Are you saying that above a ligand-to-protein ratio of 0.5 no additional ligand binding was observed? If so, this could be for one or more of 3 reasons: (1) there is one ligand binding site per protein dimer (in which case the protein concentration should be halved); or (2) half the protein is unable to bind the ligand due to denaturation or occupancy of the ligand binding site; or (3) at least one of the concentrations of the species is inaccurate.

Thanks for the response Adam. It is Protein-ligand ratio (~1:0.5). I mean, there is a saturation at ~1:0.5 ratio (0.1 um of protein: 0.05 um of Ligand) . Did I answered your view?

Hi Senthil, sorry I am a (very) old school biochemist.

To determine Kd of [P]+[L]-> [PL], the absolute [P] is irrelevant as long as it is constant and  [L] is >>than [P] providing you can determine either [PL] and/or [L] with varying [L].

Being an old dude, I would like to see some binding curves with constant [P], with varying [L] plotted with changing signal.

I would like to see a saturation curve with constant [P] with varying [L].

Thank you for the answer Steingrimur. Yes, The concentration of protein [P] is constant with varying [L] concentration. Now, in order to plot the curve,  the [L] concentration is in the negative range ( the Kd value is around sub nanomolar). Now, my problem is how to plot the curve ?  Probably i may need some dedicated software ( like origin-lab) for curve fitting. 

Hi Senthil, I think we are talking different ligand binding curves here. In saturation binding curves of protein binding, the [L] can never be negative.

Kd measurements for protein/ligand (P/L) can be similar to Km measurements for enzyme/substrate (E/S).  

Simple Km estimations for enzyme/ligand, the [E] is constant and [S] is varied (x-axis). The enzyme activity, generally expressed as V (substrate conversion/time) is expressed on the y axis.

This means that an enzyme can only operate at a certain V (Vmax). The [E] will saturate meaning adding more [S] will not increase V past Vmax. Km is the 1/2 of Vmax.

Similarly, simple Kd estimations can be done by varying [L] (x-axis) and measuring the binding (y-axis). [P] will saturated. 1/2 of the saturation will be a Kd.

As in Kd or Km measurements, us old guys prefer to see simple saturation curves with positive values.

Thanks for the comment. I agree with your explanation and I am trying to solve the issue.

Hi, 

Not sure if you are still trying to do this. IMO first goal is to get good protein signal (can go above 0.1uM!) as binding events will split the signal intensity of peaks you see. Then make up say 10 aliquots of protein (for 10 data points on curve) at double this concentration (say 2uL aliquots at  2 x 'good intensity' concentration). Then to these aliquots add 2uL aliquots of ligand in MS buffer at different concentrations, whatever range you need to go from just visible binding to saturation 1:0.01 -> 1:0.5 perhaps?). Measure the spectrum for each concentration of ligand then plot data to get binding curve.

If you export the spectral data as text you could use some software called Unidec http://unidec.chem.ox.ac.uk/ to automatically extract peaks, plot and calculate Kd based on different binding models.

Check out Hoper et al. Mass Spectrometry Quantifies Protein Interactions—From Molecular Chaperones to Membrane Porins Angew. Chem. Int. Ed. 2014, 53, 14002 – 14015 and refs within for a good starting point 

@Joseph Gault, Thanks for the answer. I have managed to do the Titration and even I found the Kd. Thanks for the tool.