What is the Dissociation Constant (KD)?
Understanding the interactions between proteins and other biomolecules is a cornerstone of biochemistry and pharmacology. Whether it is a drug binding to a receptor or an enzyme binding to its substrate, determining the strength of interactions (or binding affinity), is a critical parameter. One of the most widely used metrics of binding affinity is the dissociation constant (KD). KD is defined as the concentration of ligand at which half of a partner’s binding sites are occupied when at equilibrium. A low KD constant (typically in the picomolar and nanomolar range) describes a tight binding interaction and a high KD constant describes a weak interaction (typically in the micromolar range or higher).
Measuring the Dissociation Constant
There is a wide range of technologies that can determine affinity; many of these methods require incubation of a titration of ligand with a binding partner and measuring a concentration dependant change in an outcome variable. This enables a binding curve to be plotted, and the KD constant can be calculated mathematically or observed visually as its inflection point (Figure 1).
Figure 1. An example of an equilibrium binding curve measured with the Fluidity One-M. Microfluidic Diffusional Sizing (MDS) measures the change in hydrodynamic radius, Rh, (that results from binding) of a fluorescently labelled probe with increasing concentrations of ligand to determine KD constants and other interaction metrics.
Watch our tutorial to explore what KD is and how it is derived
Whilst the dissociation constant is clearly a powerful and important metric for characterising molecular interactions, like all metrics, it has its limitations. In this blog post, we will list three of the most common mistakes that scientists make when examining their KD constants, so you are able to avoid them in the interpretation of your data.
Three Common Mistakes to Avoid When Interpreting the Dissociation Constant
1. Ignoring experimental conditions
Dissociation constants (KD) are not universal values but are highly dependent on the experimental conditions under which they are measured. As well as assay temperature, particular consideration should be given to the buffer in which the interaction is measured, including its components and their concentrations (i.e detergents and additives), pH and ionic strength [1]. Manipulation of these factors can substantially influence affinity values by altering the stability of biomolecules and strength of interactions [2]. This variability underscores the importance of reporting detailed buffer conditions in publications and standardizing protocols during drug screening. Without consistent experimental parameters, comparing KD constants across investigations can lead to misleading conclusions.
Further to this, it should never be assumed that KD constants reported in laboratory buffers are unanimously replicated in complex biological samples. For instance, the behaviour of a drug binding to a receptor can differ drastically between aqueous buffers and serum [3]. As such, testing the affinity of drug candidates to their targets, across biologically relevant backgrounds, early in the development process can save pressure time and resources. Surface-based methods, e.g. Surface Plasmon Resonance (SPR), Biolayer Interferometry (BLI) and Enzyme-linked Immunosorbent Assay (ELISA), struggle to work with undiluted biological samples due to non-specific binding. However, in-solution methodologies, such as Microfluidic Diffusional sizing (MDS) is able to rapidly determine KD constants of targets across any background, including undiluted serum, cell media and cerebrospinal fluid. As well as enabling early testing of candidates in near native conditions, this capability allows the influence of different buffers to be tested and optimised for a range of experimental protocols.
2. Employing 1:1 binding models for complex interaction behaviours
Another frequent mistake when interpreting the KD constant is the assumption that every binding interaction follows a simple 1:1 model, where one ligand binds to one binding partner. While the 1:1 model is the most straightforward and is often used for convenience in early analyses, it does not accurately represent many biological systems.
In reality, many molecules have multiple binding sites, may undergo cooperative binding, or may form multimeric complexes, all of which deviate from a 1:1 model [4, 5] (Figure 2). Misapplying a 1:1 model can lead to reporting of incorrect KD constants, as well as masking important biological behaviours. It’s crucial to understand the biophysical nature of the interaction and select a fitting mathematical approach that reflects the underlying biology.
Figure 2: Binding may follow a 1:1 model (A) whereby binding partners have just one binding site for a ligand. However, in biology, it is much more common for binding partners to have more than 1 binding site for one ligand, leading to the formation of more complicated (and larger) complexes (B).
This highlights the importance of tools that give a fuller picture of binding events – those that go beyond the KD constant. MDS measures the change in hydrodynamic radius (Rh) of a fluorescent probe that occurs with binding to determine affinity, but also gives an automatic readout of probe size while unbound (free) and bound (in complex). Given that size is somewhat predictable, the magnitude of difference between the unbound and bound probe can be used to indicate binding stoichiometry.
For instance, if the KD constant of two hypothetical proteins is measured, each with a Rh of 1 nm, and the resulting complex size is given as 5 nm, this suggest a binding model that is greater than 1:1. If you are looking for a more quantitative metric, the NeSt assay is able to simultaneously determine the size, KD constant and interaction stoichiometry by through Bayesian inference within our AI/ML data analysis platform – Fluidity Insight.
3. Assuming KD constants determined by surface-based methods are reflected in-solution
Surface-based methods such as SPR, BLI and ELISA have been commonplace in biochemistry and pharmaceutical laboratories for many decades [6, 7] and have been developed more recently to offer high-throughput capabilities for the investigation of binding affinity. However, it should be noted that this group of technologies require one of the binding partners to be immobilized on a surface (chip, biosensor or well) – a system that is never reflected in biology. As well as being time-consuming to optimize, the immobilization of a molecule can alter its native conformation and orientation, as well as its activity in some cases due to steric hinderance [8]. In addition, these methods suffer from non-specific binding to the surface which can create artifacts [9].
Although these high-throughput methodologies remain powerful tools for the determination of KD constants, these limitations highlight the need for additional orthogonal techniques. In-solution technologies, such as MDS, offer an alternative to measure affinity, without the need for surface immobilization. As well as avoiding the limitations of this requirement, it enables interaction metrics to be determined in their near-native environments – with both binding partners in solution.
Conclusion
When used appropriately, the KD constant is a powerful and important metric in biochemistry and pharmaceutical research. However, it is important to consider all aspects of experimental design, conditions and binding behaviours when interpreting your data.
Choosing the right technologies is a crucial step for getting the most out of any analysis. MDS is an attractive technology for measuring affinity. By measuring the changes in molecular size (Rh) that occur with binding, not only does this produce an accurate KD constant, but it also gives physiologically meaningful metrics that can be used to infer binding stoichiometry. Further to this, having the capability to run complex biological samples, including undiluted serum, allows in-solution KD constants to be determined in native conditions.
References
1. Salis, Andrea, and Maura Monduzzi. “Not only pH. Specific buffer effects in biological systems.” Current Opinion in Colloid & Interface Science 23 (2016): 1-9.
2. Altschuler, S.E., Lewis, K.A. and Wuttke, D.S., 2013. Practical strategies for the evaluation of high-affinity protein/nucleic acid interactions. Journal of nucleic acids investigation, 4(1), p.19.
3. Blakeley, D., Sykes, D.A., Ensor, P., Bertran, E., Aston, P.J. and Charlton, S.J., 2015. Simulating the influence of plasma protein on measured receptor affinity in biochemical assays reveals the utility of Schild analysis for estimating compound affinity for plasma proteins. British Journal of Pharmacology, 172(21), pp.5037-5049.
4. Bellelli, A. and Carey, J. (2017). Proteins with Multiple Binding Sites. In Reversible Ligand Binding (eds A. Bellelli and J. Carey). https://doi.org/10.1002/9781119238508.ch4
5. Stefan, M.I. and Le Novère, N., 2013. Cooperative binding. PLoS computational biology, 9(6), p.e1003106.
6. Stenberg, E., Persson, B., Roos, H. and Urbaniczky, C., 1991. Quantitative determination of surface concentration of protein with surface plasmon resonance using radiolabeled proteins. Journal of colloid and interface science, 143(2), pp.513-526.
7. Liedberg, B., Nylander, C. and Lundström, I., 1995. Biosensing with surface plasmon resonance—how it all started. Biosensors and Bioelectronics, 10(8), pp.i-ix.
8. SPRpages.nl, (June 2025). Immobilization theory. [online] Available at: https://www.sprpages.nl/immobilization/theory [Accessed 5 June 2025].
9. Rich, R.L. and Myszka, D.G., 2007. Higher-throughput, label-free, real-time molecular interaction analysis. Analytical Biochemistry, 361(1), pp.1-6.