Isothermal titration calorimetry S2004 – METHODS FOR CHARACTERIZATION OF BIOMOLECULAR INTERACTIONS: CLASSICAL VERSUS MODERN Mgr. MONIKA KUBÍČKOVÁ, Ph.D. SLIDES BY: Mgr. JITKA HOLKOVÁ Outline:  Historical background  Theory  Study of interactions  Instrumentation  ITC data analysis History of calorimetry  Calorimetry – Latin calor – heat, Greek μέτρον (-metry) – to measure - thermodynamic technique based on measurement of heat that may be generated (exothermic process) or consumed (endothermic process) by sample  Calorimeter – instrument for measuring the quantity of heat released or absorbed in process of chemical reaction Calorimetric units  A single calorie is the amount of energy required to increase the temperature of 1g of water by 1°C.  A single joule is the amount of energy required to apply a force of 1 Newton over one meter of distance.  1 calorie = 4.184 J 1 Calorie= 1 kcal = 4184 J 1 J = 0.000239 kcal = 0.2390 cal History of calorimetry: „Founding Fathers“  Joseph Black (1728 – 1799) – „founder of the calorimetry“ – first scientist who recognized the distinction between heat and temperature  Antoine Lavoisier (1743-1794)  Pierre- Simon Laplace (1749-1827) First calorimeter - small guiney pig inside Calorimetry  INDIRECT CALORIMETRY – calculates the heat generated by living organism when their metabolic processes yield waste carbon dioxide  DIRECT CALORIMETRY – measures heat generated by living organism by placing the entire organism inside the calorimeter for the measurement • Native molecules in solution (biological relevance) • Very sensitive to accomodate range of affinities Microcalorimetry in cube: Microcalorimetry Broad dynamic range Ease-of-use • Direct measurement of heat change (ITC) • Direct measurement of melting transition temperature to predict thermal stability (DSC) • No labeling or immobilization necessary • Wide range of solvent/buffer conditions Information rich • All binding parameters (affinity, stoichiometry, enthalphy and entropy) in a single ITC experiment 0 1 2 -12 -9 -6 -3 0 Xt/Mt NDH,kcal/moleofinjectant Microcalorimetry  Differential scanning calorimetry DSC ▪ Biomolecular stability in solution ▪ Provides insights into mechanisms of unfolding and refolding ▪ Midpoint (Tm) determination  Isothermal titration calorimetry ITC ▪ Heat is released or absorbed as a result of the redistribution and formation of non- covalent bonds when the interacting molecules go from the free to the bound state. With isothermal titration calorimetry you can…  Get quick KDs  Measure target activity (Stoichiometry, active concentration)  Protein batch activity comparison  Confirm drug binding to target  Use thermodynamics to guide lead optimization  Measure enzyme kinetics How does it work? Reference Calibration Heater Cell Main Heater Sample Calibration Heater DP Sample The DP is a measured power differential between the reference and sample cells to maintain a zero temperature between the cells DT~0 DP = Differential power ∆T = Temperature difference Reference Basics of ITC experiment Integration of heats are used to extract affinity (KD), stoichiometry (N) and binding enthalpy (DH) using appropriate binding model Universal technique based on heat detection µcals-1 Time -> The energetics ∆G = Gibbs free energy ∆H = Enthalpy ∆S = Entropy R = Gas constant = 1.985 cal K-1 mol-1 T = Temperature in Kelvin = 273.15 + t 0C KD = Affinity ΔH, enthalpy is indication of changes in hydrogen and van der Waals bonding -TΔS, entropy is indication of changes in hydrophobic interaction and/or comformational changes The energetics -14 -12 -10 -8 -6 -4 -2 0 kcal/moleofinjectant 0 1 2 3 4  The same affinity and stoichiometry but different enthalpy  This tells us there are different binding mechanisms Ligand A into compound X Ligand B into compound X Molar ratio ITC experiment Standard set-up: Reference cell Sample cell Syringe  “Ligand” in syringe  “Macromolecule” in sample cell ▪ Reverse arrangement possible ▪ Concentration and other parameters necessary to set-up the experiment Performing an ITC experiment Competition titration: Single injection method  This setup consist of one injection of substrate into enzyme with the aim to reach Vmax as fast as possible and observe the signal decay associated with substrate depletion and product formation. Bad Good Steady state (Vmax) Assessment of protein quality by MicroCal™ iTC200 system  100% of Batch #1 protein active based on stoichiometry  23% of Batch #2 protein active based on stoichiometry Presented by L.Gao (Hoffmann-La Roche), poster at SBS 2009 Peptide binding to protein Batch #1 Peptide binding to protein Batch #2 Sample preparation Sample preparation – „c value“ C = 10-100 Great C = 5-10 and 100-500 Good C = 1-5 and 500-1000 Accuracy limited C = < 1 and > 1000 Competittive ITC experiment C = [Protein]/KD Sample preparation ► Stability of interacting molecules in conditions of the ITC experiment ► The cell and syringe buffers must be carefully matched. This is best accomplished by dialyzing both the macromolecule and the ligand in the same buffer. ► If the ligand is too small for dialysis then dialyze the macromolecule and then dissolve the ligand in the dialyze buffer ► Accurately measure protein concentration using A280nm Poor sample preparation = poor data  The data show possible difference of measurement of the sample before and after dialysis  The large peaks were due to differences in the NaCl concentration between buffers Without dialysis With dialysis Instrumentation Malvern Instruments  Differential scanning calorimetry MicroCal iTC200 MicroCal VP-Capillary DSC MicroCal VP-ITC MicroCal™ VP-DSC MicroCal PEAQTM ITC …. MicroCal PEAQ ITC Automated ►Isothermal titration calorimetry TA Instruments  Differential scanning calorimetry ► Isothermal titration calorimetry ITC – theory of data analysis What we are going to discuss? ► Analysis of raw ITC data ► Data fitting • - choice of model • - fitting procedure • - assessment of goodness of fit ► Global fit: fitting of multiple data sets Raw ITC data -3.33 0.00 3.33 6.67 10.00 13.33 16.67 20.00 23.33 26.67 30.00 9.70 9.75 9.80 9.85 9.90 9.95 10.00 10.05 10.10 10.15 Time (min) µcal/sec Make a sanity check of the raw data Buffer mismatch – no dialysis Bad quality raw ITC data -8.33 0.00 8.33 16.67 25.00 33.33 41.67 50.00 58.33 -4 -2 0 2 Time (min) µcal/sec Too low reference power 0.00 33.33 66.67 100.00 133.33 166.67 14 15 16 17 18 19 Time (min) µcal/sec Too short spacing between the injections Origin-based ITC data analysis software First steps  Adjust integrations  Check concentrations  Subtract control heats Choice of model  Models supported by Origin:  4 binding models • one set of sites • two sets of sites • sequential binding • competitive binding  a dimer dissiociation model “One set of sites” model: parameters defined by binding isotherm 0 1 2 -12 -9 -6 -3 0 Xt/Mt NDH,kcal/moleofinjectant Stoichiometry Enthalpy change Affinity Manual fit initialization : your educated guess The simpler the model, the higher the chance automatic initialization will work. If automatic initialization is not satifactory, in NL curve fitting box insert your ”best-guess” values for parameters and click Chi-Sqr button. A simulated curve will appear next to your experimental data curve. Compare and decide. Control of fitting procedure Tolerance-compare chi-sqr values between two successive iterations Delta – controls the way partial derivatives are calculated Parameter constrains – allow to exclude unphysical values of parameters (USE WITH CAUTION) Quality of fit: dependency of parameters ✓Dependency value very close to 1 indicates strong cross-correlation and over- parameterization. Quality of the fit • Chi-sq • Parameter dependence • Errors in the fitted parameters • Agreement between repeated experiments • Biochemical and experimental relevance in the parameters returned by the fit Quality of the fit: fitted parameters N, number of binding sites  “N” is the average number of binding sites per mole of protein in solution, assuming: • that all binding sites are identical and independent • that you have pure protein (and ligand) • that you have given the correct protein and ligand concentrations • that all your protein is correctly folded and active           −         ++−++ D = t nM t X t nKM t nM t X t nKM t nM t X o HV t nM Q 4 2 1111 2 Goodness of the fit: fitted parameters N, number of binding sites  If N≠ 1 • different number of binding sites • inaccurate input values for protein and/or ligand concentration • protein instability issues • compound solubility issues • binding does not fit simple independent model 0.0 0.5 1.0 1.5 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 Molar Ratio kcal/moleofinjectant 0.0 0.5 1.0 1.5 2.0 2.5 -10 -8 -6 -4 -2 0 Molar Ratio kcal/moleofinjectant 0.0 0.5 1.0 1.5 2.0 -14 -12 -10 -8 -6 -4 -2 0 Molar Ratio kcal/moleofinjectant n<1 n>1 n=1 Stoichiometry: Incorrect [Ligand] N 0.8218 K 6.899E4 DH -1.694E4 N 1.279 K 4.434E4 DH -1.089E4 N 1.023 K 5.543E4 DH -1.361E4 Stoichiometry: Incorrect [Protein] 0.0 0.5 1.0 1.5 2.0 -14 -12 -10 -8 -6 -4 -2 0 Molar Ratio kcal/moleofinjectant 0.0 0.5 1.0 1.5 -14 -12 -10 -8 -6 -4 -2 0 Molar Ratio kcal/moleofinjectant 0.0 0.5 1.0 1.5 2.0 2.5 -14 -12 -10 -8 -6 -4 -2 0 Molar Ratio kcal/moleofinjectant n<1 n>1 n=1 N 0.8181 K 5.543E4 DH -1.361E4 N 1.278 K 5.543E4 DH -1.361E4 N 1.023 K 5.543E4 DH -1.361E4 !!!  Error in syringe concentration results in error in DH, K and N  Error in cell concentration results in error in N  Put the sample of which you have most control over in the syringe and evaluate accordingly Reporting results: final figure and parameter box Data: XXXX Model: OneSites Chi^2/DoF = 2089 N 0.991 0.00373 Sites K 4.92E4 1.93E3 M-1 DH -6200 29.00 cal/mol DS 0.675 cal/mol/deg Global fit of multiple datasets collected at different experimental conditions MicroCal PEAQ software-based ITC data analysis MicroCal PEAQ-ITC software MicroCal PEAQ-ITC software MicroCal PEAQ-ITC software MicroCal PEAQ-ITC software – Design experiment Thank you for you attention…☺ ITC and DSC techniques available in CF BIC (Core Facility of Biomolecular Interaction and Crystalization) bic.ceitec.cz CF Head: Prof. Michaela Wimmerová Contact: Monika Kubíčková, monika.kubickova@ceitec.cz C04/339 Data fit: non-linear least-squares minimisation  Fitting procedure evaluates the deviation of the fitted function from the experimental data in terms of chi- squared.  In Origin ITC data analysis module minimization is performed iteratively by Marquardt-Levenberg or simplex algorithms .