Friday, April 15, 2005

Did someone say, "data"?

(I'm going to ignore the fact that it's the 15th, and I haven't yet cracked the book. I've been busy getting data. Give me a few days, mkay?)

I've been working on this yeast two-hybrid. It's the sexiest experiment ever, because it exploits the innate genetics in yeast for a truly holy purpose. And since I love it so much (I'd marry it, if I could), I'm going to explain it to you.

The point of the yeast two-hybrid is to find out whether two proteins interact (bind) with each other.

Background: genes are encoded in DNA, but genes are not always expressed. So a yeast cell can have a gene that codes for "survival in the absence of histidine" (the gene allows the yeast to make its own histidine, a nutrient), but the yeast will only survive in the absence of histidine IF the gene is expressed.

More background: What determines whether a gene is expressed? There's a promoter region next to the gene which is responsible for promoting gene expression. So what usually happens is that a protein (called a transcription factor) will go to the DNA where the gene is, bind to the promoter region, and promote gene expression.

The transcription factor has two parts to it: the DNA binding domain and the activation domain. (Go ahead, guess what they each do.) And these parts are modular, so they can be separated into two different proteins and still work, as long as they can be brought next to each other.

That is what happens in the yeast two-hybrid: You take your favorite protein A, and fuse it to the DNA binding domain. Then you take your favorite protein B, and fuse it to the activation domain. If proteins A and B bind to each other, the activation domain is brought into proximity to the DNA binding domain (which binds the promoter region), and the gene is expressed. You can then test for binding by testing whether the yeast grow in the absence of histidine. It's really very beautiful.

So I'm doing this with six proteins: A, B, C, D, E, and F. And I'm looking for interactions between each possible pair. And a little more complicated than that. But I finally have all my preliminary data for that. And I can draw an interaction map, and it makes sense! It even agrees with previously published data (which isn't as definitive as mine). So I'm really very excited.

No comments:

Post a Comment