In all living systems, biomolecular binding interactions are crucial in many biochemical processes such as cell proliferation, maintaining homeostasis or information processing. In these processes, proteins are the main players responsible for structural, mechanical or cell signaling functions. They fulfill their biological functions mainly through direct physical interaction with other molecules – proteins and peptides, nucleic acids, membranes, substrates, and small molecule ligands such as oxygen, metal ions or more complex small organic molecules.

Molecular recognition is defined as the process in which biological macromolecules interact with each other or with various small molecules to form a specific complex. This process is characterized mainly by two terms: specificity and affinity. Specificity relates to the ability to distinguish the highly specific binding partner from less specific partners, while affinity determines the strength of the interaction. The nature (enthalpic or entropic) of the forces affecting the interactions between the molecules, as well as their magnitude, are of great importance in the discovery and design of ligands in drug development, and facilitates further optimization of the compound structures.

Drug discovery campaigns comprise several phases, e.g. primary screening, structure optimization or pharmacokinetics/pharmacodynamics analysis, leading to selection of the most promising molecules (so called “hits” or “leads”). Revolutionary high-throughput screening (HTS) or ultra HTS, which test large chemical or biological compound libraries against specific biological targets, became an extremely promising method in drug discovery in recent years. After this very first experimental process, the identified hits must be validated. In order to confirm specificity of the hits, additional screening techniques exploring binding interactions between the biological macromolecules or with small-molecules are necessary.

One of the secondary screening methods allowing to move forward with the hits selected by HTS is isothermal titration calorimetry (ITC). By direct monitoring of the heat changes during binding events, ITC enables precise measurement of affinity of studied molecules, giving information about binding constants (KD), reaction stoichiometry (n), enthalpy (∆H) and entropy (ΔS). Such data allow for the rapid and facile identification of non-specific binders, targets with multiple binding sites and unstable protein targets. A single ITC experiment provides a sensitive and accurate means to study the complete basic thermodynamic profile of binding reactions involving small organic molecules (e.g. potential medicinal compounds) or larger macromolecules (such as proteins or nucleic acids). Moreover, ITC goes beyond binding affinities and can elucidate the mechanisms underlying molecular interactions. This deeper understanding of structure‑function relationships enables more confident decision making in hit selection and lead optimization.

The ITC analysis is based on the thermodynamic nature of binding of molecules. An interaction between two molecules is always associated with release or absorption of heat, resulting in temperature changes in the environment of the reaction. Detection of these changes is a principle of ITC technique. Several different ITC devices are available on the market. They differ in sensitivity, sample volume, degree of automatization or throughput. However, the main components are similar and are presented on the picture below.

Figure 1. A general schematic of an ITC instrument

All above are crucial components of the core of ITC device, where molecules are mixed, interact with each other and the changes of temperature are recorded. Beyond these basic components, several others might be included – like computer unit, washing module or automatization station.

In practice, the ITC measurement is a fairly simple and not overly time‑taxing procedure. It starts with the preparation of buffer and solutions of both molecules of interest,  usually a protein and a small compound. One challenge is to prepare both solutions in exactly the same buffer. Ideally, they should differ only in tested molecule, which can be achieved by dialysis of the samples. In the second step, a reference cell is filled with degassed water or buffer, the sample cell with a solution of larger molecule, and the titration syringe with compound dissolved in the buffer. In commonly used devices the further, last action is just initiation of the protocol; the software manages automatic addition of compound solution to sample cell and the changes in temperature are recorded by extremally sensitive sensors. Finally, the data is processed and determined parameters (like ∆H or -T∆S) might be extracted for further analysis.

To further elaborate on the ITC principles, the example of analysis of interaction between two small molecules – EDTA and Ca²⁺ is presented. The binding of divalent ions by chelating agents such EDTA or EGTA is one of the most important aspects of molecular biology research and is also used for calibration and maintenance of ITC devices. An example of such experiment is presented in Figure 2.

Figure 2. Chelating of Ca²⁺ ions by EDTA. Upper panel: raw heats released by binding events; lower panel: the graph enabling calculation of molar ratio of binding. Each peak showed on the first graph represents single portion of Ca²⁺ added to EDTA solution. The system always equilibrates until the next addition of compound.

Analyzing above charts we can conclude that:

• the interaction between EDTA and Ca²⁺ is exothermic, which is exhibited as a decrease of compensation heat
• a stoichiometry of the binding is one molecule of EDTA to one molecule of Ca²⁺

Going forward with the analysis, we can calculate thermodynamic parameters, which allows us to determine preferable interactions between tested molecules. Figure 3. showed that both hydrogen bonding (represented by ΔH) and hydrophobic interactions (-TΔS) are preferable, thus binding of EDTA and Ca²⁺ is thermodynamically favorable.

Figure 3. A bar chart presenting calculated thermodynamic parameters.

The above experiment is a very simple example of applying the ITC technique to the analysis of biomolecule interactions. ITC analysis is advantageous also in drug discovery campaigns where several promising compounds, selected during HTS, need to be validated and their structures optimized for boosting potency of drug candidates. One of the approaches for structure optimization is analysis of enthalpy, entropy and their contribution to binding. The strategy can be utilized in Structure-Activity Relationship approach by simply designing compounds with different substituents, analysis of thermodynamic parameters by ITC and selection of the compound with the most favorable binding properties. The great example is First in Class to Best in Class progression of HIV-1 protease inhibitors, presented on Figure 4. Despite of highly favorably entropic contribution to binding, for first drugs approved by FDA the enthalpy is positive hence the hydrogen bonding is rather unfavorable interaction. Further optimization of drugs structures resulted in increase of the enthalpy contribution to binding, leading to boosting of affinity and selectivity of the drugs.

Figure 4. First in Class to Best in Class progression of HIV-1 protease inhibitors.

All of the screening methods employed in drug discovery campaigns have pros and cons. As a positive side, ITC allows for the measurement of all binding and thermodynamic parameters in a single experiment, which is a significant advantage over the competition. ITC is fundamental, but at the same time very sensitive method, which can be used to study complex intermolecular interactions, often employing multiple binding sites and quantifying multiple KDs, which is difficult to accomplish by other methods. Identifying non-specific binding and eliminating false‑positive results are also hallmarks. ITC measures a relatively broad range of binding affinities and has basically no molecular weight limitations of the studied molecules, as well as an ability to measure native molecules in solution, in a range of aqueous solvent and buffer conditions. The ease of use and simple setup are also important factors as in most cases the assay requires neither any development or optimisation steps, nor immobilisation or labelling of the protein. An ability to measure protein purity also plays a non-trivial role here, as it allows determination of the quality of identified hits in HTS campaigns. 

As with any other method, however, ITC has some limitations. First and foremost, it is not suitable for primary HTS screening. High consumption of source material (protein and/or other studied molecules) is still an important drawback, despite this requirement reducing significantly in recent years. Also, the slowness, as well as a low throughput of the measurement caused by the necessity to perform separate experiments for each binding target, might be an issue. Another notable disadvantage is that ITC cannot determine reaction kinetics (i.e. association or dissociation rate constants), resulting in a need for the employment of a secondary method, should that information be required. Moreover, it should also be noted that since the ITC signal is directly proportional to the binding enthalpy, the non-covalent interactions may show small binding enthalpies.

Table 1. A general comparison of biophysical methods measuring biomolecular interactions.

Regarding methods alternative to ITC, which provide the researcher with similar information, the Surface Plasmon Resonance (SPR) technique seems to be the main contender. SPR is also a label-free technique, which allows measuring biomolecular interactions in real time. It is based on immobilisation of one molecule onto the sensor surface (which is usually covered with a thin layer of gold), while the analyte is passing free over that surface. Binding of the analyte causes the refractive properties of the surface to change, which are then measured optically. SPR is a sensitive and accurate method, which in contrast to ITC can also measure kinetics in addition to the biding affinities and thermodynamic parameters, with considerably higher throughput. On the flip side, performing SPR analyses requires expensive instruments, sensors and costly maintenance, as well as a highly trained researcher. It is, however, the only technique that fulfils the regulatory requirements of FDA, EMA and other similar organisations. There exists a less expensive variant of the SPR, called Localised Surface Plasmon Resonance (LSPR), where the continuous film of silver or gold on the chip surface is substituted with metal nanoparticles. It still however requires an immobilisation of the binding targets and has a lower throughput than the more advanced devices.

There are two other methods, which are often compared to ITC and SPR. One is Biolayer Interferometry (BLI), where the biomolecules of interest are usually adsorbed to the tips of optical fibers, so it requires immobilisation. The interference of white light reflected from the fiber‑biomolecular layer and biomolecular layer‑buffer surfaces is then measured, providing information about concentrations, direct binding affinities and interaction kinetics. Crude samples can also be used. Unfortunately, the temperature control is limited, not allowing thermodynamic characterization, and the sensitivity is rather in the lower spectrum. However, if one can surpass the inability to measure thermodynamic parameters, BLI has the highest throughput of the discussed technologies.

Besides BLI, there is also a method called Microscale Thermophoresis (MST), which is somewhat similar to the ITC in that it is performed using thin capillaries filled with a free solution. A microscopic gradient of temperature is then applied, and the movement of molecules detected. MST might be used to study complex mixtures, consumes minimal amounts of the sample, and immobilisation is not required. It provides an information about binding affinities, thermodynamics and stoichiometry, and is relatively easy to perform. On the other hand, it doesn’t measure kinetics, requires fluorescent labelling of the sample, and non-specific binding might occur due to the labelling.

To conclude, there is no single universal biophysical method for measuring biomolecular interactions. The ITC analyses provide a lot of unique insights, which no other techniques can, however, they all have their merits and flaws. Ultimately, the responsibility lies on the researchers to choose the best-suited analytical technique for their current scientific needs.

• Authors:

Mateusz Biernacki, Ph.D., Scientist
Dorota Trzybulska, Ph.D., Scientist
Marcin Markiewicz, Scientist

To contact the authors please email selvita@selvita.com

• References:
  
  1. A Wei, Chin-Chuan, et al. “Isothermal Titration Calorimetry and Macromolecular Visualization for the Interaction of Lysozyme and Its Inhibitors.” Journal of Chemical Education, 92.9 (2015): 1552-1556.
  2. Feig AL. “Studying RNA-RNA and RNA-protein interactions by isothermal titration calorimetry.” Methods Enzymol. 2009; 468: 409–422.
  3. Chin-Chuan Wei, Drake Jensen, Tiffany Boyle, Leah C. O’Brien, Cristina De Meo, Nahid Shabestary, and Douglas J. Eder. “Isothermal Titration Calorimetry and Macromolecular Visualization for the Interaction of Lysozyme and Its Inhibitors.” Journal of Chemical Education 2015 92 (9), 1552-1556.
  4. Drescher DG, Ramakrishnan NA, Drescher MJ. “Surface plasmon resonance (SPR) analysis of binding interactions of proteins in inner-ear sensory epithelia.” Methods Mol Biol. 2009; 493: 323–343.
  5. Jerabek-Willemsen, Moran, et al. “MicroScale Thermophoresis: Interaction analysis and beyond.” Journal of Molecular Structure 1077 (2014): 101-113.
  6. Abdiche, Yasmina; Malashock, Dan; Pinkerton, Alanna; Pons, Jaume. “Determining kinetics and affinities of protein interactions using a parallel real-time label-free biosensor, the Octet”. Analytical Biochemistry (15 June 2008), 377 (2): 209–217.
  7. Selecting the Best HTS Hits to Move Forward: ITC Ligand Binding Characterization Provides Guidance Ronan O’Brien and Richard Brown.
  8. Xing et al., Insights into Protein-Ligand Interactions: Mechanisms, Models, and Methods, Int J Mol Sci. 2016 Feb; 17(2): 144.
  9. Wafaa Al-Zuhairi. Revisiting Lysozyme and Tri-N-acetylglucosamine Interaction: Preparation and Characterization. 2014. Master Thesis, Graduate School Southern Illinois University Edwardsville.
  10. Freire E. Do Enthalpy and Entropy Distinguish First in Class From Best in Class? 2008. Drug Discovery Today 13(19-20): 869–874.