GABA(A) - the A should be subscript - binds two molecules of GABA and then changes its shape slightly to form a pore that allows chloride ions to pass through. Since the concentration of chloride in extracellular fluid is generally always higher than the concentration inside the cell, this results in the diffusion of Cl- ions into the cell (which normally has a negative charge to begin with). Resting potentials vary from cell to cell throughout the nervous system, but tend to be around -70 millivolts. When GABA channels are activated the cell hyperpolarizes to about -80 milllivolts. In order to fire action potentials and signal other cells, a neuron must depolarize - to a threshold of about -50 mv in mammalian cells. Thus, the GABA receptor prevents action potentials, and is considered inhibitory.
The other type of GABA receptor is GABA(B), which binds two GABA molecules and then through a series of intermediary proteins opens up other potassium channels, which pump K+ *out* of the cell. (Potassium is maintained at a higher concentration inside the cell than outside). The net effect is the same, although it's achieved a little slower.
In our lab we work with the GABA(A) receptor from rats. These receptors are formed by five subunits which, when viewed outside the cell, would probably look something like a pie with five wedges and a big hole in the middle. There are several classes of subunits - alpha, beta, gamma, rho, and epsilon. There may be more, but those are the only ones we work with. And within each class there are several subunits. The GABA(A) receptors we study contain the following subunits: beta2, alpha1, beta2, alpha1, gamma1. Much of the structure of this receptor is unknown, but we do know that the binding "pocket" for GABA molecules is formed at the interface between a beta and an alpha subunit. Thus there are regions on both that have binding properties. Our lab has already investigated some regions on each subunit, but there are a few that remain. One of which is a particular protein loop (called loop X ) on the Y subunit. This is my domain :-)
The problem with studying membrane-bound proteins is that it is almost impossible to visualize them with x-ray crystallography, a technique commonly used on intracellular proteins. I say almost because a lab recently cut off a snippet of the acetylcholine receptor and visualized that successfully. So anyway, we have to use tricky indirect methods to get at the structure of the GABA receptor. What we do is to go through the particular region of protein we're interested in, and create a series of mutants. In each mutant, one amino acid has been changed to cysteine. Once we have the mutants we want (I do not have all ten of mine just yet, but *hopefully* they'll be ready by next weekend), the next step is to produce RNA in vitro. We then inject that RNA into single-celled, unfertilized frog eggs, called oocytes. These oocytes then start expressing the mutant gaba receptor proteins.
From there we can do lots of things to gather data. Since I haven't gotten to that point myself I am not terribly certain on what exactly we will do. The most basic procedure, though, is to see if the functioning of any mutant receptors is drastically different from normal. We do this by attaching patch clamps to an oocyte to read current. We then "puff" various concentrations of GABA molecules onto the receptor, and construct graphs of concentration of GABA versus current evoked. Mutant receptors will always have slightly different graphs from normal ones, but sometimes it sticks out like a sore thumb (lets say that 1,000 times more GABA is required to evoke the same current). Basically, that would tell us that the amino acid that we mutated into cysteine is a direct part of the binding pocket.
Once I have all of my mutants ready there will be a ton of data to gather, and the gathering is pretty tricky. So some of the other grad students and/or postdocs in the lab will help out with it. The end result, though, is that there a paper will [hopefully] be published with my name on it. Phew!