More Recent Comments

Saturday, August 01, 2009

Perspectives on the Tree of Life: Day Two

The second day began bright and early with a talk by Jeffrey Lawrence, professor of biology at the University of Pittsburgh, Pittsburgh (PA, USA).

The title of his talk, Fragmented speciation in bacteria: The failure of the coalescent model doesn't really reflect its significance. Jeff showed us his careful analysis of speciation in bacteria focusing on a small clade of bacteria including Escherichia coli, and its closest relatives, Citrobacter kosteri, Salmonella enterica, Enterobacter sp., Klebsiella pneumonia etc.

When "speciation" occurs in bacteria there is considerable lateral gene exchange that obscures the actual relationship. What this means is that there really is no single common ancestor of most pairs of species and, given that it is almost impossible to define a species in the first place, a single phylogenetic tree of organisms is not a correct depiction of the tree of life.

Imagine a situation where our human ancestor was hanging out with the ancestors of gorillas and chimpanzees and a lot of gorilla ancestral genes made it into our genome. We wouldn't be able to depict the true representation of our evolution by just showing a simple bifurcating tree where we are closest to chimpanzees and more distant from gorillas. This treelike representation would ignore the transfer of genes between our ancestor and that of gorillas. You need a net or a web to show the actual path of descent. That's the situation in bacteria, as we saw clearly in the presentation.

But the failure of genomes to coalesce to a single ancestor doesn't make the whole exercise useless. It's still possible to say that Escherichia coli, Citrobacter kosteri and Salmonella enterica are more closely related to each other that any of them are to Klebsiella pneumoniae. Jeff gave us some estimates of the frequency of lateral gene transfer but I didn't write them down. The general consensus at the meeting is that DNA from another species is incorporated at the rate of once every few million years. This is enough to seriously compromise a treelike phylogeny.

Greg Morgan is a philosopher who is currently at a college in New Jersey. He gave the second talk. His title was Defining biodiversity in a world with horizontal gene transfer. He was mainly critiquing a book by the philosophers Maclaurin and Sterelney on biodiversity. Their concept of a species doesn't take lateral gene transfer into account and it doesn't really pay attention to the microbial world and the definition of species. Greg emphasized that there are many ways to define biodiversity and it makes a difference if you are interested in preserving biodiversity.

Laura Franklin-Hall is a philosopher at New York University in New York (NY, USA).Her presentation was an attempt to root for the underdog by defending trees: Scientific models and the history of life: Deep disagreement or mere misunderstanding?

I wish I had time to explain her description of scientific models and how they work but I didn't take enough notes and, besides, I'm still digesting the information. The idea is that no model is perfect—they all suffer from various weaknesses—and we shouldn't necessarily opt for the model that incorporates the most detail. In other words, while a web or a net may be a more accurate description of descent with modification, there's still a lot to be said for a tree if it represents some underlying, but partial, truth.

Maureen O'Malley is a philosopher at the University of Exeter and a former colleague of Ford Doolittle's. She is one of the behind-the-scenes organizers of the meeting. Her talk was on The philosophy of evolution, Ernst Mayr and the tree of life. She gave a really interesting summary of Mayr's views on evolution—a view which obviously didn't consider microbes and didn't concentrate on genes ("beanbag genetics").

I think its fair to say that most evolutionary biologists in the room found it interesting to review the history of Mayr's thoughts, but that's all it was; history. The philosophers, on the other hand, were much more attentive. Apparently Mayr is thought of as one of the founding fathers of the philosophy of biology and an "attack" on Mayr is close to blasphemy. That's shocking to me and it suggests that philosophers of biology are very much out of touch with modern evolutionary biology. This observation is consistent with some of the things that went on at this meeting although I hasten to add that most of the philosophers here seem to have a much more modern perspective. They weren't the least bit unhappy to see Mayr dethroned.

(I gather that photographs of Maureen are unusual. That's why the group photo at the top of the page is so important. She's right there in the second row.)

Eric Bapteste is a philosopher at IHPST in Paris (France). He told us how important it is to incorporate Lateral thinking about trees into our view of evolution. By "lateral thinking" he means the idea that genes can be inherited horizontally and not just vertically.

Lisa Gannet is a philosopher at St. Mary's University here in Halifax (NS, Canada). Her talk was on the difficulties in reconstructing the history of human populations and the various ways of representing it. Much of her talk was peppered with references to racism and it became apparent that she is very concerned about whether a tree of humans populations might support racism. After her talk I asked her about this and it turns out she is one of those people who don't think that human races exist.

The next talk was by James Mallet, a biologist at the Galton Laboratory, University College London, London (UK). The title of his talk certainly got everyone's attention: Was Darwin wrong about the nature of species and speciation?. Jim began by quoting Ernst Mayr, who in several papers, made the point that Darwin was wrong about speciation. Mayr quoted directly from Darwin's works to support the idea that he (Darwin) had an incorrect view of speciation.

Jim demolished that idea by showing that Mayr was as guilty of quote-mining as our typical creationist friends. It turns out that Darwin's ideas on species and speciation were far more accurate than Mayr was willing to admit and it's time we acknowledged that. Jim went on to illustrate real examples of speciation by looking at various races and varieties of butterflies in Venezuela. Just as Darwin imagined, there's a smooth continuum from varieties, races, and subspecies to real speciation events and it's often difficult to distinguish between races and species. Many species interbreed to form hybrids, as he showed using his data, but the hybrids might be quite rare indicating that gene flow between the species is almost nonexistent.

The last talk was by Richard (Dick) Burian, a philosopher at Virgina Tech in Blacksburg (VA, USA). He spoke about "Some conceptual issues deriving from the challenges to tree thinking. The emphasis was on identifying the "process" that can explain the "pattern" of evolution. This becomes difficult in the face of chimeras and symbiosis. Dick tried to establish the criteria that these processes had to fulfill in order to be legitimate contenders in the explanation of evolution. One of the criteria was selection so I asked him whether random genetic drift was a process that could be used to explain the pattern of evolution. The answer was "no." After the talk we had a discussion about this and we both agreed to think about it.

We all went off to dinner at The Cellar restaurant. Much alcohol was consumed and this contributed greatly to the discussions. As usual. we went home satisfied that we have solved most of the problems and woke up forgetting the solutions. We'll try again tonight. I'm more convinced than ever that this is the best meeting I've ever attended. I'm learning things at such a furious rate that my brain is beginning to feel the pain.



14 comments :

John S. Wilkins said...

It's Dick Burian, not Burain.

JLT said...

"Much alcohol was consumed and this contributed greatly to the discussions. [...] I'm learning things at such a furious rate that my brain is beginning to feel the pain."

I'm not sure that you've identified the real cause for you hurting brain there ... ;)

DK said...

The general consensus at the meeting is that DNA from another species is incorporated at the rate of once every few million years.

Sounds like underestimate by many orders to me. Judging from the fact that simply taking E.coli and mixing it with DNA (no other treatments) about 1 in 10^10 cells takes up the DNA.

Bayesian Bouffant, FCD said...

It's Laura Franklin-Hall, not Laure.

"The general consensus at the meeting is that DNA from another species is incorporated at the rate of once every few million years."

Is that figure restricted to unicellular prokaryotes (the context of the paragraph in which it appears), or is it meant to apply across the spectrum?

Rosie Redfield said...

@DK: It's not enough to bring a DNA fragment that contains novel genetic information into the cell. The fragment has to escape degradation by intracellular exonucleases (very active in E. coli), be incorporated into the genome by non-homologous recombination (very inefficient), and persist in the genome for millions of years. The later may be the big obstacle, as most foreign DNA is useless at best, and random incorporation is likely to interfere with genes already present in the chromosome.

I'd trust the estimates derived from real evolved genomes more that those based on what can happen in the lab.

DK said...

Rosie:

If I am not mistaken, several viruses carrying DNA from other species made it into H.sapiens genome during my lifespan. Somehow then, I think I have a valid reason to doubt the "once in a million years" number.

The 10^(-10) I mentioned refers to transformation, so the DNA has already survived and persisted for many generations. The rest you mention (incorporation into chromosome and persistence for a million years) don't strike me as particularly rare. In many species, direct ligation/illegitimate recombination is ~ 1000X more frequent than homologous recombination. And judging from how much junk DNA is there in genomes that can afford it, persistence is obviously not a problem.

I'd trust the estimates derived from real evolved genomes more that those based on what can happen in the lab

I wouldn't. At least not without seeing how the estimates were made. Just think of a number of E.coli cells that existed on planet Earth during 10^6 years. I have a feeling it is a very large number, likely ~ 10^20.

I am not a specialist and I may very well be wrong but based on what I know, my gut feeling is that the estimate is wrong. Calls for authority won't change this feeling, too.

Anonymous said...

In fact, the very existence of these potent exonucleases (RecBCD) in E. coli argues that E. coli is regularly bombed with foreign DNA such that an effective defense is required. Presumably there was a time when these exonucleases weren't as efficient as they are today, yet you'd expect the rate of foreign DNA update wasn't correspondingly reduced in those earlier times.

Rob Beiko said...

It's Eric Bapteste, not Baptiste. I wasn't going to point it out, but no one can resist a pile-on :^>

I would dispute the idea of a 'consensus' about the rate of interspecies DNA uptake. There is no 'the' rate, because lineages differ in their propensity to take up, integrate, and retain genetic material.

The rate will also be non-uniform with respect to time, since close proximity and adaptive potential of introgressed DNA will strongly influence the rate of 'productive' (sensu Rosie) uptake and retention. In some cases, it may be like eating a bag of potato chips: once you start taking up genes from a certain other species, you take up an ever-increasing about of sweet, sweet foreign DNA. There is something seriously weird going on in many thermophiles, for example.

Finally, what do you mean by 'species'? :^>

ERV said...

... Is this it?

No one said anything about viruses?

*sigh*

:(

Larry Moran said...

ERV asks,

No one said anything about viruses?

One person talked about bacteriophage but that was pretty much it.

We don't care about nonliving things. :-)

ERV said...

*HULK SMASH*

RAAAAAAAAR!!

LOL!

Rob Beiko said...

Eric included phage and episomes as separate entities in his vertical-free networks. And Chris Malaterre had a '0.5 living' generalization of his Many Roots ideas.

perk said...

I was to write a paper on Kenneth Schaffner's rendition of developmentalism from the point of view of bacteriophage biology. I argue that the fact that a viablephage can be produced from purified DNA and host cellularcomponents lends some support to the anti-developmentalist, ifthey first show that one can draw a principled distinctionbetween genetic and environmental effects.

Kevin said...

This treelike representation would ignore the transfer of genes between our ancestor and that of gorillas. You need a net or a web to show the actual path of descent. That's the situation in bacteria, as we saw clearly in the presentation.