaptive ones (Pfeiffer and Kirkegaard, 2005; Vignuzzi et al., 2005).

Many of the mutations tipping the balance of viral fitness are nonsynonymous, meaning they alter the amino acid composition of virus-encoded proteins. In high-stakes genetic battles for survival at host-pathogen interfaces, even single amino acid changes can be ‘game-changers,’ which modify the structure or activity of viral proteins in ways consequential for evading, countering, or exploiting host defenses. There is growing evidence that synonymous substitutions, which do not change amino acids, also have a substantial influence on viral fitness. In this issue, Lauring and colleagues report on a clever set of experiments with polioviruses differing only by synonymous changes to reveal how viral genomes may be ‘silently tuned’ for increased adaptability. They describe a case where wildtype viral populations appear to have better odds of avoiding deleterious mutations while gaining access to beneficial ones compared to another strain differing only by synonymous changes.

Historically, synonymous substitutions have been considered largely neutral in terms of impact on evolutionary fitness. However, the burgeoning alliance between evolutionary and molecular biology, bringing together models and mechanisms through hypothesis testing, continues to reveal ways in which synonymous substitution can influence RNA stability and splicing, as well as the efficiency of protein translation and folding, with impacts on fitness (Chamary et al., 2006; Wilke and Drummond, 2010). Examples include studies where recoded viruses, engineered with hundreds of rarely used codons by synonymous substitution to accentuate codon bias, resulted in highly impaired virus replication owing to decreased translational efficiency (Mueller et al., 2006; Coleman et al., 2008). In this issue, Lauring and colleagues examine a different mechanism for how synonymous variation might play a fundamental role in how viruses adapt and compete against host defenses. They consider the underlying evolutionary potential of different codons encoding the same amino acids to produce distinct mutations, both synonymous and nonsynonymous, which could differentially influence viral fitness.

In a scientific twist on the recycling ethic, the authors use synonymously recoded polioviruses that served as controls from previous studies on codon bias to gain new insight into the adaptive potential of viruses. One strain, called SD, was meticulously engineered to minimize the effects of codon bias by swapping different codons for the same amino acids between different locations in the capsid gene. The result is a virus with over 900 synonymous changes, but at first blush, seemingly wildtype characteristics. Supporting this idea, the authors show that translation of the recoded capsid protein and single cycle replication of the virus are both nearly identical to wildtype parameters. Importantly, the structural properties of the RNA genome also appear to match the wildtype strain. However, on infection, SD shows variable plaque morphology, hypersensitivity to ribavirin treatment that lowers the fidelity of replication, and performs poorly in competition assays against other poliovirus populations.

So what can explain the lower fitness of the SD virus relative to wildtype? Placed in an evolutionary framework, the exciting implication of this study is that the wildtype virus is less likely to accrue detrimental mutations and more likely to hit upon beneficial ones compared to the SD strain, as these virus populations adapt to host defenses by sampling different mutations (Figure 1). Consistent with this model, the authors perform deep-sequencing analysis of wildtype and SD virus populations after several rounds of infection and find increased differences in variation in the recoded capsid region than in the rest of the genome. These differences appear to underlie the higher fitness of wildtype virus populations, and intriguingly, include variable regions of the capsid recognized by neutralizing antibodies. Together, these results lend new experimental credence to the idea that synonymous variation in RNA viruses influences the potential of these populations to adapt rapidly to host defenses with important implications for immune evasion and viral fitness.

Figure 1.

Model for RNA virus adaptation

Although differing only by synonymous substitutions and equally capable of single cycle replication, the WT (red) and SD (blue) virus populations have different levels of fitness as they sample mutations during adaptation to host defenses. The differing ‘clouds’ of variation in virus populations are indicated for WT (light red) and SD (light blue) poliovirus strains.

This work adds to a fascinating and growing body of research describing how cryptic genetic variation plays important roles in evolutionary adaptation. One recent example involving small catalytic RNA species revealed that benefits of underlying genotypic variation emerged when populations adapted to new substrates (Hayden et al., 2011). An emerging theme of these studies is the idea that populations are ‘pre-adapted’ to new environments owing to otherwise unappreciated genotypic differences revealed by natural selection. One of the strengths of the current work is the ability to compare the fitness of poliovirus synonyms – wildtype and SD strains – against the strong selective challenge imposed by host immunity. Extending their experiments beyond cell culture infections, Lauring and colleagues made the key observation that SD virus is also at a significant disadvantage in mouse models of infection and virulence relative to the wildtype strain. These results support an important role for cryptic variation under the selective intensity of host-pathogen evolution in the context of infection.

As with many thought-provoking studies, this investigation of poliovirus evolution also raises interesting questions. The ‘sequence space’ available to viral genomes for sampling synonymous variation is enormously vast and has been barely explored, even when including every study of synonymous variation ever conducted thus far. For instance, when the authors tested another poliovirus strain called Max, which was optimized for host codon pair usage with more than 500 synonymous capsid mutations, they found no reduction in fitness compared to wildtype. These results hint at a dynamic complexity underlying adaptive outcomes that could be balanced by multiple opposing consequences of synonymous variation on fitness. Future studies placing viruses in different locations of ‘synonymous space’ to test fitness will provide a more complete picture of the contributions of synonymous variation to adaptive potential. For example, what role does synonymous variation play in altering the probability of generating certain mutations with either beneficial or negative consequences due to epistasis, as was recently examined with an in vitro evolution study of antibiotic resistance (Salverda et al., 2011). Taking advantage of the ability to conduct prospective studies of experimental evolution with virus populations like these will provide opportunities to determine if and how viruses evolve back to synonymously optimized locations (Bull et al., 2012), and determine the impact of synonymous variation following long term adaptation to new hosts.

In terms of evolutionary ‘bang for the buck’, acquisition of adaptive nonsynonymous mutations can have a significant impact on viral fitness. But, as this work beautifully demonstrates, some virus populations seem well-positioned via ‘silent’, synonymous optimization to minimize mutational risk while increasing the odds of landing adaptive combinations of old-fashioned, nonsynonymous haymakers, which can be crucial for knocking-out host defenses.