Biology 203...

I noticed my buddy Andrew has a link to this blog on his website, complete with a large title of "SCIENCE". I figure with such a kind endorsement, I really ought to post something about science. So let me tell you a bit about some recent experiments I've been doing.

Western Blot

Hopefully you remember the leukemia studies I'm working on. I've been busy trying to characterize these cancers, and today I'm going to describe a very powerful tool for doing that called Western blot. Before I discuss Western blotting, let me start with a related procedure called the Southern blot. I happen to be running both Western and Southern blot experiments, but the Southern is arguably more confusing, so I'll save the details for another day. But I mention it now because it was the first of the "blot" experiments to be invented.

Southern blot is unexcitingly named after its inventor, Dr. E. Southern. He developed a way to detect and quantify particular sequences of DNA in cells. Before long, the technique was altered to detect RNA, and was named "Northern blot" -- not because there's a Dr. Northern, but because of the play on Dr. Southern's name. Eventually, variations and combinations of the procedures led to Western, Eastern, and Southeastern blots as well.

The Western blot is the one I'll focus on today, and it too resembles the Southern and Northern blots, but it specifically detects protein instead of DNA or RNA. In my experiments, I'm interested in how leukemia differs from normal cells. If I can identify abnormal levels or actions of key proteins, they can potentially be used as treatment targets down the road. I won't get into extreme details, but let's say I think these cancer cells lack an important regulatory protein we'll call protein X (pX). pX restricts the action of other growth proteins, and thus without pX, the growth proteins go crazy. To test my hypothesis (that the cancer cells lack pX), I'll employ the Western blot to measure the levels of the protein.

Remember my post about antibodies, and how there can be an antibody specific for any possible protein? Western blotting takes advantage of this specificity by spreading out all the proteins in a tissue, and probing them with antibodies. Here's how it's done:

I start by making a soup of protein from my cells of interest. Specifically, I'm looking at protein in pre-B lymphocytes, but you can use any cell you want. I extract the proteins by breaking open the cells with a particular buffer solution. One problem with this is that the insides of cells also contain enzymes called proteinases that degrade proteins, and breaking the cells open exposes the two, and your proteins break down. To prevent this from happening, I add proteinase inhibitors to the solution.

After isolating the proteins, I treat them with reducing and denaturing techniques to uncouple and unfold the complex shapes proteins have. Incidentally, one of the reducing agents is beta-mercaptoethanol -- a chemical that smells so ridiculously awful, the bottle has a warning the says it is toxic because of the smell (think stinkbombs). Next, I separate the protein on an acrylamide gel by putting the samples at one end of the gel and running an electrical current through. The negative charge on proteins causes them to migrate toward the positive pole in the current, and as they move through the gel, they separate based on size (large proteins move more slowly through the gel than small proteins).

Now the proteins are spread all over the gel, but I can't really do much with them. If I want to analyze them, I need to get them out of the gel by transferring them to a special membrane. This is done in almost the same way as I made them migrate trough the gel, except this time, I use the electrical current to move them up and out of the gel. So if I put the membrane on the surface of the gel and apply the current, the proteins will lift out and stick to the membrane in the same pattern as they were before (sort of like lifting a drawing off of paper with Silly Putty). Now I have my proteins spread out by size, stuck to the membrane, and I can manipulate it however I like.

Now comes the "blot" of Western blot. I cover the membrane with a primary antibody for pX. In theory, it will bind only to pX, but proteins are notoriously sticky, and will bind to anything they can get their grubby little alpha-helices on. To prevent this, I give them a bunch of irrelevant dummy proteins to play with -- essentially "blocking" the nonspecific binding of sticky proteins. That leaves the antibody free to find pX. In the laboratory, we accomplish this by using a state-of-the-art protein compound called powdered milk.

Okay, I've covered the membrane with antibodies, and they have bound to my protein of interest, pX. Unfortunately I don't have a way to see where the antibodies have bound. Now I have to come in with a secondary antibody to find the primaries. This secondary antibody has a "marker" or "flag" stuck on the end that lets me see it. This flag can be a couple different things, such as a bioluminescent molecule, radioactive material, or -- in my case -- a fluorochrome. A fluorochrome is a molecule that emits light at a certain wavelength when stimulated by a different wavelength, thereby rendering the antibody detectable. Like this:



On the left of the picture is a "ladder" of known standard protein sizes, so I can confirm my protein is the right size. It just so happens that pX is ~62 kDa (kilodaltons), and as you can see, the more total protein I start with (10, 20, or 40 ug), the more pX I can detect.

And that, folks, is how a Western blot works.