Microscale thermophoresis

Measure binding in solution using very little sample.

KDSimulator readyCore simulatorEngine: Binding, biophysics & structure data

Our ruleSimulations prepare you for the bench but don't replace hands-on practice.

1Molecules that move in a temperature gradient

Most molecules drift along a temperature gradient, usually from hot to cold. The effect, thermophoresis, depends on the molecule's size, its charge and above all the layer of water and ions bound to its surface, so it changes when a ligand binds even if the ligand is small.

2Designing the titration

Two concentrations decide whether the curve reports the affinity. The labelled partner should sit at or below the expected KD; if it is far above, almost every ligand molecule added binds until the target runs out, the curve's midpoint falls at about half the target concentration whatever the KD, and the affinity hides in the rounding of a corner.

label

Fluorescence

200–2000 counts is the usual working range: dim samples are noisy, and the label should report the same in every capillary.

readout

Fnorm

Hot over cold fluorescence, in ‰. The binding amplitude is often only 5–20‰, so noise of a fraction of a ‰ matters.

check

Capillary scan

The cold fluorescence of every capillary; it should agree within about ±10% and show clean, symmetric peaks.

fit

KD

Quadratic fit with the labelled concentration known; report it with a confidence range from independent series.

3How it goes wrong

MST responds to anything that changes how the labelled molecules move or how brightly they shine, which is its strength and its main risk. Protein sticking to the glass lowers and scatters the cold fluorescence from capillary to capillary and adds a fraction that heats without moving; a little detergent or a carrier protein in the buffer, or coated capillaries, usually cures it.

Thermophoresis

Directed movement of molecules in a temperature gradient, sensitive to size, charge and hydration shell.

Fnorm

Normalised fluorescence: hot-region fluorescence divided by the initial cold fluorescence, in parts per thousand.

Temperature jump (T-jump)

The fast fluorescence change within the first second of heating, caused by the dye's own temperature dependence rather than by movement.

Capillary scan

A scan along the capillaries before heating that records each one's fluorescence, used to spot adsorption, aggregation and pipetting errors.

Ligand depletion

Removal of free ligand by binding to the labelled partner, significant when that partner is not far below KD; handled by the quadratic fit.

MST power

Infrared laser power, which sets how far the spot heats; higher power gives a larger signal but more heating.

SD test

A control in which samples are denatured with SDS and heat before measurement, to tell a ligand-induced fluorescence change from a loss of protein.

When to use it

  • Measuring KD in solution when protein is scarce: microlitres per capillary and nanomolar labelled partner.
  • Interactions that are hard to immobilise, such as membrane proteins in detergent or nanodiscs, or binding in serum, lysate or other complex liquids.
  • Small-molecule or fragment binding to a protein, including weak binders up to the millimolar range when the ligand is soluble enough.
  • Quick checks of affinity across mutants or buffer conditions before committing material to calorimetry.

Compared with alternatives

Isothermal titration calorimetry →

Label-free and gives ΔH and n as well as KD, but needs far more protein; prefer it when the thermodynamics matter and material allows.

SPR & BLI →

Gives on- and off-rates, which MST cannot, at the cost of immobilising one partner on a surface.

Fluorescence polarization →

Plate-based and higher throughput when a small labelled ligand binds a much larger partner; MST works whichever partner is labelled.