Mar 012008
ResearchBlogging.orgContinuing the banner day, Daniel G. Gibson and colleagues have published a rather subdued article in Science. In this truly impressive experiment, a group from the J. Craig Venter Institute synthesized and assembled an entire 583 kilobase-pair genome of Mycoplasma genitalium (1). Besides being a fairly cool piece of work all on its own, this outcome has profound implications for the future of bacterial research and the development of custom organisms

The stated purpose of the research is to investigate M. genitalium itself, which has the convenient property of possessing a very small genome. As the authors note (emphasis mine):
Approximately 100 of its 485 protein-coding genes are nonessential… it is not known which of these 100 genes are simultaneously dispensable. We proposed that one approach to this question would be to produce reduced genomes by chemical synthesis

One might reasonably conclude from this passage that JCVI has a dearth of small thinkers: this is akin to proposing to light your yard by shifting the position of the moon.

Yet Gibson et al. pulled it off using a tiered assembly method. They synthesized 101 overlapping pieces of DNA ~6 kb long. Using in vitro recombination they assembled groups of these into 24 kb fragments and propagated them using bacterial artificial chromosomes. Using a similar strategy they managed to continue assembling fragments up to about 144 kb in size; beyond this, however, the in vitro method did not work. So they turned to the yeast Saccharomyces cerevisiae to assemble the larger “C” fragments into half-genome and even whole-genome constructs. They then checked the whole genome for errors using the famous “shotgun” approach. Amazingly, their first run produced only two errors, both human in origin. After some wrangling they ironed it all out and ended up with precisely the genome they wanted, complete with special “watermarks”.

So they managed to construct an enormous hunk of DNA. What’s the benefit? Well, they can do those crazy experiments that they’re planning, for one thing. For another, assuming they can figure out how a mycoplasma works, they can design a new mycoplasma that does something special. For instance, they could design a new genome containing genes that allow the bacterium to eat toxic materials, or purify uranium out of the soil, or turn biomass into octane.

Granted, it is possible to get some of these genes into a bacterium using plasmids or BACs. However, what goes in can also go out—ejecting a plasmid is relatively easy, especially if it doesn’t provide any selective advantage. Completely ejecting a genomic cassette, while possible, is much more rare. Moreover, if you can custom-design the genome, then the custom enzymes can be interspersed with essential genes, making deletion that much more unlikely. Thus, once the genome can be manipulated at will, the only limitation on what an organism can be made to do is the extent of our ability to find or design proteins that fulfill the desired function. Figuring out which of a hundred-odd genes from M. genitalium can be done away with is just an efficiency step to make space for new genes to do whatever we want.

This approach will be a powerful tool for other experiments as well, don’t get me wrong. For instance, you could use the approach to place custom tags around certain genes so that they get excised at particular times or in a conditional manner, or you could swap sequences around to get a better idea how important gene arrangement is. But there’s no denying the enormous advance that this success makes possible, despite the authors’ decision not to speculate in the paper. The genome-synthesis technique is an essential step in the construction of a completely custom organism.

1. Gibson, D.G., Benders, G.A., Andrews-Pfannkoch, C., Denisova, E.A., Baden-Tillson, H., Zaveri, J., Stockwell, T.B., Brownley, A., Thomas, D.W., Algire, M.A., Merryman, C., Young, L., Noskov, V.N., Glass, J.I., Venter, J.C., Hutchison, C.A., Smith, H.O. (2008). Complete Chemical Synthesis, Assembly, and Cloning of a Mycoplasma genitalium Genome. Science, 319(5867), 1215-1220. DOI: 10.1126/science.1151721

 Posted by at 2:20 AM
Nov 212007
Blogging on Peer-Reviewed ResearchNot all papers are revolutionary—some merely provide an entertaining test of common assumptions. A recent paper by Taylor, Wedell, and Hosken in Current Biology is a case in point. One assumption of evolutionary biology is that males who are attractive mates produce attractive offspring; this is supposed to underlie evolutionary developments such as the peacock’s tail. Despite its intuitive appeal, this principle has not been put to the test very often. Taylor et al. made use of the fruit fly Drosophila simulans to investigate whether the assumption held up.

Their experiment was relatively simple. Isolated male flies were bred with 2-3 female flies, and the copulation latency, or the time between pairing and successful mounting, was recorded. This is a reasonable measure of male attractiveness because in D. simulans the female is in charge of the sex event—a male cannot copulate with an unwilling female. Granted, this was something of a desperation measure, the question being asked is, essentially, “If this was the last guy on earth, how long would it take for him to convince you to sleep with him?” A more convincing demonstration might have been to have females choose from different males, but as I am not a fly geneticist I’m not sure whether one could set up such an experiment. Additionally, it seems possible to me that the copulation latency might be affected by the attractiveness of the female, but I am not sure whether male flies are choosy. At any rate, the experiment might have been improved if repeated experiments with different females had been performed for each male to minimize bias arising from female attractiveness. This was done in the first part of the experiment, but not the second.

After maturing in isolation, so as to eliminate the possibility of learned behaviors affecting the results, the male progeny from these breeding events were isolated with females from a different population. The copulation latency was again recorded (no word on whether alcohol or dance music were supplied to help get things going). From these results, the authors concluded that the attractiveness of the father, as measured by the latency, contributes significantly to the attractiveness of the sons.

The authors did not provide much information on the nature of heritable features that give rise to attractiveness in their fruit flies, except to exclude body size as a factor. I myself don’t really know what the Brad Pitt of the fly world would look like, though maybe cousin Kathy has some idea, but wingspan, pheromones, and maybe some kind of auditory factor could all contribute and conceivably be heritable and account for the observations. A knowledge of the specific means by which “attractiveness” is transmitted isn’t relevant to the larger question of whether attractiveness is heritable, but matters when one wants to know whether the results in this model system can reasonably be extended to, say, birds or mammals.

Does this experiment tell us anything about the transmissibility of attractiveness in humans? Not necessarily. Mating behavior in humans is as much bound up in learned social behaviors as with unlearned biological ones. Many genes with the potential to result in male attractiveness are certainly heritable, but the diversity of lifestyles among humans means that the actual development of physical attractiveness from these potentials is uncertain. External non-heritable features (wealth, education, musical talent) also contribute significantly to human mating, which further complicates the issue. So I guess Dad is off the hook for this one, at least for the time being.

The article itself is relatively bare-bones and does not include any figures of raw data, which is something I never like to see. I don’t mean to impugn the authors—the journal may have decided not to make the expenditure. I think it would have been beneficial to see the correlation, and I can’t imagine any scientific reason not to include such a figure. Maybe it’s some biology thing.

Taylor, ML, Wedell, N, and Hosken, DJ. “The heritability of attractiveness.” Current Biology. 17 (2007) pp. R959-R960.

 Posted by at 2:52 PM
Sep 272007
As will be reported in this week’s Nature, French and Italian scientists have completely sequenced the genome of the common pinot noir wine grape (subscription required). As most wine connoisseurs would expect, the vitis vinifera genome is fairly complex, containing about 30,000 genes. While this is several thousand more than human beings, it is significantly lower than the number of genes noted in some earlier plant genomes, such as the poplar or rice. The most relevant comparisons cannot yet be made because the grape is the first fruit plant to be sequenced. However, some aspects of the annotations that can be made stand out noticeably. Relative to known genomes, the grape has a large number of stilbene synthases (responsible for the synthesis of resveratol) and terpene synthases (which produce compounds that give rise to wine’s complex flavors).

For the present, this sequencing effort is unlikely to produce any great revolutions in the wine industry—a significant body of work is required to relate flavor profiles to genetics, especially so considering the influence that soil chemistry, water abundance, and climate have on grapes and the resulting wines. As this research develops, however, it may prove possible by transplanting cassettes of synthetic genes from one grape to another, to introduce flavors currently unique to poorly-dispersed or finicky strains into hardier grapes that can be grown anywhere. Moreover, this research may provide a means to protect grape monocultures from diseases or climate change by identifying (and possibly correcting) genetic weaknesses.

The thing that strikes me is how overdue this kind of research is. I’ve already written in this space on the importance of agricultural research in the context of climate change; this is particularly true with regards to fruits. Most grains are grasses, and while this does not mean they are invincible, it does mean they’re likely to be relatively hardy—this is part of why they became so widely cultivated in the first place. By contrast, fruits and vegetables can be difficult to cultivate even in relatively good conditions. While grains are sufficient to sustain life, to ensure proper nutrition it will be important to make fruits and vegetables available. Understanding the genetics of these organisms is essential to that task.

It’s important to emphasize that sequencing a genome is only the beginning of understanding an organism’s genetics. Knowing what the genes you’ve sequenced produce, what stimuli control their expression, and what can be safely done to manipulate those stimuli or genes is the true aim of a genetic study. In this regard, plant studies lag significantly behind those in animals—the fact that many animals are significant models for human illness is a major reason for this. However, plants are the basis of our entire food supply, and therefore understanding their biochemistry may be of much greater importance in the near future. Research funding bodies, especially at the government level where the results will not be proprietary, should begin putting a greater emphasis on crop plant research in order to ensure the integrity of the food supply in light of climate change, declining diversity, and other challenges.

 Posted by at 6:46 PM