Pursuing a reliable calculation of binding affinity.

Eva Muñoz Case of use

The relevance of uncertainties.

 

The reliable determination of binding constants (Ka) and enthalpies (ΔH) from Isothermal Titration Calorimetry depends, among other aspects, on the quality of the thermogram recorded where a high signal-to-noise ratio is pursued to get a trusty isotherm. Low signal-to-noise ratio conducts to less precise integration of the thermogram peaks and, consequently, to a binding isotherm defined by points with larger related uncertainties. Logically, the uncertainty associated to the integral points will significantly affect the determination of the thermodynamic parameters and should not be set aside through the data analysis.

Here´s an illustrative example of two ITC experiments of the interaction between Carbonic Anhydrase II (CA) and 4-carboxybenzenesulfonamide (4-CBS) that yielded thermograms of different quality (Fig 1*).

Fig.1 ITC titrations of 4-CBS into CA. a) Exp1, higher signal-to-noise ratio; b) Exp2, lower signal-to-noise ratio.

Data fitting to a 1:1 binding model was performed for both experiments and for comparison purposes the analyses were conducted twice, taking into account the uncertainties associated to the calculation of the peaks integral (weighted fitting) and not considering them (all the points having the same weight).
The results show a larger discrepancy in the Ka values obtained from Exp2, depending on whether or not the uncertainties are included in the fitting. Besides, the Ka obtained from the weighted fitting of Exp2 is significantly closer to the corresponding Ka value of the weighted fitting of Exp1 (Figure 2).
Fig.2 KA, DH and rM (correction parameter for CA concentration in the cell) values from fitting of Exp1 and Exp2 a) without uncertainties and b) with uncertainties. The bar diagrams show the comparison between KA values of Exp1 and Exp2.

 

The main origin of this difference is the large uncertainty associated to the mid-titration point, which consideration has an effect on the final Ka value (Figure 3).
Fig.3 Fitted isotherm from Exp 2. Redline: theoretical curve calculated when considering uncertainties; black line: theoretical curve calculated with no uncertainties.

 

Being aware of the relevance of weighting data points based on their uncertainty AFFINImeter includes, into the raw data processing step, the automatic determination of uncertainties in the peaks integration and the possibility to perform weighted fitting for a more reliable analysis. Another good reason to work with AFFINImeter!

 

Eva Muñoz

Dr Muñoz has worked as a researcher for over 15 years in different international centers in Spain, Germany and the USA, focusing on the characterization of complex biomolecular interactions by different experimental techniques such as NMR, FT-IR, fluorescence and Surface Plasmon Resonance (SPR). She has a high publication record in this area as well as two patents on biotechnological developments. Eva is working with us in the validation and design of AFFINImeter as well as in the implementation of functionalities for the analysis of SPR measurements.