« May 2010 | Main | July 2010 »

June 30, 2010

Thanks, Olivia!

Olivia Judson is ending her always-interesting weekly posts (often with evolutionary themes) at the NYT. Her penultimate post nominated yeast as Life-form of the Month. I second the nomination and look forward to blogging about some experimental evolution we're doing with yeast, once we publish.

June 25, 2010

Evolution meetings in Portland

The Society for the Study of Evolution, the Society of Systematic Biologists, and the American Society of Naturalists are meeting together over the next few days, in Portland, Oregon. I'm not going this year, unfortunately, but my grad students Ryoko Oono and Will Ratcliff will both be speaking.

Once again this year, there are no sessions scheduled on atheism, pornography, abortion, etc., just evolution. It's almost as if evolutionary biology were a science, maybe even a branch of biology! Reminds of a job fair long ago, where they told me "ecology" was too broad a specialty, but "biology" was OK.

Are scientists smarter than squirrels?

"Monkey-watchers often use the word "aunt" for an adopting female." -- Richard Dawkins, The Selfish Gene
The willingness of animals to adopt and care for orphans has been shaped by past natural selection. Often, Dawkins suggested, adoption represents "misfiring of a built-in rule... a mistake that happens too seldom for natural selection to have 'bothered' to change the rule by making the maternal instinct more selective." This seems a reasonable explanation for the failure of bird parents to kick "brood parasites" out of their nests, a situation I discussed recently.

But this week's paper, by Jamieson Gorrell and colleagues, seems to show that squirrels have a more-sophisticated understanding of selfish-gene theory than I would have expected. "Adopting kin enhances inclusive fitness in asocial red squirrels" was recently published in the new online journal Nature Communications. The authors analyzed five cases of orphaned squirrels being adopted, all by close relatives, and two cases where they were left to die, even though a relative had a territory nearby. In each case, they asked whether adopting would likely increase or decrease the frequency of the adopter's genes in future generations.

Closely related individuals tend to share gene variants (alleles) even if those alleles are rare in the overall population, so adopting a younger sister or a nephew who would otherwise die could increase one's genetic representation in future generations. On the other hand, adding an orphan to one's litter puts one's own offspring at somewhat greater risk. The authors were able to estimate this risk and compare it to the increased survival chances of the adoptee, weighted by its relatedness to the foster mother. If this benefit exceeds the risk, then Hamilton's rule (the fundamental equation of social evolution) predicts adoption. All of the adoptions that did occur met this criterion -- two cases were right on the line -- whereas the two potential adoptions that didn't occur failed the Hamilton's-rule test. Yet another example of squirrels solving challenging problems.

At least, that's what the data seemed to show. But the "relatedness" term in Hamilton's rule isn't necessarily equal to the relatedness we could calculate from a family tree or from genetic similarity. It would be, if helping an orphan had no negative effect on anyone outside one's current litter. But if there are more red squirrels than red-squirrel territories, then a surviving orphan may end up displacing another squirrel. So the question is, how closely related is that displaced squirrel likely to be to the adoptive mother? In the cases studied, 1/4 to 1/2 of the lactating females nearby were kin to the adopting mother. If that's a representative sample, then a surviving orphan might often end up displacing another squirrel that was as closely related to the mother as the orphan was. In such cases, the mother would have exposed her own litter to increased risk, without doing much to increase her genetic representation in future generations. Even so, the adoptive mothers aren't acting as maladaptively as Dawkins suggested (as if they adopted orphans at random), but their behavior wouldn't be optimal (by Hamilton's rule) unless there were unoccupied territories available nearby. Thanks to Dr. Carin Bondar, whose blog alerted me to this interesting paper.

Meanwhile, over at Science, Jeff Smith and colleagues propose "A generalization of Hamilton's rule for the evolution of microbial cooperation." When one cooperative act (releasing an expensive enzyme, say) benefits all microbes nearby, it's common to assume we can add up all the costs and benefits over a population. But what if twice the enzyme gives three times the benefit? The authors developed some high-powered math to deal with such problems and concluded that certain kinds of cheaters would have a harder time getting established than we would have expected from the simpler version of Hamilton's rule. Scientists are definitely smarter than squirrels, but they can't jump as well.

June 11, 2010

No evolution this week

Actually, evolution is continuing, but I don't have time to write about it this week. I'm still working on my talk for the North American Symbiotic Nitrogen Fixation Conference next week and I need to read parts of a couple theses before I leave, so my two PhD students can graduate on schedule.

June 4, 2010

Cancer's deep evolutionary roots

This week's paper, "Phylostratigraphic tracking of cancer genes suggests a link to the emergence of multicellularity in metazoa" was published in BMC Biology by Tomislav Domazet-Lošo and Diethard Tautz. Each of our cells is descended from an unbroken lineage going back to the first living cell. Most cells in an adult, however, are at the end of the line and will have no descendants. Exceptions include sex cells, stem cells, and cancer cells.

We consider cancer an aberration, but think back to the first multicellular life, which may have resembled Trichoplax. A Trichoplax has an upper and a lower layer of cells, and not much in between. They can reproduce by dividing in half, producing two offspring with hundreds of cells each (video). Or they can bud off propagules containing a small number of cells. They also seem to be able to reproduce sexually, from a fertilized single-cell egg, although complete development from an egg hasn't been documented. A Trichoplax can reform from separated cells, sometimes combining cells from two individuals. In such a chimeric organism, cells with different genotypes could compete for resources and reproductive opportunities, undermining collective success. Similar problems can occur when social amoebae get together to form a stalk for their spores. Even in a genetically uniform organism, a mutant cell could start reproducing (perhaps generating many propagules) at the expense of the whole. Today, we call cells that reproduce at the expense of the whole cancers, but something similar would presumably have been a problem for the earliest multicellular organisms.

Presumably? The authors of this week's paper used "phylostratigraphic tracking" to see when the ancestors of our cancer-suppressing genes evolved. Sure enough, there was an evolutionary burst of such genes right around the time when multicellular animals first evolved.