I’m going to get off a quick summary of this afternoon’s talks, then I have to run down to the poster session to find out what the grad students have been doing. Are we having fun yet? I’m going to collapse in bed tonight, and then unfortunately I have to catch an early flight back home, so I’m going to miss a lot of cool stuff tomorrow.
First up after lunch was Deneen Wellik, who summarized her work with Hox genes and patterning the vertebrate axial skeleton. I’m spared some effort here — I already wrote up her paper, so read that for the whole story. In short, one of the confounding things about studying the Hox genes is that there are four banks of them, so there is a lot of redundancy; knocking out just one has fairly subtle effects. Wellik has been deleting whole sets of Hox genes in mice, and finding that their loss creates homeotic transformations of regions of vertebrae into more anterior regions.
Daniel Meulemans described Gene regulatory novelty and the origins
of the vertebrate head skeleton. Much of the vertebrate skull is the product of neural crest cells and ectodermal placodes. Meulemans was looking for the origins of these cranial innovations by comparing gene expression in the cephalochordate Amphioxus and the lamprey, and found that certain classes of genes, specifically those involved in specifying neural crest tissue, were expressed in lampreys but not in Amphioxus at the time and place where neural crest would form. This was not just a descriptive, correlational study, though — he also carried out overexpression and rescue experiments. In the lamprey, one experiment is to over-express genes involved in neural crest specification, which causes more cells to be recruited into the neural crest population and produces an assayable surge in those cells. Injecting extra copies of the Amphioxus SoxE and Snail genes did the same thing, producing more neural crest. In addition, there is a zebrafish mutant that has lost the homologous Sox10 gene — these mutants have no pigmentation or enteric neurons, both of which are derived from neural crest. Injecting Amphioxus SoxE rescues the zebrafish mutant; it appears that the Amphioxus gene is fully equivalent to the zebrafish gene. So why does Amphioxus lack neural crest? One reason is because other regulators are not present in Amphioxus to activate SoxE and other neural crest associated genes.
Detlev Arendt told a familiar story about Evolution of the central nervous
system in animals. He works on a polychaete worm, Platynereis, and his story was that we have astonishing similarities between chordates and this worm in the molecular markers that distinguish certain specific cell types in our brains. I’ve also written up a couple of his papers, one on the homologies in the CNS of worms and chordates, and another on his work with evolution of the eye. This was good stuff, but I’d been conditioned at this time by some of the earlier talks to think that we definitely need more taxa in these comparisons — it’s a strong story that will be even stronger with some more intermediates.
Fred Nijhout sort of blew me away with Multiple and diverse mechanisms for
robustness in a complex network. It wasn’t at all what I expected: it was solid biochemistry. Whoa. I really had to strain my brain to keep up, but it was fascinating stuff that I’ll have to digest for some time. One of the big issues in developmental biology is robustness or canalization, a property of many systems which exhibit remarkable consistency and flexibility in the face of environmental and genetic challenges. You can knock out significant genes, or expose them to utterly horrible environmental conditions, and embryos tend to go ahead and produce fairly normal organisms (in most cases; they can also fail spectacularly). So the question is how do embryos do that?
Nijhout took a very different approach to this problem. What we need is a specific developmentally significant pathway which is well understood, complex but not too complex, and that can be approached mathematically. He’s looking at one carbon metabolism. Metabolism! It’s biochemistry! He was talking about enzyme kinetics! I was flashing back to my junior year in college, when I lived immersed in charts of biochemical pathways … it wasn’t an entirely pleasant feeling.
But it was OK, there was some familiar ground. The one carbon metabolism pathway is a set of reactions that are important in DNA synthesis and repair, DNA and histone methylation, dopamine and serotonin metabolism, and responses to oxidative stress — it’s a pathway with fingers in lots of pies, being involved in cancer, neurobehavioral disorders, cell division, and development. This is the pathway in which folate is central; it’s why pregnant women need to take a sufficient amount of folic acid, to prevent neural tube defects. So what Nijhout has done is to quantitatively model the whole pathway, using the known properties of the enzymes, substrate concentrations, etc., and showed that it exhibits developmental robustness. Plug in extra amino acids, just as typically happens after you eat a meal, and some parts of the system respond with wild variations in concentrations and reaction rates, but key parts are stabilized, with, for instance, purine and pyrimidine synthesis chugging along at a steady pace. And of course the real virtue is that he could trace the process through the pathway and show exactly and quantitatively how the biochemistry dictated that behavior. It was hard stuff for someone unaccustomed to talks in biochemistry, but a good way to stretch the mind.
OK, now the posters. This is going on until 10:00 local time, which is midnight to my midwestern clocks — I’m going to have to hope my stamina holds up here.