Why is my Kd and Ki different - Homogeneous Radioactive Competitive Binding Assay

[Laynne]

Hello!

I'm new here and I've come across a question that I'm unsure about and have done some research but I couldn't find the answer.

In my experiment, I've used single concentration radioactive raclopride and increasing concentrations non-radioactive raclopride ligand
Effectively, my Kd should be the same as my Ki

However, when I've calculated my Ki of non-radioactive raclopride it is different from the Kd of radioactive raclopride

Why is that?

Many thanks for your help

 

[alkap555]

just a shot in the dark but i've had similar problems with ligand depletion from physical adsorption to the actual 96 well plate (plastic).

It could be that as you add increasing amounts of cold ligand, a higher proportion of the hot ligand gets displaced and binds non-specifically, thereby giving you false affinities, since non-specific binding is assumed to be proportional to free ligand in soln.

Or you could be saturating some non-specific binding site in your 96-well plate/brain homogenate/cell lysate etc with your higher concentration cold ligand, effectively giving you a falsely high Ki (now i'm starting to confuse myself!).

The (radio)ligands I work with are notoriously lipophillic [similar to N-desmethyl (on the benzothiazole) thioflavine T derivatives...so they are neutral compounds] and LOVE to associate with all sorts of plastics, so I've had to switch to glass or pre-coating my plastic 96-well plates with some type of silicone oil IIRC--supposed to help, haven't tried it yet.

Hope that makes sense

Edit: Are you using tritiated raclopride?

 

[Laynne]

Thanks for you answer, I kinda understand where your coming from and it sounds right from what you've written there.

So overall, its due to these non-specific binding - such as binding to filters and equipment and also binding to cell membranes of the HEK cells were using will cause the difference between the Kd and Ki values of the radioligand and the non-radioligand of raclopride - Is that correct?

Yes I was using tritiated raclopride.

 

[alkap555]

Yes, I think that could be an issue, principally due to the fact that the hot vs cold ligand concentrations vary in terms of concentration.

Granted I am no expert in binding assays, but I have done my fair share, however they have all been done on either crude brain homogenate from Alzheimer's (and other neurodegen. disease) patients, and purified proteins.

If you google "ligand depletion" there are some FAQ/troubleshooting websites that may be able to point you in the right direction.

http://graphpad.com/curvefit/ligand_depletion81.htm

http://graphpad.com/curvefit/ligand_depletion7a.htm

http://www.pdg.cnb.uam.es/cursos/ho...t_Lab/Radioligandos/Mailman_Boyer/lesson2.htm

I'll keep you posted if I come up with anything else.

Good Luck

Edit: other things to consider...I have also had problems of ligand forming aggregates resembling micelles at higher concentrations, which get stuck on the filter and give you extremely high readings. Best course of action is probably seek out someone in pharmacology dept or something and see if they can help troubleshoot, or depending on the equipment manufacturer of your equipment they can have good technical support.

 

[Laynne]

Wow thanks very much for your help I really appreciate it haha.
I just noticed I had written so many typo's in my previous reply haha, lack of sleep eh.

But yeah, I'll be writing up a lab report soon about this assay and was concern of the difference if Ki and Kd so would be a rather good topic to bring into my report

Thanks very much again!

I'll post whatever I encounter while writing my report :D

 

[Hyperthesis]

Is use of a detergent an option? I think about Triton X-100 or alike. This could prevent formation of precipitates and unspecific binding. We use this material regularly in our microplate-based assays (...targeting enzymes, not receptors, though) to avoid false-positive hits by 'promiscuous ligands' (see eg. Nature Protocols 2006, 1(2): 550–553; DOI: 10.1038/nprot.2006.77).

 

[endotropic]

In our competition binding assays we pre-coat all of our plates and pipette tips with a compound called Sigmacote (http://www.sigmaaldrich.com/catalog/product/sigma/sl2?lang=en&region=US) to prevent our drugs sticking to the plates/tips. This is probably the "silicone oil" alkap mentioned.

The stuff isn't indicated for plastics, but we've found it's effective in reducing non-specific binding of notoriously "sticky" compounds. I have no connection with Sigma by the way, this is just something that my lab uses for our binding assays.

 

[alkap555]

All good points, I forgot to mention non-ionic detergent--we use that as well depending on the assay. Also typically run about 5% EtOH but that is due to the poor water solubility of our ligands. Tween-20 usually at like 0.01% IIRC (that might be the conc for co-immunoprecipitation with Prot A/G Dynabeads)

BTW ill go ahead and plug Invitrogen's Dynabeads for (co)immunoprecipitations on solid phase support. They beat the hell out of sepharose beads cause you separate them with a magnet rather than centrifugation, so you can work on crude homogenates (insolubles), rather than supernatants only. Not to mention way lower non-specific binding and overall convenience. (And no I have no affiliation with Invitrogen).

Yea Sigmacote is what I was talking about. According to the Sigma website its a "special silicone" (lol) soln in heptanes.

 

[Laynne]

Wow you guys are awesome! haha.

I've decided to pre-coat all my plates and have repeated the experiment the same way as explained above. I don't know if it will make difference but I should have included more information when asking a question.

The experiment started with expressing HEK cells expressing D2 receptors only. (via transfection methods)
The next step was to observe the inhibitory effects of the non-radioactive raclopride in increasing concentrations, with a single concentration of radioactive raclopride.
The IC50 was determined and used to calculate the Ki of the inhibitory ligand.

With comparing the Ki and Kd of non-radioactive and radioactive raclopride they observed to be different, when in theory, they should be the same.

So what we've established so far is that the different in Ki and Kd is due to:
1. Non-specific binding to cell membranes
2. Non-specific binding to experimental filters and equipment when recording the results
3. Poor seperation phase of bound and unbound radioactive raclopride ligand

Would there be any other possible factors that could deviate my Ki and Kd values?

 

[nuke]

Ki should never be the exact same as Kd, I had thought (it's early in the morning, so sorry if this is stupid). Ki is based around effective inhibition/excitation of a biological function (Cheng-Prusoff equation), whereas Kd (for enzymatic reactions: Km) is a measure of affinity of a ligand/substrate for a receptor/enzyme. Hence Ki is arbitrary and centered around the measurement of some biological effect; for instance, lisuride will always have the same Kd for the same cloned 5HT2A receptor, however the Ki will depend on how you measure receptor activity, eg phosphatidylinositol turnover or alternatively diacylglycerol release.

 

[endotropic]

Actually nuke in homologous competition experiments the Kd and Ki should theoretically be identical, since your hot and cold ligands are essentially identical. In a heterologous competition with different hot and cold ligands you're absolutely right.

Also in a radioactive competition assay the Ki is just the concentration of the unlabeled ligand required to compete off half of the labeled ligand, so it's not necessarily the "inhibition of a biological function." The labeled ligand could be an antagonist and it wouldn't matter, all you're measuring is how effectively you can compete it off of the receptors.

 

[alkap555]

^^Yep.