« June 2008 | Main | August 2008 »

July 26, 2008

Would the host want anyone to starve?

This week’s paper (Ratcliff, et al., in press) is from my lab.

The relationship between a legume plant and the rhizobium bacteria living in its root nodules is usually beneficial to both. Rhizobia convert nitrogen from the atmosphere into a form their plant host can use to make proteins. Rhizobia benefit from infecting plants because they reproduce more inside nodules than they would in the soil. They can also acquire certain resources in nodules. In particular, they can accumulate large amounts of energy-rich PHB – up to 50% of their own weight!

But there can also be conflicts of interest between rhizobia and their host plants. Resources used to make PHB could have been used, instead, to acquire more nitrogen for the plant. Therefore, mutant rhizobia that don’t make PHB provide their hosts with more nitrogen (Cevallos, et al., 1996). So why do most rhizobia make PHB? Our hypothesis has been that the rhizobia themselves benefit from having more PHB. This contrasts with an earlier hypothesis that rhizobia store PHB so that they can use the energy for the benefit of the plant. Consistent with our hypothesis, mutants defective in PHB metabolism reproduce less (Cai, et al., 2000). However, mutations can have complex interacting effects, making it hard to be sure that differences in PHB were the only cause of differences in reproduction. So Will Ratcliff decided to compare rhizobia that were genetically identical but had different amounts of PHB per cell.

He started by supplying rhizobia with a mix of resources that created favorable conditions for reproduction (by growing, then dividing) and also accumulating PHB. He started with a single cell, so all the resulting rhizobia were genetically identical, mostly. (A few mutations probably occurred, even in a few days, but conditions were similar to those under which the rhizobia had long grown in the lab, so any mutants would not be likely to increase by natural selection.)

So the rhizobia were almost all genetically identical, but did they vary in PHB per cell? To find out, Will added Nile red, which binds to PHB and fluoresces. Then he ran his cells through a flow cytometer. As each cell passed through a laser beam, its Nile-red fluorescence was measured, giving an estimate of PHB per cell. Analyzing many thousands of cells – this takes several seconds – showed that there was indeed a great deal of variation in PHB per cell.

The next stage was to separate the rhizobium cells into low-, medium-, and high-PHB groups. Will did this using a centrifuge, which spins to create a strong gravity-like field inside a tube. The tubes contained the rhizobia, plus a liquid with particles that separate in the field so that the liquid becomes denser with depth. Under these conditions, the rhizobia float up or down until they reach a depth where their density matches that of the liquid. (Similarly, someone with less body fat could float at the same depth in salt water as someone with more body fat floats in fresh water.)

Next, he starved each of the three rhizobium groups separately, for 160 days. We predicted that the high-PHB cells would survive longer. This turned out to be true. Not only that, but over the first month they actually reproduced. There was almost no food in their culture liquid – low-PHB cells did not reproduce – so we assume the energy for reproduction came from the PHB they started with.
There’s more, both in the paper and in terms of interesting results since then, but these results were enough to answer our main question. As we predicted, rhizobium cells can benefit from having more PHB. Therefore, the previously published tradeoff between hoarding more PHB versus supplying the plant with more nitrogen does indeed create a conflict of interest between plant and rhizobia. They also have shared interest, of course. Plants with more nitrogen can grow bigger and support more rhizobia. But what if this benefit is shared with competing strains of rhizobia infecting the same individual plant? We have shown previously (West, et al., 2002) that this could select for rhizobial cheaters, who invest in their own current or future reproduction (by hoarding more PHB, say) rather than in acquiring nitrogen for the plant. Cheaters would spread, we predicted, unless plants impose sanctions that decrease their survival or reproduction. So Toby Kiers, previously in my lab, checked for sanctions and found them (Kiers, et al., 2003). There are still plenty of unanswered questions, both for rhizobia and for other root-associated beneficial microbes. We will discuss some of these questions in a review coming out next year (Kiers and Denison, in press) and try to answer some of them in our ongoing research.


Cai, G.Q., Driscoll, B.T., Charles, T.C., 2000. Requirement for the Enzymes Acetoacetyl Coenzyme A Synthetase and Poly-3-Hydroxybutyrate (PHB) Synthase for Growth of Sinorhizobium meliloti on PHB Cycle Intermediates. Journal of Bacteriology 182, 2113-2118.

Cevallos, M.A., Encarnación, S., Leija, A., Mora, Y., Mora, J., 1996. Genetic and physiological characterization of a Rhizobium etli mutant strain unable to synthesize poly-ß-hydroxybutyrate. Journal of Bacteriology 178, 1646-1654.

Kiers, E.T., Denison, R.F., in press. Host sanctions, cooperation, and the stability of plant-rhizosphere mutualisms. Annual Review of Ecology, Evolution, and Systematics .

Kiers, E.T., Rousseau, R.A., West, S.A., Denison, R.F., 2003. Host sanctions and the legume-rhizobium mutualism. Nature 425, 78-81.

Ratcliff, W.C., Kadam, S.V., Denison, R.F., in press. Polyhydroxybutyrate supports survival and reproduction in starving rhizobia. FEMS Microbiology Ecology .

West, S.A., Kiers, E.T., Simms, E.L., Denison, R.F., 2002. Sanctions and mutualism stability: why do rhizobia fix nitrogen? Proceedings of the Royal Society of London B, Biological Sciences 269, 685-694.

July 17, 2008

More talks from Evolution 2008

I’m done with two grant proposals, revising a book chapter, and checking the final version of a review article. I still have a pile of interesting reading and writing to do before I can get back into the lab – actually, I did help Ryoko set up an experiment yesterday – but no more looming deadlines for awhile. So, here are two more summaries of talks from Evolution 2008.

Do I know you?

The ability to tell other individuals apart by their faces is presumably maintained by natural selection, so you can recognize and avoid bad guys. But is there also selection for looking different enough to be recognizable? Or is it better to blend in with the crowd, so you can get away with stuff?

Michael Sheehan and Elizabeth Tibbetts are studying individual recognition in wasps (Tibbetts and Dale, 2007). Their hypothesis is that distinctive-looking individuals benefit, because they get in fewer fights over dominance.

It’s better to be the top wasp, of course, but the worst situation would be having to fight to establish dominance, every time you ran into another wasp. With individual recognition, you only have to fight each wasp once; then you know which is dominant and don’t have to fight again. That was their hypothesis, anyway. To test it, they painted wasp faces, so that three in a group of four looked the same and one was different. Consistent with their hypothesis, the different one was attacked less by the others. Differences among individuals in appearance (or in calls, in birds) are more common in species that have more social interactions. Things get more complicated if you need to recognize individuals but also membership in some group (fellow colony members, close relatives, etc.). Then two or more different kinds of signals may be required.

Tibbetts, E.A., Dale, J., 2007. Individual recognition: it is good to be different. Trends in Ecology & Evolution 22, 529-537.


Your local food cooperative

Members of one species often have effects, positive or negative, on nearby members of other species. One example is “cross feeding?, where each species produces something the other needs. Such interactions may be mutually beneficial, but mutual benefit does not guarantee that an interaction will persist. Will Harcombe has developed a nice experimental system to study cross-feeding, using two species of bacteria that can, to varying extents, benefit each other. E. coli benefits Salmonella by breaking down lactose into a form Salmonella can use. E. coli needs methionine but can’t make it. Will started with a Salmonella strain that makes a little methionine, but not enough for maximum growth of E. coli. The Salmonella bacteria would benefit, collectively, if they invested some of their energy into making methionine, because that would stimulate E. coli growth and break down more lactose for them to eat.

But what if a mutant Salmonella pays the cost of making more methionine, but the benefits are shared with Salmonella cells that don’t pay that cost? Which genotype will win? Will hypothesized that, with a more structured environment, the benefits of making more methionine would be recycled locally, leading to more cooperation between Salmonella and E. coli, including more methionine production. To test this hypothesis, he allowed the two species to evolve on culture plates, where cells interact mostly with neighbors, and in mixed liquid culture, where methionine and other resources are equally available to all cells. On culture plates, Salmonella evolved to make more methionine, as predicted. In liquid culture, Salmonella mutants that made no methionine spread, leading to the extinction of E. coli. There was also an affect of crowding. When competition for resources was intense, there was less cooperation.

This system is simple enough that it can be modeled mathematically, for detailed comparison of theoretical and actual results. One simplifying assumption is that the cost of methionine production is constant; that is, making twice as much methionine costs a cell twice as much glucose (or whatever). But Will expects that natural selection could eventually favor mutants that make the same amount of methionine at lower cost, rather than those that make less methionine. I’ll be looking forward to reading about this work in the future.

July 9, 2008


I'm busy this week. Our last two grant proposals were rejected by NSF -- funding rates are down around 10% -- and my lab is almost out of money. (Some other time, I may address the question of whether there are too many good proposals, not enough money, or nonoptimum distribution of grants.) So I'm working on revised versions of those two proposals this week, with help from the grad students they would support. Then I have an overdue book chapter to revise before I can take time to blog, probably discussing more interesting talks at Evolution 2008. After that I have two interesting manuscripts sitting on my computer waiting for my input, one from a grad student in my lab and one from an Australian colleague I haven't met yet, before I can get into the lab and start some long-delayed experiments.

For those considering a faculty position at a research university, you do know that you will spend summers writing papers and grant proposals and (if you're lucky) doing research, not vacationing, right? On the other hand, I am rarely bored.