Grad publication: Mike Dorrity
The June issue of Plant Physiology features work I did in my previous lab at University of California, Davis. It's all about root development, and the main question it asks is: what is the genetic architecture underlying the length, angle, and cellular structure of the developing root?
In other words, what groups of genes are responsible for controlling a complex, pleiotropic trait like root length? We investigated this using two species of tomato; a domesticated breed (M82) and a wild relative (Solanum pennellii). S. pennellii is native to the coastal cliffs of Peru, and you may not even recognize it as a tomato relative considering how much time we’ve spent breeding varieties with the plump, red fruits we love (see photo). We’re talking centuries upon centuries here. The seedling root morphologies of these two species differ especially; M82’s roots are much longer than its wild relative, and S. pennellii’s roots consistently grow at a sharp angle compared to straight-shooter M82. To identify loci responsible for these differences, we took advantage of a population of ‘introgression lines’ (ILs) that were generated by crossing (and backcrossing and backcrossing and backcrossing...) M82 with S. pennellii. Each one of these true-breeding ILs contains a small fragment of the S. pennellii genome, while the remainder of its genome represents an M82 background. This concept is always a little hard to explain without a chalkboard, but these ILs are a very valuable tool; here’s an example to explain why. Imagine one IL (one with an S. pennellii introgression on part of chromosome 1) has a very short root. Now imagine a second IL (whose piece of wild-relative genome is in a different part of chromosome 1) with a root length very similar to its parent line, M82. Right off the bat, we can see that the region of chromosome 1 defined by the first IL contains something responsible for that line’s shorter root. Having the S. pennellii version of that chromosomal region makes that IL’s root morphology look more like the wild relative.
We took that idea and ran with it. We looked at early developing seedlings, and grew them on plates with plant medium for two reasons: 1. we cut out variation in environmental conditions to more accurately identify genetic components and 2. we can grow a lot of these guys when they’re little. We need those numbers for resolving phenotypic differences. Differences between wild-type M82 and S. pennellii root length, root angle, and cellular structure are very apparent (see pictures). However, the large population of ILs will represent a continuum of these trait values as genetic components are exchanged between the two species. As you might expect, we certainly found QTL responsible for significant chunks of M82 and S. pennellii’s differing morphology; I’ll let you investigate giant barplots yourself. Here’s one tidbit from the phenotypic analysis to think about if you’re interested, though. We see many ILs with ‘transgressive’ phenotypes, meaning these lines had trait values greater than their parents’ (an IL which had an even longer root than M82, for example). Interestingly, the wild variety’s trait values were never surpassed in the IL population. This observation could mean that S. pennellii’s short, angled root represents a kind of a phenotypic barrier encoded at the genetic level. If you’ve got any ideas about this, tell me about them; I love talking about this stuff.
You might be thinking to yourself: all that’s been identified here are these chunks of DNA that control these traits, big deal, what genes are responsible? What specific genes can, say, adjust the root growth angle? Well, that’s the next step we took (and will eventually publish). Using a whole-transcriptome approach, we sampled ‘snapshots’ of mRNA from all of these lines, and we sequenced it. With these data, we have access to the expression level of nearly every gene in the tomato genome (whose sequence was recently published). The very high similarity of the M82 and S. pennellii genomes makes this whole process possible. This approach (with lots and lots of replication) gives us the ability to identify changes in gene expression between these ILs and parents, and attempt to associate those with significant changes in a given trait value. Those genes which have the biggest expression differences are likely culprits, so we can generate mutants and check them out.
To move from QTL down to the gene level on the order of a few years is a very exciting prospect for identifying genetic underpinnings of complex traits. This has exciting potential both on the basic science level and application. S. pennellii retains drought, salt, and disease tolerances that are lost in domestic tomato. If we can identify genetic factors responsible for those traits or others, we may be able to inform breeding choices that support more robust agricultural varieties. Now if we could just generate a line that could flourish in the unkempt yard of my place in shady, cool Seattle...
-- Mike Dorrity
Mily Ron, Michael W. Dorrity, Miguel de Lucas, Ted Toal, R. Ivan Hernandez, Stefan A. Little, Julin N. Maloof, Daniel J. Kliebenstein and Siobhan M. Brady. (2013) Identification of Novel Loci Regulating Interspecific Variation in Root Morphology and Cellular Development in Tomato. Plant Physiology June 2013 vol. 162 no. 2 755-768.