There is a methodology that was used first in fluorescence spectroscopy and later in isothermal titration calorimetry (ITC) consisting of performing a binding experiment titrating the potent ligand into a solution of the binding molecule in the presence of a weak ligand. In fluorescence spectroscopy either the potent or the weak ligand have a fluorescent label exhibiting different intensity when bound or free; then, there is a need for fluorescent labels. In the case of ITC there is no need for labels, but the enthalpy of binding for both ligands must be significantly different. The important point is that given a potent ligand, it is easy to get a weak competitive ligand.

When the potent ligand binds it must displace the weak ligand from the binding site. The reduction in the apparent affinity of the potent ligand due to the presence of the weak ligand is a function of the affinity and the concentration of the weak ligand.

Please, check these references:

http://www.ncbi.nlm.nih.gov/pubmed/10625516

http://www.ncbi.nlm.nih.gov/pubmed/17406231

http://www.ncbi.nlm.nih.gov/pubmed/10543978

In addition, this methodology can also be used for determining very low affinity:

http://www.ncbi.nlm.nih.gov/pubmed/9716416

http://www.ncbi.nlm.nih.gov/pubmed/18228446

In addition, you may also use differential scanning calorimetry. When a ligand interact preferentially with the native state of a protein, it will stabilize the protein against thermal denaturation. From the increase in stability it is possible to estimate the binding affinity. Athough this methodology allows determining ultra-high affinities, it is somewhat imprecise, because you must extrapolate in temperature in order to calculate the binding affinity at lower temperature (that is, you need the binding enthalpy and the binding heat capacity for the protein-ligand interaction, as determined by ITC).

Please, check this reference:

http://www.ncbi.nlm.nih.gov/pubmed/2204424

Hi. If you can immobilize your protein on microtiterplates, you could try doing competitive binding assays using radiolabeled ligand at a fixed conc. and varying the conc. of cold competitor. I did this some time ago and it works well to determine Kd in the pM to nM range.

Article Plasminogen Activator Inhibitor-1 Contains a Cryptic High Af...

You could also try surface plasmon resonance (SPR). You can determine Kd in the pM range. The main advantage is the ability to study molecular interactions in a label-free setting. Obviously you need access to a SPR instrument, eg. Biacore from GE Healthcare.

You could try using a quartz crystal microbalance (QCM). I'm not sure if it gives a resolution on the picomolar/femtomolar range, but I think it's quite accurate.

There is a methodology that was used first in fluorescence spectroscopy and later in isothermal titration calorimetry (ITC) consisting of performing a binding experiment titrating the potent ligand into a solution of the binding molecule in the presence of a weak ligand. In fluorescence spectroscopy either the potent or the weak ligand have a fluorescent label exhibiting different intensity when bound or free; then, there is a need for fluorescent labels. In the case of ITC there is no need for labels, but the enthalpy of binding for both ligands must be significantly different. The important point is that given a potent ligand, it is easy to get a weak competitive ligand.

When the potent ligand binds it must displace the weak ligand from the binding site. The reduction in the apparent affinity of the potent ligand due to the presence of the weak ligand is a function of the affinity and the concentration of the weak ligand.

Please, check these references:

http://www.ncbi.nlm.nih.gov/pubmed/10625516

http://www.ncbi.nlm.nih.gov/pubmed/17406231

http://www.ncbi.nlm.nih.gov/pubmed/10543978

In addition, this methodology can also be used for determining very low affinity:

http://www.ncbi.nlm.nih.gov/pubmed/9716416

http://www.ncbi.nlm.nih.gov/pubmed/18228446

In addition, you may also use differential scanning calorimetry. When a ligand interact preferentially with the native state of a protein, it will stabilize the protein against thermal denaturation. From the increase in stability it is possible to estimate the binding affinity. Athough this methodology allows determining ultra-high affinities, it is somewhat imprecise, because you must extrapolate in temperature in order to calculate the binding affinity at lower temperature (that is, you need the binding enthalpy and the binding heat capacity for the protein-ligand interaction, as determined by ITC).

Please, check this reference:

http://www.ncbi.nlm.nih.gov/pubmed/2204424

I agree with Adrian, the best way is ITC and competitive binding.

In my opinion, the best method for subnanomolar ligands is by thermal shift assay (ThermoFluor). There are no limits in affinity and no concentration limitations as you would have in fluorescence quenching. Check our Bioorganic & Medicinal Chemistry 21 (2013) 2093–2106, where 0.1 nM has been measured.

Hello Guus,

The determination of high affinity (sub nanomolar) interactions is clearly difficult and whatever technique you are using will require that you/the operator is really knows how to run the instrument. I have measured a 22pM interaction in SPR/Biacore but it was pure luck that the kinetic binding constants were within the instrument specification. Now I work with a different instrument, LigandTracer, which allows for much longer detection times, longer than 10 hours and the strongest (public) interaction determined so far is 6 pM.

The problem with using concentration at or below the affinity for sub-nanomolar interactions is the time to equilibrium - it can be days or even weeks - so any end-point assay must be incubated properly to provide accurate results. Higher concentrations speed up the interaction, but require a time-resolved interaction analysis method (because the equilibrium value will be the same, saturated value for such concentrations).

Personally I would go for a time-resolved approach at concentrations ranging from approximately KD to 10-30 times KD. You can actually do a manual time resolved binding assay by running a plate based method at different incubation time, or rely on some automated approach.

LigandTracer (www.ligandtracer.com) is I believe one of the few technologes that can to such long-term kinetic interaction measurements, but I am affiliated with the company that makes the machine so please consider that statment as biased.

I really recommend the following papers:

http://www.ncbi.nlm.nih.gov/pubmed/20132208 General paper on ligand binding assays

http://www.ncbi.nlm.nih.gov/pubmed/23166716 The paper with the 6 pM interaction discussed above.

Karl

In my oponion if you could use labbellling protein with either iodine or other radioactive material and the bind it to your microlement what ever it is but with different concentration in the present of cold ligand calculate total ,nonspecific binding and then find specific binding.Using scatchard plot calculate your Ka or Kd from slope of the plot.

Hi Guus,

Karl rightly says that these high affinities are difficult to measure with equilibrium measurements because of the long incubation time needed, but inline measurements giving rate constants as can be done with SPR can do the job. Nevertheless, as in SPR your signal relates to mass, the MW of your molecules and the dynamics of the interaction, will determine whether you are going to be successful. Have a look at our paper on the cytokine TRAIL and its receptor (http://www.ncbi.nlm.nih.gov/pubmed/20852190) and contact me if you think we should give it a try. You know how to reach within the RUG.

Cheers, Robbert Cool

Using scatchard plot it is possible to calculate binding constant

Hi Guus,

you can use Microscale Thermophoresis (MST) to measure picomolare Kd's. MST is a free soultion (immobilization free) method which consumes 1000 times less material than ITC:

http://www.ncbi.nlm.nih.gov/pubmed/23270813

With our new Monolith NT.115 Pico instrument you can easily and directly measure Kd's down to 10pM. http://www.nanotemper-technologies.com/products/instruments/monolith-nt115/article/monolith-nt115pico-instrument-g006/

Best, Philipp

for tight binding ligands, one can use a specific equation that is analogous to the Morrison equation when doing enzyme kinetics. I believe it is the same formula, just that E+ I is changed to Ligand plus receptor. The equation takes into account things when E and I are above or near the KD and tight binding occurs. I know also, people can measure the off rates, and use those to calculate KD or Ki. The competition with a weak ligand also gets a thumbs up. This has been done with fluorescence displacement assays. I guess I did do this a long time ago, but you can look at the paper in analytical biochem, because I don't remember too much.

Sorry. I hope this helps.

The nice thing is, with this equation, you can use your same setup, ie the tryptophan fluorescence as a readout.

I. 2: The Quadratic Velocity Equation for Tight-Binding Substrates

depts.washington.edu/wmatkins/kinetics/quadratic.html

I. 2: The Quadratic Velocity Equation for Tight-Binding Substrates ... approximation, the free ligand approximation and the rapid equilibrium approximation.

Graph pad in Prism has all the info for the equations. Or check on the Internet under tight binding inhibitors and equation as the search words.

Isothermal titration calorimetry is the best.

SPR, photo-luminescence and electrochemical techniques are very useful for direct detection of these interactions. An example of all three in one system is attached.

Article Comparative study of Surface Plasmon Resonance, electrochemi...

MIcroScale Thermophoresis is definitely a very good option. It can measure pM interactions with low sample consumption in solution with free choice of buffers.