You should try to doing a saturation binding first. Try incubating the labeled and the unlabeled peptides together (1:1). and separately incubate labeled and the control - solution without test peptide (1:1) For this assay use a high potent unlabeled peptide to get a better saturation curve. You can make a series of dilution with the labeled peptide and keep the unlabeled peptide concentration constant. Generate the total and nonspecific binding curves and check whether the labeled peptide has a lot of non specific binding. If so you can not expect to see a decrease in binding as you mentioned. Also when you add the labeled peptide later, the receptors could have recycled hence show no difference in biding. Hope this give you some idea.

As Dilani Dehigaspitiya said, you should do in vitro work first before moving into cells, where there are a lot more unknowns. Does the fluorescently-labeled peptide changes its emission properties upon binding to the receptor (i.e. does it increase its emission intensity?). Once you know that, you can make a titration with the labeled peptide and obtain your first Kd. Then you should move to the competition experiment itself. I would prepare a solution of the receptor and the labeled peptide at the lowest concentration of the peptide at which the signal is saturated, and then progressively add the unlabelled peptide, monitoring the changes in fluorescence. For competitive 1:1 binding the analysis is relatively straightforward to get the second Kd.

Keep in mind that if you are not seeing any changes upon addition of the second peptide, then it means that the second peptide binds more weakly.

Wang, Z. X. FEBS Lett. 1995, 360, 111–114.

For starters, your assay is not ideal. A competition assay assumes equilibrium or near equilibrium conditions. Binding the probe before the competitor pretty much assures that you are not in equilibrium. Peptide binding to specific receptors is usually pretty tight and this means very slow dissociation. I would recommend adding both peptides simultaneously.

Second, microscopy is not the best way to measure this, in particular if you are using your eyes to detect the differences. Human eyes are designed to detect contrast, not light intensity, and a 2-fold change is surprisingly difficult to see. You would be much better off using a plate reader with the proper filters.

Third, the assay you are describing is pretty much the same that has been used for the past 50 years, only using fluorescence rather than radioactivity. The way I would do it is:

Treat cells in a multicellular plate with a fixed concentration of Fluor-peptide in the presence of increasing concentrations of unlabeled competitor. Do not do it the other way around. Increasing the concentration of the labeled peptide has the unwanted consequence of increasing your background. I would cover two orders of magnitude (or more) of the competing peptide concentration. Do the incubation in the cold (15 degrees C is fine) to avoid endocytic traffic of the labelled peptide. After a while (1-2 hours) wash the cells with cold PBS 3-5 times and measure the fluorescence using a plate reader. Plot fluorescence intensity vs concentration of the unlabeled peptide. The midpoint should be your dissociation equilibrium constant.

It is unnecessary to purify the receptor, which is a rather difficult thing to do anyhow.

Correction: it says above "multicellular plate"; it should say multiwell plate. Blame the autocorrect function of my stupid tablet.

Guillermo, The description you make of the cell assay seems fine, but I think he should characterize the system in vitro before moving to cells. On the other hand, simultaneously adding both labeled and unlabeled peptides to the receptor does not make much sense to me. Tight binding and preformation of the complex does not imply that he is not in equilibrium conditions and has no relation with binding kinetics; we have worked with slow binders with microM binding and fast low nM binders.

A couple of good references about competitive titrations:

Hulme, E. C.; Trevethick, M. A. British J. Pharmacol. 2010, 161, 1219–1237.

Thordarson, P. Chem. Soc. Rev. 2011, 40, 1305–1323.

Roehrl, M. H. A.; Wang, J. Y.; Wagner, G. Biochemistry 2004, 43, 16056–16066.

Eugenio, this has nothing to do with slow or fast binders. It is rather about slow or fast dissociators, so to speak. If you prebind a peptide that dissociates very slowly ( like insulin, or parathyroid hormone, for instance), it will take hours before you reach an equilibrium with the competitor. Just do the math. Ninety (actually 87.5) percent equilibriation is achieved in 3 half-lives of the slowest reaction in a one-step, two-component system. PTH dissociates from the receptor with a koff of about 10^-5 at neutral pH. The half-life of the reaction is therefore over 1 hour. So, if you prebind some labeled PTH to your receptor and then add a 1000 fold excess of the competitor but you only wait 1 h, you will exchange less than half the prebound PTH. How is that an equilibrium?? Adding them simultaneously is not only better. It is the only correct way to do it, and my lab has been doing it this way for about 35 years.

Furthermore, how is it simpler to do an in vitro experiment unless you purify to at least some extent the receptor? Crude membranes are just a little less complicated than whole cells, and a lot more artificial. Pure receptors require detergents, and then you have to assume that the detergent does not alter the binding properties. Or alternatively you must test many detergents to find something that does work -and to know that it does work, you have to compare it with cells, which is what you set out to do in the first place.

Do the cells first and spend your time wisely troubleshooting and optimizing the procedure.

Guillermo, I think your systems tend to be much, MUCH slower tan ours, therefore our approaches are also different. With small peptides we've always seen almost immediate equilibration, even with tight binding peptides with low nM binding constants. Only with large (>50 aa peptides) we've had some problems with folding/binding and equilibration times, so each point had to be measured after 15 min of equilibration. With small peptides we've never had trouble pre-forming the complex with the probe and then displacing it with the silent ligand. This is the standard procedure for us and has been reported many times also by other groups. Then again, your system might not allow that approach due to kinetics.

Of course if he does not have the receptor isolated, then cell assays are more straightforward, my mistake on that :-)

Dear All

Thanks for your really useful suggestions. Guillermo, I incubated the cells with the unlabeled peptide (25 and 50 µM) for 1 hr and after incubation washed and incubated with labeled peptide (1 µM) for 1 more hour. I could not see any difference in microscopy. I decided to go with flow cytometry but the histogram showed a very minor difference in the positive cell population (5-10%). However when I checked the mean intensity of the sample, it was a huge difference between competition and un competition sample (40%). I further incubated the cells with unlabelled and labelled peptides simultaneously and interestingly the mean intensity decrease by 90% and the histogram also showed a very significant shift to left. I guess the flow cytometer counts a cell as positive whether it has 100 labeled peptide bound to it or 1000. However the mean intensity clearly showed the difference.

Furthermore the differences in mean intensities of sequential or simultaneous assays might be due to receptor recycling. Please give your opinion on this.

I am planning to start with fluorometry. I am not experienced and I have very limited quantity of peptides. One well of 96 well culture plate per peptide would be sufficient to detect differences or I need to culture cells in 2-3 wells and then lyse and transfer them in one well before measurement.

Please suggest