This post is the fourth in a series that aims to educate readers about the tools that are used in neuroscience research. Previously we discussed Novel Object Recognition, Calcium Imaging and EEG.
Neurons communicate with one another by releasing small molecules called neurotransmitters that bind to specific receptors on adjacent neurons. The interaction between the receptor and the neurotransmitter (or ligand) is very strong and the two are said to fit together like a lock and key. An interesting feature of the brain is its ability to regulate the number of receptors expressed at the surface of the cell. For instance, exposure to nicotine results in robust increases in the number of receptors at the cell surface. So while initial exposure to nicotine may stimulate the majority of receptors, a subsequent dose of equal size will only stimulate a fraction of the upregulated receptors. A mechanism like this may underlie some features of addiction such as tolerance and withdrawal.
Because neurotransmitter binding to receptors play such a crucial role in brain function, scientists need tools to quantify them. A classic way of doing this is by exposing ground up brain tissue to radiolabeled molecules (or radioligands) that bind to complimentary receptors found in the sample. Because the difference between a radioligand and a normal neurotransmitter is very small (only a single extra neutron) and the radioactive signal can be detected very easily with a scintillation counter, they are ideal probes for quantifying receptor binding. By incubating the brain tissue with the radioligand, washing the reaction mixture to remove any excess and measuring the remaining radioactivity, you can quantify binding. Repeating this assay with multiple concentrations yields a binding curve. Replicating each assay condition 2-3 times and averaging will also improve the quality of data dramatically.
Here is a made-up example of what binding assay data looks like. Our radioligand will be H3-spiperone, which binds strongly to dopamine D2 receptors. On the x-axis we have the various concentrations of radioligand tested, ranging from 0.5 nanomolar to 1 nanomolar. On the y-axis we have the number of binding sites labeled at that concentration, expressed in femtomoles/microgram of protein sample:
But there’s a problem. While the majority of the radioligand will be bound to your receptor of interest, some will bind non-specifically to other proteins in the ground brain sample. In most cases, even multiple washes cannot remove this nonspecific binding. Furthermore, non-specific binding will prevent us from reaching a binding plateau because, unlike specific binding, non-specific binding does not saturate with higher concentrations of radioligand. Looking at our fictional data, we can see that the y-values do not reach an maximum but rather continue to rise. How do we calculate specific binding from this data? By measuring the nonspecific binding and subtracting this from total binding, it is possible to calculate specific binding. In other words, Total Binding – Nonspecific Binding = Specific Binding.
Non-specific binding can be measured by displacing some of the radioligand by incubating with very high concentrations of another molecule that binds to all your receptors of interest. Unlike the first ligand, this one is not radio-labeled. For instance, we might use Sulpriride, another specific ligand for D2 receptors. Those molecules of radioligand that represent D2 receptors should be displaced by Sulpriride. When the brain tissue is washed, the only radioactivity that remains corresponds to the nonspecific binding because it didn’t bind another known ligand for that receptor. Because the unlabeled competitor is present in great excess, it out-competes the radioligand for a limited number of binding sites.
After replicating each assay condition with and without an unlabeled competitor, we can subtract the two curves to find specific binding. Here is what this data might look like:
Binding increases rapidly at the lower concentrations while reaching a plateau at the higher concentrations. This maximum number of binding sites is usually referred to as the Bmax value and it is expressed as moles per microgram of protein sample. Here we can see that binding levels off somewhere between 400 and 500 fmol/ug. The concentration of radioligand at which binding reaches half the Bmax is known as the Kd. The lower the Kd for a given ligand, the higher it’s affinity for the receptor in question; the higher the Kd, the lower the affinity.
Once you have a specific binding curve, all that is left to do is fit the equation using a non-linear regression. If all goes well, the data should fit the following equation:
Y = (Bmax)*(X)/(Kd + X) where X = radioligand concentration
When X is equal to the Kd, the term (X)/(Kd+X) will equal 1/2 and therefore the Y value will be half of the Bmax, as expected. When X becomes very large the same term will approach 1 and the Y value will approach the Bmax value. For our fictional data, the a non-linear regression reveals that the Bmax is 500 fmol/ug and the Kd is approximately 0.1 nanomolar. This makes sense because at 0.1 nM 3H-Spiperone, binding is at 250 fmol/ug.
Once you have the method down (easier said than done), it’s possible to investigate how the receptor’s properties change with different experimental treatments. For instance, the literature reports that chronic treatment with the antipsychotic haloperidol upregulates D2 receptors. This kind of change in the quantity of binding sites could be detected as a change in the Bmax value, as measure with the binding assay discussed above. In contrast, an experimental treatment that altered the affinity of the binding sites for H3-Spiperone would be detected as a change in the Kd value. A shift toward the left would correspond to increased affinity (lower concentration sufficient to occupy half the binding sites) while a shift toward the right would correspond to decreased affinity (higher concentration necessary to occupy half the binding sites).
Interested in learning more about binding assays? Here are some links that might be of interest:
- Our fictional experiment has been done by real scientists. Find out how their results compare to ours.
- Here is a paper about upregulation of nicotine receptors that makes use of saturation binding assays.
- And here’s one about upregulation of D2 receptors.
- Here is a helpful website for learning about the analysis of radioligand binding data.
- And here you can find out how to use Microsoft Excel to perform non-linear regressions. I also have a pre-made spreadsheet for this purpose which I can share upon request. (In case you haven’t figured it out yet, most of these Method of the Month features are about techniques that I have or currently do make use of in my labwork.)
- Millipore is the manufacturer of choice for binding assay equipment. A little on the pricey side, but my guess is that obtaining radioactive compounds is even more prohibitive for the home tinkerer.