Fitting of heterogeneous ligand model vs 1:1 Binding model

Hi, Around

Thank you for always giving me lots of advice.

I am currently analyzing "Receptor - Ligand" binding using CM5 chip.

Theoretically, since the two bindings are 1: 1 binding, set the Sensorgram fitting model to 1: 1 binding and proceed kinetics calculation.

When I analyzed it, I was worried because it was not always fitted neatly.

a. A result seems to have a bulk effect even though the solvent is exchanged.
b. Fitting does not fit generally in the dissociation phase.
c. Fitting is better when there is no saturation in the association phase
(means short contact time, low analyte concentration make better sensorgram fitting)

When I analyzed it using the Heterogeneous Ligand model, I found that the sensorgram was fitted almost completely regardless of the concentration - high or low.

In my opinion, this phenomenon may be caused by the lack of directionality of the amine coupling method.
(I think that I might be wrong)

In this reason, I would like to ask you for some advice.


1. Is it possible to use Heterogeneous Ligand method in this case?

: Theoretically, I think it is appropriate to use a 1: 1 binding model but heterogeneous model because my ligand is homogeneous (consist of one protein). I wonder if it is possible to use a heterogeneous ligand model in my case.

2. How to improve analysis with a 1: 1 Binding model

: In my case, I want to use CM5 chip because of price and convenience. How can I solve this problems?

Thank you for reading.

Have a nice day!

Hi,

Thank you for your kind words.

When I analyzed it, I was worried because it was not always fitted neatly.

This is a vague statement. With an example it is better to understand how the fitting is deviating from your measured curves.

a. A result seems to have a bulk effect even though the solvent is exchanged.
b. Fitting does not fit generally in the dissociation phase.
c. Fitting is better when there is no saturation in the association phase

a. don't worry about small bulk deviations (<5% of response). In general bulk should be compensated for with the reference surface. An extra compensation can be the 'volume exclusion correction' ( www.sprpages.nl/experiments/data-processing ). Make a series of the same dilutions as with your ligand but now with the dialysis buffer since that is where the bulk difference is likely to come from.
b. This is possibly pointing to a heterogeneous interaction, but can be rebinding of the analyte. Try increasing the flow rate and check if the dissociation is faster.
c. Do you mean saturation (reaching Rmax or reaching steady state / equilibrium) ( www.sprpages.nl/sensorgram-tutorial/a-curve )? High concentrations of analyte can give heterogeneous curves since at high concentration the low 'affinity' sites will also give some response. At low concentrations these sites will give negligible responses. The best curves to analyse have some steady state but you don't have to saturate the ligand (reach Rmax) to get meaningful results.

When I analysed it using the Heterogeneous Ligand model, I found that the sensorgram was fitted almost completely regardless of the concentration - high or low. In my opinion, this phenomenon may be caused by the lack of directionality of the amine coupling method. (I think that I might be wrong)
1. Is it possible to use Heterogeneous Ligand method in this case?

The heterogeneous model introduces two sets of kinetic parameters (k_a, k_d, Rmax) which per definition will make your fitting better. You should careful look at the numbers that are reported. For instance the Rmax values can give you an indication of how much each of the two interactions is adding to the total response. When the second is very low compared to the first you possible have some non-specific interaction in the background distorting your curves and fitting. But when the total response is more or less equally divided between the two interactions you cannot say you have a one to one interaction. The problem with the heterogeneous model is to assign which interaction is the main one and which one is the non-specific one. It is wrong to analyse with a heterogeneous model and only report one set of kinetic values.

2. To improve the curves, look at the experiments section on the SPR-pages. You could try the directional immobilization of the ligand by first immobilizing an antibody that recognises your ligand but is not interfering with the analyte binding. The capture the ligand and perform the experiment. Depending on the regeneration you may have to load the ligand for each cycle (or try single cycle kinetics).

Kind regards
Arnoud