How are you measuring your bound ligand? Turbidimetry, ELISA, ITC, spr or something else?

What is your ligand or receptor? are either easily quantified?

for equilibrium times it depends on what method you are using.

for Turbidimetry you can simply centrifuge the complexes and measure whatever is left in solution. This should correspond to [free]

Similar for elisa try to measure what you discard.

If you are using ITC or SPR the signal should be directly proportional to ligand bound.

If you are outside the range of most interaction methods I recommend DSC

It will depend on how tight the binding is and how low the total concentration of enzyme/receptor is in the experiment. It will depend, therefore, on the experimental technique. In ITC you can determine Kds in the nanomolar range, while in fluorescence you may go to the subnanomolar range.

It is not always possible to lower enough the concentration of enzyme/receptor. Displacement experiments with ligands with moderate affinity will allow determining very high affinities (very small Kds) or very low affinities (very large Kds).

Which technique are you using?

Are you measuring Kd of an inhibitor through enzymatic inhibition curves?

Why don't you use ligand depletion in your analysis? That is always correct, no matter how tight/weak is the binding.

Hi Adrian, thanks for your response.

I am using FP to measure Kd of my protein towards a RNA sequence (fixed RNA conc., titrating the protein). My RNA concentration is 35nM, Kds were around 18-50nM across several conditions. I want to account for ligand depletion straight away, but GraphPad Prism only allows me to use this approach with a radioligand (when trying to fit this equation it requests the specific activity of the radioligand, which I dont have obviously). Do you know of a better way to proceed with the ligand depletion analysis?

My other option is decreasing the RNA concentration 5-10X below Kd and repeat the experiment, but I have already decreased it to 5nM and still cannot assume free ligand equal to total ligand in my system. I am planning on reducing it further, perhaps to picomolar concentrations (still need to check the sensitivity of my plate reader to see if I can do this), and that assuming I don't get non-specific binding to the plate...

Thanks!

Ramon

You can try to apply the quadratic tight-binding equation to your titration. This can extend the range of Kd that can be measured by the FP experiment, but not indefinitely, since the accuracy of the calculation becomes limited by the scatter in the data. Maximize data quality by averaging a large number of flashes/measurement and multiple replicate samples.

Lowering the fluorescent ligand concentration as much as possible is a good strategy to measure lower Kds. This is obviously limited by the ability detect the fluorescence. With a good plate reader, optimized geometry and filters, and a high-quantum yield dye like TAMRA, you should be able to measure 1 nM, perhaps even less.

To avoid loss of the ligand to surface adsorption, include 0.01% Triton X-100 detergent, or some similar detergent, as a blocking agent, or use non-binding surface plates (expensive).

Thanks Adam, that is the same strategy I had in my mind, the only thing left would be to apply the tight binding equation, for which Prism requires the presence of a radioligand. Do you know of a way to get around this?

So I have defined a similar quadratic equation for tight binding ligands on Prism and got a very similar Kds compared to the standard saturation curve. The model that I used is: y= offset + a * ((RNA + x + Kd)-((RNA+x+Kd)^2-4*RNA*x)^0.5)/(2*RNA), where offset is Y-intercept in the absence of protein and a is defined as signal change. Is this correct?

I was confused by the fact that Prism mentions that both axis should be in the same units and that this cannot be applied to non-radiolabelled ligands. The fact that both models give me similar values, does this mean that the first one (standard binding curve) already takes ligand depletion into consideration? Or am I missing something in the tight binding equation?

Which are the values you are getting from both fittings?

I am not familiar with GraphPad Prism, so I do not know if the built-in fitting function would account for ligand depletion.

Highly similar, in the same decimal. Also very similar fitting, like the standard saturation curve had already accounted for such tight binding, or there's something missing in the tight binding equation. Which software do you use to analyse tight binding ligands?

Origin from OriginLab

The form of your tight binding equation looks right, I think. I don't know about what Prism is doing. If you are getting the same result with the tight binding equation as with the simple binding isotherm equation, then you are not working under tight binding conditions. You can tell from the fit of the data to the simple binding isotherm equation. If it is a good fit with no systematic error, then it is not a tight binding situation. Perhaps your RNA or protein concentration is inaccurate.

I have attached a snapshot of my data. This fitting has been performed using the standard hyperbolic function (Kd of 10nM, +/- 1nM for green curve, left graph), and also using the quadratic equation (Kd of 6.6nM, +/- 2nM, right graph). Fitting is very similar, perhaps the quadratic equation curve bends slightly more before reaching plateu, whereas the standard curve is basically a straight line up at low concentrations and then saturates immediately, forming a horizontal line.... I do think the concentrations are accurate, protein stocks were measured using extinction coefficient and our RNA was acquired from an external provider. Please share your thoughts!

The Kds are quite close to each other using the two fitting methods. Considering that there are relatively few data points in the rising part of the curves, I would say the Kds are indistinguishable. I suggest repeating the titration with more coverage of the rising part of the curve.

Hi,

I am not so familiar to your software. In my study, Kd values were calculated using Igor Pro (WaveMetrics, Lake Oswego, OR, USA) and I get Kd about 100nM, then it is confirmed by ITC with Kd 65, 80, 75nM (3times), so if you do not mind about the inconvenience please try it.

In case, if you think to change the experiment setup, there is way to find the reliable Kd with competitive titration experiment. First, you could titrate with a low affinity ligand and then competitive with high affinity ligand. This way, it would be possible to find the kd of tight binding ligands.

Shakir Tuleab, are you copy-pasting other people's answers as your own aswers through different posts?

VERY SORRY this mistake DEAR

i am new member in RG

MANY THANKS

I REMOVE THE COMMENT