In general, there are no limitations for the assay buffer used in MST. This includes buffer viscosity. Buffer components like glycerol are therefore not a problem. Importantly, the buffer composition needs to be constant throughout one MST experiment. Please take care not to introduce gradients of e.g. glycerol when creating a serial dilution of one of the interaction partners.

However, viscosity effects introduced by high ligand concentrations are usually not an issue in MST. For example, maltose causes no change in the MST signal of a non-binding protein (lysozyme) even when used at concentrations as high as 500 mM. More experiments showing the very limited influence of high concentrations of common buffer additives and ligands were performed and the results are outlined below.

If viscosity effects arise, they can be demonstrated to be unspecific since they never reach saturation.

If concentration gradients of solvents and other additives do arise, how will they affect the MST signal?

To demonstrate the effect of different solvent concentrations on binding signals and to show the robustness of MST towards changes in solvent or additive concentration, we applied a wide range of different concentrations of three commonly used solvents/additives (DMSO, glycerol and sucrose) and tested the thermophoresis of fluorescently labeled BSA:

The MST signal of BSA is constant over a wide range of increasing concentrations of DMSO, glycerol and sucrose, underlining that MST is a very robust method and not affected by slight changes in additive or solvent content. DMSO or glycerol affect the MST signal only at very high concentrations of > 2 % (w/v) or > 5 % (w/v), respectively. Sucrose only slightly influences the MST signal of BSA (≈ 3-4 F_norm) at concentrations ≥ 100 mM.

Please note: Measurements at very high additive/solvent concentrations (> 2 % DMSO, > 5 % glycerol, > 100 mM sucrose) are of course possible without limitations when applied at constant concentrations throughout the dilution series. Note, however, that high amounts of organic solvent might lead to issues with the temperature gradient since the IR laser with a wavelength of 1480 nm is primarily absorbed by water molecules.

Do high concentrations of ligands, like proteins or compounds, affect MST indirectly by changing viscosity?

To illustrate that the MST signal of a target is not affected indirectly by a changing viscosity of the solution, we analyzed the MST signal of labeled BSA with increasing concentrations of different non-binding proteins and small molecules. These increasing concentrations are comparable to the ligand concentrations in a typical MST experiment.

Protein concentrations up to 70 mg/ml typically do not affect the MST signal of your target of interest at all.

Only at concentrations ≥ 100 mg/ml, some proteins might have a slight effect on the MST signal. Keep in mind that those concentrations are extremely high (e.g. 100 mg/ml lysozyme corresponds to ≈ 7700 µM), and are rarely used in binding experiments. The slight change of the MST signal upon addition of high protein concentrations can also be attributed to unspecific binding to BSA.

Small molecule compounds usually do not affect the MST signal of labeled proteins by unspecific viscosity effects at all.

In some cases, for example when using unusually high ligand concentrations or high molecular weight ligands, control experiments can be helpful to exclude viscosity artefacts. To this end, titrate the same ligand against an unrelated target, or alternatively, titrate a similar but non-binding ligand to the target of interest.