Kinetic Analysis of a Protein-protein Complex to Determine its Dissociation Constant (K_D) and the Effective Concentration (EC₅₀) of an Interplaying Effector Molecule Using Bio-layer Interferometry

Materials and Reagents

1. Microplate, 96-well plate, PS, F-Bottom (Chimney well) Black, Non-Binding (Greiner Bio-One, catalog number: 655900)

2. HEPES (Thermo Scientific, Fischer Bioreagents, catalog number: BP-310)

3. KOH (Merck, Millipore, catalog number: 1.05033)

4. Nonident-P40 (Thermo Fischer Scientific, Fluka, catalog number: 74358)

5. KCl (Carl Roth, catalog number: 6781.1)

6. MgCl₂·6H₂O (Merck, Millipore, catalog number: 7791-18-6)

7. ATP (store at -20°C) (Carl Roth, catalog number: K054.3)

8. α-ketoglutaric acid disodium salt hydrate (store at 4°C) (Thermo Fischer Scientific, Fluka, catalog number: K3752) – another name for 2‐oxoglutarate (2-OG)

9. Glycine (Merck, Sigma-Aldrich, catalog number: 33226)

10. NiCl₂·6H₂O (Carl Roth, catalog number: 4489.1)

11. Kinetics buffer (see Recipes)

Equipment

1. Ni-NTA (NTA) Dip and Read^TM Biosensors (tray) (Sartorius, FortéBio^®, catalog number: 18-5103)

2. FortéBio^® Octet K2 System (Sartorius)

Software

1. Data Acquisition 11.0.0.64 (FortéBio^®, https://www.sartorius.com/en/products/protein-analysis/octet-systems-software)

2. Data Analysis HT 11.0.0.50 (FortéBio^®, https://www.sartorius.com/en/products/protein-analysis/octet-systems-software)

3. Prism, Version 6.01 or higher (GraphPad Software Inc.)

Procedure

For this protocol, we used purified 6×His-tagged P_II protein as the bait protein (ligand) and a Strep-tagged PirC protein (analyte), as described previously (Orthwein et al., 2021). It is recommended to use 5-25 µg/ml of ligand per well of a 96-well plate when loading the proteins onto the Ni-NTA biosensors. A purified ligand concentration of 0.5 mg/ml in 500 µl at a minimum is needed to perform 50 measurements if 25 µg/ml is used in the assay. This provides an acceptable dilution of the kinetics buffer with a sufficient amount of protein. A minimum analyte molar concentration that is 5-times higher is needed.

Note: It is possible to store your proteins in a buffer containing glycerol.

An overview of the method: The ligand is loaded specifically onto the Ni-NTA sensor tips via the 6× His-tag tail (loading) followed by a washing step in kinetics buffer. The loaded tips are then dipped into the analyte solution to measure the complex formation (association), indicated by an increasing signal. Afterward, the sensors with the complex are dipped back into the kinetic buffer (decreased signal dissociation). Before and after the procedure, the sensors are washed and regenerated with an acidic 10 mM glycine solution (pH 1.7) and a 10 mM NiCl₂ solution. The software calculates the response in equilibrium (R_eq) using the association and dissociation signal. The values of R_eq for each analyte concentration are plotted in a graph (R_eq against the analyte concentration), which is used for the determination of the kinetics parameters. For this purpose, the measurement is performed with different concentrations of the analyte to determine the K_D and with different concentrations of the effector to determine the EC₅₀. The R_eq values are then plotted against the related analyte concentrations. For EC₅₀ determination, the values are transformed and the relative response is plotted against the log₁₀ of the analyte. Afterward, kinetic analysis with a statistical software (e.g., GraphPad Prism) is performed for the calculation of the kinetic parameters.

1. Bio-layer interferometry kinetic binding assay

   The assay was performed using the FortéBio^® Octet K2 System (Sartorius).

   Note: Make sure that other tags used for the analyte do not interact with poly histidine (possibly metalloproteins) or bind non-specifically to Ni-NTA. It is possible to use a His-tagged protein as an analyte after the cleavage removal of the tag; however, it is not recommended. Test this via trial loading experiments of the analyte onto the biosensor. If there is a signal, there is unspecific binding of the analyte. If there is non-specific binding of the analyte, it is recommended to load the remaining loading positions with a His-tagged protein that does not interact with the analyte. For triplicates, measure three times with newly prepared 96-well plates.

   1. Switch on the machine and prewarm the sample holder for a minimum of 30 min. Adjust the working temperature of choice.Note: We used 30 °C because of the growth temperature of cyanobacteria.

   2. Prepare the kinetics buffer, the sensor washing solution, i.e., 10 mM glycine pH 1.7 (adjust pH with HCl), and the regeneration solution, i.e., 10 mM NiCl₂.

      Note: The optimal conditions for interaction depend on the proteins under study. A preliminary buffer screening test (e.g., different pH and salts) may be required to determine the optimal conditions for protein-protein interaction. We used the kinetics buffer described in the recipes section.

   3. Dilute the ligand in the kinetics buffer for the Ni-NTA sensor tip loading.

      Note: The optimal concentration of the ligand should be tested before the start of the kinetic protein-protein interaction assay. For this, perform a trial loading experiment with different ligand concentrations. The manufacturer protocol recommends a ligand concentration between 5-25 µg/ml. We used concentrations that did not exceed a binding signal of Δλ = 5 nm.

   4. Prepare a dilution series of 7 different concentrations of the analyte.

      Note: We recommend starting the dilution with ~10x the expected K_D, as described in the manufacturer protocol.

   5. Fill the vials of the 96-well plate with 200 µl of solutions as shown in Figure 1.

      
      

      
      Figure 1. Pipetting scheme of the 96-well plate for a kinetic assay. Column 4 contains the dilution series from the highest (A) to the lowest concentration (G) of the analyte. Vial 4H contains the control without analyte. Adapted from the manufacturer protocol - technical note Ni-NTA biosensor kinetic assays (FortéBio, 2019).

      
      

   6. Place your plate in the machine for prewarming.

   7. Define the experiment in Data Acquisition 11.0.0.64 (Sartorius, FortéBio^®) by following the top menu and start the measurement.

      1. Plate definition: Define the plate position and enter additional information in the table, as shown in Figure 2.

         
         

         
         Figure 2. Plate definition and sample information table. Definitions: (1A-H) Buffer, (2A-H) Loading, (3A-H) Buffer, (4A-H) Sample, (10A-H) Wash, (11A H) Neutralization, and (12A-H) Regeneration. Concentrations are example values.

         Note: Use the concentrations and MW of your analyte.

      
      

      2. Assay Definition. Define the assay as shown in Figure 3.

         Note: The time of the association and dissociation steps depends on the proteins under study. For kinetic investigation, we suggest determining the optimal association and dissociation times by performing preliminary assays using different association/dissociation times ranging between 2-10 min and using saturating concentrations of the analyte (~ 10× the expected K_D). Define the assay steps method for row A and B. Mark the assay step list and click “Replicate” to clone the steps for the other rows (> Replicate> Replication type: add a new assay > Offset step > vertically > OK).

         
         

         
         Figure 3. Assay definition for the kinetic binding assay

      
      

      3. Sensor Assignment. Assign the Sensors in the correct line.

      4. Review Experiment. Review your experiment.

      5. Run Experiment. Define the localization, name the results and start the experiment.

         Note: If you wet your biosensors immediately before, use click on “delayed experiment starts: 600 s”. Do this also if you did not prewarm your plates previous to the experiment.




2. Determination of the EC₅₀ of a small effector molecule (effector) that interferes in the protein-protein complex

   1. Prepare a dilution series of the investigated effector.

      Note: In the example, we used the 2-OG metabolite as it inhibits the P_II-PirC complex. Prepare a dilution series of the effector without the analyte for the pre-washing step (to be used to fill the wells in column 3) and a dilution series that contains the analyte (constant amount of analyte in each dilution of the effector) for analyte binding (to be used to fill the wells in column 4). The analyte should be used in amounts enabling saturation of the ligand as determined from the previous kinetic data.

   2. Fill the vials of the 96-well plate with 200 µl of solutions as shown in Figure 4.

      
      

      
      Figure 4. Pipetting scheme of the 96-well plate for a determination of the EC₅₀ for an effector that affects the stability of the protein complex. Columns 3 and 4 contain dilution series of the effector from highest (A) to lowest concentrations (G), without and with constant amount of analyte, respectively. Vial 4H contains the control without the effector [adapted from the manufacturer protocol – technical note Ni-NTA biosensor kinetic assays (FortéBio, 2019)].

   
   

   3. Repeat Steps A6 and A7.

Data analysis

1. Kinetic binding assay

   1. Start the Data Analysis HT 11.0.0.50 (Sartorius, FortéBio^®) software and open the results of the measurement.

   2. Go to preprocessed Data > Data Correction and adjust the settings as follows:

      1. Align Axis: Align Data to: Average of Baseline.

      2. Inter-step Correction: Align Data to: Association Step.

      3. Activate Savitky-Golay Filtering.

   3. Go to kinetics analysis and adjust the settings as follows:

      1. Step to Analyze: Association and Dissociation.

      2. Binding Model: depending on protein.

         The 1:1 model is recommended for protein complexes where the analyte binds to one specific binding site of the ligand. If the analyte binds at two independent ligand sites, use the 2:1 (heterogeneous ligand) model. The Mass Transport model is recommended for bindings that are limited by diffusion. The bivalent model fits the binding of one analyte that interacts with two sites of the immobilized ligand. Detailed information is provided in the manufacturer’s user manual (Page 114, Pall FortéBio LLC, Octet System Data Analysis User Guide).

      3. Fitting.

         Type: Global (Group)

         Group by: Sensor Type

         R_max Values: Linked

      4. Steady State Analysis: R equilibrium.

      The software calculates the response values in the equilibrium and creates a steady-state graph of your measurement, as shown in Figure 5.

      
      

      
      Figure 5. Kinetic Analysis page in the data analysis software. Graphs (fit graph) below the function tab show how the preprocessed data is fitted with the settings in step 3. The list below shows the calculated R_eq values and additional information about the sensors. On the right edge, the Steady-State calculation of one measurement row is shown. The calculated K_D is also reported.

      Note: You can use this calculation to determine the K_D, but the combination of several measurements is not possible.

   
   

   4. Export the data by clicking on “Excel Report.”

   5. Open Prism and create a new Table and Graph Project.

      Enter and import data:

      1. Click on XY.

      2. Enter three replicate values in side-by-side sub columns.

         Note: It is also possible to use a different software for the calculation and fitting of the binding kinetics. Make sure that the software contains the correct binding kinetic equations.

   6. Transfer the R_eq values into the GraphPad table as Y-values.

   7. Add the interactor concentrations as X Values.

   8. Click on the automatically created graph and perform the “Analyze” function.

      1. Nonlinear regression (curve fit).

      2. Binding – Saturation.

      3. One site – Specific binding.

         The software fits the graph and creates a results file with the calculated B_max (= R_max) and K_D, as shown in Figure 6.

      
      

      
      Figure 6. Representative Data of Analysis for kinetic analysis of protein interactions for the P_II-PirC example. (A) Fit of the binding (R_eq) against the concentration of the analyte (Strep-PirC). (B) Prism document (results file) of calculated kinetics in the results folder. The window shows the maximal binding (B_max), the calculated K_D, and the standard error of both. Confidence intervals and the quality of the fit are also calculated.




2. EC₅₀ determination

   1. Repeat the steps of Data analysis A1-A4.

   2. Add the effector concentrations as X Values.

   3. Export the data by clicking on “Excel Report” and transfer the R_eq values into a GraphPad table as Y-values.

      Note: For triplicates, measure three times with newly prepared 96-well plates.

   4. Perform the “Analyze” function “Normalize” with y = 0.0 as 0% and the largest value in each data set as 100%.

   5. Perform the “Analyze” function “Transform” and transform X values using X = Log[X]. In the original data table, change the concentration from 0 to 1.

   6. Perform the “Analyze” function “Nonlinear Regression (curve fit)”

      1. Dose-response – inhibition (negative effector; IC₅₀)/Dose-response – Stimulation (positive effector).

      2. Log(inhibitor) vs. normalized response/Log(agonist) vs. normalized response.

      The software fits the graph and creates a results file with the calculated EC_50. The example IC₅₀ is shown in Figure 7.

      
      

      
      Figure 7. Representative Data Analysis for determination of the IC₅₀ of 2-OG in the P_II-PirC interaction. (A) Fitted data of a triplicate measurement of P_II-PirC binding for determination of the IC₅₀ of 2-OG. X-Axis logarithm of 2-OG concentration (dilution series of 2 mM, 1 mM, 0.5 mM, 0.25 mM, 0.125 mM, 0.0625 mM, and 0.03125 mM), and Y-Axis: relative response in equilibrium. (B) Prism document (results file) of calculated IC₅₀ in the results folder. The window shows the logarithmic IC₅₀, the calculated IC₅₀ and the logarithmic standard error of both. Confidence intervals and the quality of the fit are also calculated.

Recipes

1. Kinetics buffer

   20 mM HEPES pH 8

   150 mM KCl

   5 mM MgCl₂

   1 mM ATP

   0.005% Nonident P40

   KOH for pH adjustment

Acknowledgments

This work was funded by grants from DFG to K.F. (Fo195/9‐2, Fo195/21-1) and from Cluster of Excellence ’Controlling Microbes to Fight Infections’ (EXC 2124 – 1-06.010_0) to K.A.S. Furthermore, we would like to acknowledge Janette Alford for her assistance in writing the first draft and Libera Lo Presti for critical scientific and linguistic editing of the manuscript, and the infrastructural support by the Cluster of Excellence ’Controlling Microbes to Fight Infections’ (EXC 2124) of the German research foundation (DFG).

  This protocol was based on our original research article published in the Journal of Biological Chemistry: Santos, A. R. S., Gerhardt, E. C. M., Parize, E., Pedrosa, F. O., Steffens, M. B. R., Chubatsu, L. S., Souza, E. M., Passaglia, L. M. P., Sant'Anna, F. H., de Souza, G. A., Huergo, L. F. and Forchhammer, K. (2020). NAD⁺ biosynthesis in bacteria is controlled by global carbon/nitrogen levels via PII signaling. J Biol Chem 295(18): 6165-6176.

Competing interests

The authors declare no competing interest.

Ethics

K.A.S. conceived and initiated the project. T.O. performed the experiments, wrote the first draft and edited the manuscript. L.F.H., K.F. and K.A.S. commented and edited the manuscript. All authors have read and approved the final version of the manuscript.

References

1. Bolay, P., Rozbeh, R., Muro-Pastor, M. I., Timm, S., Hagemann, M., Florencio, F. J., Forchhammer, K. and Klahn, S. (2021). The Novel PII-Interacting Protein PirA Controls Flux into the Cyanobacterial Ornithine-Ammonia Cycle. mBio 12(2): e00229-21.
2. Forchhammer, K. and Selim, K. A. (2020). Carbon/nitrogen homeostasis control in cyanobacteria. FEMS Microbiol Rev 44(1): 33-53.
3. FortéBio. (2005-2006). Octet System User's Guide.
4. FortéBio. (2019). Ni-NTA Biosensor Kinetic Assays.
5. Orthwein, T., Scholl, J., Spät, P., Lucius, S., Koch, M., Macek, B., Hagemann, M. and Forchhammer, K. (2021). The novel P_II-interactor PirC identifies phosphoglycerate mutase as key control point of carbon storage metabolism in cyanobacteria. Proc Natl Acad Sci U S A 118(6): e2019988118.
6. Pall FortéBio LLC. (2011). Octet^® System Data Analysis User Guide. Release 7.1.
7. Santos, A. R. S., Gerhardt, E. C. M., Parize, E., Pedrosa, F. O., Steffens, M. B. R., Chubatsu, L. S., Souza, E. M., Passaglia, L. M. P., Sant'Anna, F. H., de Souza, G. A., Huergo, L. F. and Forchhammer, K. (2020). NAD⁺ biosynthesis in bacteria is controlled by global carbon/nitrogen levels via P_II signaling. J Biol Chem 295(18): 6165-6176.
8. Scholl, J., Dengler, L., Bader, L. and Forchhammer, K. (2020). Phosphoenolpyruvate carboxylase from the cyanobacterium Synechocystis sp. PCC 6803 is under global metabolic control by P_II signaling. Mol Microbiol 114(2): 292-307.
9. Selim, K. A., Ermilova, E. and Forchhammer, K. (2020). From cyanobacteria to Archaeplastida: new evolutionary insights into P_II signalling in the plant kingdom. New Phytol 227(3): 722-731.
10. Shah, N. B. and Duncan, T. M. (2014). Bio-layer interferometry for measuring kinetics of protein-protein interactions and allosteric ligand effects. J Vis Exp(84): e51383.
11. Shang, J., Wan, Y., Luo, C., Ye, G., Geng, Q., Auerbach, A. and Li, F. (2020). Cell entry mechanisms of SARS-CoV-2. Proc Natl Acad Sci U S A 117(21): 11727-11734.