There are various methods that have been developed over these years to study protein-protein interactions (PPIs). PPI plays a big role in the cell-signalling cascade; for instance, dephosphorylation of glycogen synthase by protein phosphatase-1 results in glycogen synthesis. To know whether a specific protein binds to its partner, for example, whether TFIIH interacts with TFIIE or TFIIF to complete the pre-initiation complex in transcription, different methods such as co-immunoprecipitation (co-IP), glutathione-S-transferase (GST) pull down assays, yeast-two-hybrid (Y2H) assays, isothermal titration calorimetry (ITC), surface plasmon resonance (SPR), nuclear magnetic resonance (NMR) spectroscopy and etc. can be use to validate PPIs. Yet, doing one experiment using one method is not enough to validate the PPI between two or more proteins. Factors such as overexpression of proteins and manipulation of the agents used in the experiment could result in a bias data. Thus, the results should be unbiased by incorporating different methods in the experiment to validate the PPI. In this essay, the different methods will be described and the factors that cause the different methods giving rise to different results will be discussed.Order now
Co-IP is the most commonly used methods to verify protein-protein interactions (Berggård et al., 2007). Antibodies that are specific to the bait complexes are used to capture the bait complexes in a cell lysate shown in Fig. 1. The antibody is immobilized on Protein A/G, which is covalently bound to the agarose beads. Since the antibody is specific to only the bait complex, the antibody will not bind to other proteins found in the cell lysate, and hence, these proteins will be wash off. The antibody-bait complex can be eluted after washing. The bound proteins in the bait complex can be identified by using mass spectrometry (MS) or by immunoblotting (Berggård et al., 2007). One major disadvantage of co-IP is the tendency of washing off interacting proteins together with unbound proteins, affecting the experiment. One recent study have shown to overcome this by introducing two-step chemical crosslinking by co-IP coupled with tandem MS to identify PPIs, and also to allow better study on weakly bound PPI (Huang & Kim, 2013).
GST pull-down assay is an in vitro method that is widely used to purify specific protein in a cell lysate, and the recombinant protein is often overexpressed in the cell to aid in the purification. GST fusion proteins are commonly expressed from E. coli and being purified through immobilized glutathione-coated beads matrix (Panchenko & Przytycka, 2008). Only proteins that contains GST-tagged will be able to bind to the matrix and unbound proteins will be washed off. Once GST fusion protein bound to the matrix, the prey protein solution can be added to matrix and only those proteins that interact with GST fusion protein will bind to the GST fusion protein on the matrix and unbound proteins will be washed off as shown in Fig. 2 (Panchenko & Przytycka, 2008).
The yeast-two-hybrid (Y2H) system is based on the idea that transcription factor have two distinct functional domains that can be spliced into two, the DNA binding domain (BD) that binds to the upstream activating sequences (UAS) and an activating domain (AD) which activates transcription (Osman, 2004). Without the presence of either domain, transcription of the gene cannot take place. However, if both domains are placed close to each other, it is enough to restore a functional transcription factor and thus, activating transcription of a reporter gene (Osman, 2004). In Y2H system, there are two plasmids being constructed. The first plasmid contains the bait protein genomic sequence fused to BD sequence and the second plasmid contains the prey protein sequence being fused to the AD sequence (Berggård et al., 2007). Both plasmids are inserted to the yeast cell, where the bait-BD protein and prey-AD protein are targeted to the nucleus. The Y2H system is described in the legend of Fig. 3. One major advantage of using Y2H system is the ability to detect PPIs in vivo as compared to co-IP and GST-tags. However, Y2H system often results in the data having large number of false positives, causing the protein interactions that are being identified to be unreliable or is questionable (Deane et al., 2002). Hence, it is important to validate the protein interactions identified in Y2H using other validation methods.
The thermodynamics of protein-protein interactions can be measured using isothermal titration calorimetry. By keeping the sample and reference cell at constant temperature, any changes in heat can be detected by the feedback heater as shown in Fig. 4 (Cooper, 2011). When two proteins interact, there are changes in the thermodynamic potentials, allowing the enthalpic and entropic contributions of the binding process to be measured (Campoy et al., 2004). The binding reaction produces heat pulses and is recorded as shown in Fig. 5. As more binding sites of the protein are occupied, the heat pulses get smaller (Cooper, 2011) and resulted in dilution peaks (Campoy et al., 2004). The heat after each injection can be obtained by calculating the area under each peak (Leavitt & Freire, 2001). Thus, ITC can be use to validate protein-protein interactions. Many studies uses ITC to validate PPIs, for instance, Zhou et al. (2005) shows that xanthine oxidase (XO) binds to copper, zinc superoxide dismutase (Cu, Zn-SOD) with high affinity due to hydrogen binding using ITC and fluorescence spectroscopy. In the experiment, a large favourable enthalpy decrease coupled with a large unfavourable entropy reduction drives the binding of XO to Cu, Zn-SOD (Zhou et al., 2005).
Surface plasmon resonance (SPR) sensing is been commonly use to evaluate protein-ligand or PPIs (Daghestani & Day, 2010). SPR relies on the phenomenon attributed to the thin metal film, usually gold, and any changes on the biological layer (dielectric) surface will results in a change of the angle of reflected light where this signal will be recorded (Berggård et al., 2007) as illustrated in Fig. 6. The response detected is proportional to the number of analytes bind to the immobilized ligand. Using this technique, the interaction partner can be immobilized on the dielectric surface and the other interaction partner can be injected into the flow cell (Berggård et al., 2007) as shown in Fig. 6B. Other than association of the analyte to the immobilized ligand, dissociation of the analyte from the ligand can also be monitored and measured. In Fig. 6D, the binding of the analyte to the immobilized ligand changes the angle of reflective light, and hence, increases the resonance angle. By changing the analyte flow to buffer flow (Berggård et al., 2007), the analyte begins to dissociate from the immobilized ligand, causing a drop in the resonance signal illustrated in Fig. 6D. The dissociation of the analyte from the immobilized causes the angle of reflected light to return back to its original angle once all the analyte has been washed off (Daghestani & Day, 2010). Recently, the information that is obtained from SPR can be coupled with MS information, giving both quantitative and qualitative information on the protein-ligand or PPIs (Daghestani & Day, 2010).
NMR spectroscopy provides important information on the molecular structure of the protein. As a drug discovery tool, it also provides the information on interactions between the drugs and their targets (Takeuchi & Wagner, 2006). As the ligand binds to a protein, it changes the protein’s structure and dynamics, which can be detected using NMR spectroscopy. By introducing isotope labeling (15N, 13C or 2H) to the protein, properties of NMR such as chemical shifts, relaxation rates (Zhang et al., 2006) can be obtained by analyzing the change in protein structure and dynamics upon binding to a ligand (Takeuchi & Wagner, 2006). Living prokaryotic cells, such as Escherichia coli, can be used to analyse the changes in protein structure and dynamics by NMR spectroscopy in vivo (Selenko & Wagner, 2006). Overexpression methods are usually used to express large amounts of isotope labeled proteins. Recombinant proteins are produced in a growth media that is rich in isotope-substituted precursors (Selenko & Wagner, 2006). Changing to an unlabeled growth media, the interacting protein can be expressed and binds to the labeled protein in the cytoplasm of the living cell (Burz et al., 2006). Due to the binding of the ligand, there is a difference in the chemical-shift values of the residues involved in the binding, and can be measured by NMR spectroscopy (Selenko & Wagner, 2006). The technique using living cell coupled with NMR spectroscopy is also known as in-cell NMR spectroscopy, STINT-NMR. One major advantage using STINT-NMR is the ability to study PPIs under physiological conditions without the need to purify the protein, which can be time consuming (Zhang et al., 2006). However, due to the limitations of NMR spectroscopy, the technique cannot analyse large protein complexes that are ≥35kDa (Yu, 1999).
Berggård T., Linse S. and James P. (2007). Methods for the detection and analysis of protein-protein interactions. Proteomics. 7 (16), pp. 2833-2842
Burz D.S., Dutta K., Cowburn D. and Skekhtman, A. (2006). In-cell NMR for protein-protein interactions (STINT-NMR). Nature Protocols. 1 (1), pp. 91-93
Campoy A.V., Leavitt S.A. and Freire E. (2004). Characterization of Protein-Protein Interactions by Isothermal Titration Calorimetry. In: Fu H. Protein-Protein Interactions: Methods and Applications Volume 261. New Jersey: Humana Press. pp. 35-54.
Chen X., Chang J., Deng Q., Xu J., Nguyen T.A., Martens L.H., Cenik B., Taylor G., Hudson K.F., Chung J., Yu K., Yu P., Herz J. and Farese R.V. (2013). Progranulin Does Not Bind Tumor Necrosis Factor (TNF) Receptors and Is Not a Direct Regulator of TNF-Dependent Signaling or Bioactivity in Immune or Neuronal Cells. The Journal of Neuroscience. 33 (21), pp. 9202–9213.
Cooper A. (2011). Thermodynamics and interactions. In: Cooper A. Biophysical Chemistry. 2nd ed. Cambridge: The Royal Society of Chemistry. pp. 199-121.
Daghestani H.N. and Day B.W. (2010). Theory and Applications of Surface Plasmon Resonance, Resonant Mirror, Resonant Waveguide Grating, and Dual Polarization Interferometry Biosensors. Sensors. 10 (11), pp. 9630-9646.
Deane C.M., Salwiński Ł., Xenarios I. and Eisenberg D. (2002). Protein interactions: two methods for assessment of the reliability of high throughput observations. Molecular & cellular proteomics. 1 (5), pp. 349-356.
Gutierrez-Uzquiza A., Colon-Gonzalez F., Leonard T.A., Canagarajah B.J., Wang H.B., Mayer B.J. and Hurley J.H. (2013). Coordinated activation of the Rac-GAP β2-chimaerin by an atypical proline-rich domain and diacylglycerol. Nature Communications. doi:10.1038/ncomms2834
Huang BX, Kim H-Y (2013) Effective Identification of Akt Interacting Proteins by Two-Step Chemical Crosslinking, Co-Immunoprecipitation and Mass Spectrometry. PLoS ONE 8(4): e61430. doi:10.1371/journal.pone.0061430 Last accessed: 4 April 2014 6.12pm
Leavitt S. and Freire E.. (2001). Direct measurement of protein binding energetics by isothermal titration calorimetry. Current Opinion in Structural Biology. 11 (5), pp. 560-566.
Osman A. (2004). Yeast Two-Hybrid Assay for Studying Protein-Protein Interactions. In: Melville S.E. Parasite Genomics Protocols Volume 270. New Jersey: Humana Press. p. 403.
Panchenko A.R. and Przytycka T.M. (2008). Protein-protein interactions and Networks: Identification, Computer Analysis, and Prediction (Computational Biology). Springer. p. 10.
Selenko P. and Wagner G. (2006). NMR mapping of protein interactions in living cells. Nature Methods. 3 (2), pp. 80-81.
Takeuchi K. and Wagner G. (2006). NMR studies of protein interactions. Current Opinion in Structural Biology. 16 (1), pp. 109-117.
Tang W., Lu Y., Tian Q.Y., Zhang Y, Guo F.J., Liu G.Y., Syed N.M., Lai Y., Lin E.A., Kong L., Su J., Yin F., Ding A.H., Zanin-Zhorov A., Dustin M.L., Tao J., Craft J., Yin Z., Feng J.Q., Abramson S.B., Yu X.P. and Liu C.J. (2011). The Growth Factor Progranulin Binds to TNF Receptors and Is Therapeutic Against Inflammatory Arthritis in Mice. Science. 332 (6028), pp. 478-484.
Wissmueller S., Font J., Liew C.W., Cram E., Schroeder T., Turner J., Crossley M., Mackay J.P. and Matthews J.M. (2011). Protein-protein interactions: analysis of a false positive GST pulldown result. Proteins. 79 (8), pp. 2365-2371.
Yu H. (1999). Extending the size limit of protein nuclear magnetic resonance. Proceedings of the National Academy of Sciences. 96 (2), pp. 332-334.
Zhang X., Tang H., Ye C. and Liu M. (2006). Structure-based drug design: NMR-based approach for ligand-protein interactions. Drug Discovery Today: Technologies. 3 (3), pp. 241-245.
Zhou YL., Liao JM., Du F. and Liang Yi. (2005). Thermodynamics of the interaction of xanthine oxidase with superoxide dismutase studied by isothermal titration calorimetry and fluorescence spectroscopy. Thermochimica Acta. 426 (1-2), pp. 173-178.