Tuesday, December 2, 2008

Why put an Arabidopsis gene in a carrot?

There was a post on Gristmill a few months ago criticizing a Wired article on some research published in PNAS by scientists in Texas* which showed that a genetically modified carrot contained more bioavailable Calcium. I'm not going to focus on the main criticism of the gristmill post: that the increase in Calcium isn't enough to be nutritionally relevant and is being oversold as a way to promote GMO (genetically modified organisms). Instead I want to focus on this part:

By the way, guess which organism provides the gene that gives us this nutritional breakthrough? Arabidopsis thaliana, in the Brassica family -- a cress. Maybe we should eat more Brassicas: kale is pretty darn high in calcium. Nah, let's stick them Brassica genes somewhere exciting -- down where the sun don't shine (where carrots grow).

because it illustrates a disconnect between the perceived motivations for choosing the particular gene/food combination and the actual reasons that they were chosen by the researchers.

Now let me first concur with your mother that you should always eat your vegetables, and lots of them. Carrots, Kale, cress, you name it, their good for you. But thats beside the point. The story of this work goes back to 1996, when the Kendal Hirschi, the PI on this work, discovered this gene, a calcium transporter in Arabidopsis, by searching through libraries of yeast, each expressing a single Arabidopsis gene. His lab went on to publish many more papers on this gene and the gene family it belongs to (see a comprehensive list here), including expressing the arabidopsis gene in Tobacco.

Now I have already mentioned three organisms, lets look at why they were chosen for this research. All three of them are "model" organisms, they have been used to answer lots of questions and researchers: know a lot about them and know how to manipulate them easily.

Yeast is a workhorse of molecular biology. Arabidopsis is a model plant because (among other things) it is small, grows quickly, and can self fertilize so you can get a line which has the same genotype generation after generation. These characteristics make it easy to do genetics, just like mendel laid it out several centuries ago. Mendel had time and space to let his peas grow, and could get one generation a year, so his experiments took 7 years. We don't like to wait that long. Arabidopsis also has a small simple genome, which is why it was the first plant to be sequenced.

Tobacco has been used as a plant to express other genes because it is easy to transform (put a piece of foreign DNA into the genome) and easy to culture tobacco cells in a petri dish.

So the organisms were selected for their suitability for the experiments that were going to be conducted. Through all of this, the researchers learned more about the gene through a wide variety of experiments. From a researchers point of view, you want to move forward with the reagent (the gene in this case) that you know the most about. So even though there might be similar genes in other species, you don't know what might be different about how the gene works.

One of the things that the authors noted when they put the gene into tobacco is that there was much more calcium accumulation in the roots than in the shoots. So if you want to improve the calcium content of a food using this gene, you need a plant where the roots get eaten or you need to spend a lot more effort to figure out how to make them put more calcium in the leaves, which could take a lot more work. Before you do that work, it would be good to know if the increase in calcium would help the people who eat it (bioavailable is the term, animals can't take up some minerals if they are in the wrong chemical form).

So if the authors wanted to test if the increase in calcium was would be bioavailable for people who ate it, they needed to put the gene in a transformable food plant where people ate the roots. And that is why they chose carrots.

*The PI on this work is a friend of mine, so if you think I'm biased about this, well, you might be right.