« Troll refuge may prevent local extinction | Main | Splitting species: sneak attacks from strategic hamlets »

Less-vicious viruses evolve in viscous cannibal populations

This week’s paper is “Local interactions select for lower pathogen infectivity? by Michael Boots and Michael Mealor, University of Sheffield, published in Science (vol. 315, pgs. 1284-1286) and suggested by my wife.

The evolution of greater or lesser infectiousness in pathogens has important implications for health of plants and animals, including humans. Evolution is a process that follows its own rules and humans can’t control it completely, but we can sometimes influence it, just as we may be able to constrain the course of a river or limit the spread of a forest fire.

One factor over which we have some control is the ease with which a pathogen spreads from one host individual to another. For example, a bacterium on the skin of one patient in a hospital can’t jump to another patient in a different room, but it may be able to hitch a ride with a doctor or nurse who forgets to change gloves between patients. Intestinal bacteria reach new hosts easily if untreated sewage is dumped into the same river used for drinking water, even if the bacterium is so virulent that the host is too sick to walk around and infect others.

Paul Ewald has suggested that easy transfer between hosts favors the evolution of greater virulence (Oxford Surveys in Evolutionary Biology 5:215-245). For example, cholera spreading through South America in 1991 evolved greater virulence in countries with poor water supplies, but lesser virulence in countries with better water supplies. So not only were more people infected in countries with polluted water supplies, but the infected people were sicker.

Virulence (disease severity) isn’t the only trait that can evolve, of course. What about transmission, which depends on the number of bacteria or virus particles released by a host, the chances that each will reach another host, and infectiousness (the chance that contact will lead to infection)? What we need is an experimental system that lets us control the ease of pathogen movement between hosts while measuring evolution of all these traits. Any volunteers to "host" this experiment? I didn’t think so, and anyway it would take a village.

So Boots and Mealor used caterpillar larvae that get infected with a virus, mostly by eating dead infected larvae. They controlled movement by controlling the viscosity of the medium in which the larvae lived. (A “viscous population? is one that doesn’t mix with neighbors much, for whatever reason, but in this case movement was controlled by physical viscosity of the medium.) After about eight generations, they harvested virus from high- and low-viscosity populations and measured their ability to infect a standard caterpillar population. The virus populations in low- and medium-viscosity media hadn’t evolved, but those swimming in oobleck had evolved much lower infectivity.

Boots and Mealor suggested that, when movement is limited, more infective viral genotypes would quickly run out of uninfected (and therefore susceptible-to-infection) hosts. (The caterpillars and viruses in a local area might all die out before another host wanders by, just as measles died out in the rarely visited Faroe Islands between 1781 and 1846.) A less-infective mutant, on the other hand, would make the limited local supply of hosts last long enough for a ship – I mean a caterpillar – to arrive from elsewhere.

Can we extrapolate from one study with cannibal caterpillars to humans or wheat? Not with much confidence. On the other hand, if a large number of experiments, conducted by different scientists, using different species and different methods, find the same general pattern, then the results might apply to our crops and perhaps to ourselves. Another study consistent with that of Boots and Mealor was published last year in Nature (442:75-78) by Benjamin Kerr and others, here at the University of Minnesota. It’s past its “use-by date? for This Week in Evolution, but maybe I’ll discuss it in my other blog, The Comedy of the Trojans, since it's about a "tragedy of the commons" for viruses.

Even if there were no effects on the evolution of infectiousness and/or virulence, reducing pathogen transfer between hosts is usually something we would want to do anyway. In the case of humans, this could involve hand-washing, face masks, and monogamy or condoms rather than restrictions on travel per se. Adding in the evolutionary benefits could make improvement of public water supplies, screening windows against mosquitoes, and enforcing glove-changing rules in hospitals even more cost-effective.

The same issue of Science has a good discussion of Boot’s and Mealor’s paper by Angus Buckling (we’ve both published with Stuart West, but hasn’t everyone?), an article on brain evolution, an experimental molecular biology paper on the evolution of a membrane protein that confers antibiotic resistance in bacteria (gene duplication is involved, as in the hormone-receptor paper discussed last week), and newly discovered 500-million-year-old fossils from the Burgess Shale that apparently shed light on the evolution of mollusks.

Next week: "Adaptive evolution in humans revealed by the negative correlation between the polymorphism and fixation phases of evolution"


Very fascinating. Thank you for the summary, as always.

Rumors of my author infidelity are greatly exaggerated

So basically the disease/parasite that has less opportunity to spread has less inclination to kill its current host. Is that what this demonstrates or did I miss the point?

Good question. Your summary of the opening discussion is on target, so long as you understand that this is an evolutionary change in the relative frequency of genotypes in a pathogen population over generations, not a change in an individual pathogen. Unlike humans and scrub jays (http://www.sciencenews.org/articles/20070224/fob5.asp), viruses can't think through the future consequences of their actions. So it would be more accurate (but longer) to say that the pathogen population contains different genetic variants, some of which are more virulent than others (i.e., more likely to kill the host, maybe because they reproduce so rapidly in the host). When transfer to another host is rare, the more-virulent variants are more likely to die out (because, as you say, they will often kill the host before a transfer opportunity occurs) leaving the population dominated by less-virulent variants.

This week's paper shows that similar dynamics apply to infectiousness (rate of infection, rather than severity of disease), but in this case a more- infectious variant spreads so fast within a group of hosts that it may kill them all. This is OK for the more- infectious pathogen variant if it was able to transfer to another group first (low viscosity treatment) but with limited mixing between groups (high viscosity) only the slower-spreading variants will survive.

Very interesting article which makes me think about the ways our human behavior can affect larger, evolutionary trends. Thanks.

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

Type the characters you see in the picture above.