Thermal shift assays

The original plate-based thermal shift: protein, SYPRO Orange and compound in 20 µL wells of a qPCR plate, heated from 25 to 95 °C while the dye's fluorescence is read. Cheap, fast and high-throughput, with Tm taken from the first-derivative maximum of each well.

TmSimulator readyCore simulatorEngine: Binding, biophysics & structure data

Choose a sub-technique

In learning pathsMake and purify a protein · step 5 of 6 · next: Dynamic light scattering →

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

Sub-technique you chose

DSF

How it differs from the standard method

Typically 1–5 µM protein and 2–10× SYPRO Orange; the protein must not already expose large hydrophobic patches, or the baseline starts high.
Detergents, lipids and some compounds bind the dye and raise the background; compound autofluorescence can also distort curves.
The curve falls after the peak because of aggregation, so only the rising part is analysed.
Absolute Tms depend on protein concentration, dye and ramp rate; compare ΔTm within one plate and one set of conditions.

In this lesson · about 4 min read

Melting as a readout of binding
Three ways to see unfolding
Thermal shift in cells
How thermal shift assays mislead
Key terms
When to use it
Sub-techniques

1Melting as a readout of binding

Every folded protein unfolds if you heat it enough, and the temperature at which half of it has unfolded, the melting temperature Tm, is a convenient summary of how stable it is. A ligand that binds the folded state more tightly than the unfolded state pays part of the cost of keeping the protein folded, so the protein has to be heated further before it unfolds.

2Three ways to see unfolding

DSF

An environment-sensitive dye

SYPRO Orange is dark in water and bright on exposed hydrophobic surface; fluorescence rises as the protein unfolds, read in a qPCR machine across a 96- or 384-well plate.

nanoDSF

The protein's own tryptophans

Buried tryptophans emit near 330 nm, exposed ones near 350 nm; the 350/330 ratio rises on unfolding, with no dye, in 10 µL capillaries.

CETSA

What stays soluble in cells

Cells are heated, unfolded protein aggregates, aggregates are spun out and the soluble target is quantified, usually by western blot.

In differential scanning fluorimetry (DSF) the curve rises from a low baseline as dye binds the unfolding protein, peaks, and then falls as the unfolded protein aggregates and the dye is excluded or quenched. Tm is the inflection of the rising part, found as the maximum of the first derivative or by fitting a sigmoid to the rise, not the temperature of maximum fluorescence, which is set by aggregation and moves with protein concentration.

3Thermal shift in cells

The cellular thermal shift assay (CETSA) applies the same idea inside intact cells. Cells are treated with the compound or vehicle, divided into aliquots, heated for a few minutes at a series of temperatures, lysed gently and centrifuged hard.

4How thermal shift assays mislead

A control that differs from the compound wells: DMSO lowers the Tm of many proteins by a few tenths of a degree per percent, so the apo wells must carry the same DMSO.
Ligand depletion: at protein concentrations similar to the ligand concentration, most of the ligand is bound and the shift is smaller than the compound deserves.
Background from the buffer: detergents and lipids light up SYPRO Orange; high dye concentrations raise the background and can destabilise the native state.
Destabilising binders and unfolded-state binders give negative or no shifts; absence of a shift is not proof of absence of binding.
In CETSA, weak centrifugation or harsh detergent lysis leaves aggregates counted as soluble, flattening both curves.

Tm: The midpoint of a thermal unfolding transition, where half the protein is unfolded; in DSF read from the first-derivative maximum.
ΔTm: The difference between the Tm with ligand and with a matched control; positive for stabilising ligands.
Tagg: The midpoint of a CETSA aggregation curve; a kinetic, condition-dependent quantity, not a thermodynamic Tm.
Isothermal dose–response (ITDR): CETSA at a single temperature with a compound concentration series, giving an apparent potency of target engagement in cells.

Key terms

Practise these as flashcards →

Differential scanning fluorimetry (DSF)

Thermal shift assay with an environment-sensitive dye such as SYPRO Orange, usually run in a qPCR instrument.

nanoDSF

Label-free thermal shift assay following intrinsic tryptophan fluorescence at 330 and 350 nm in capillaries.

CETSA

Cellular thermal shift assay: thermal stabilisation of a target measured as soluble protein after heating intact cells or lysates.

SYPRO Orange

A dye whose fluorescence increases strongly in hydrophobic environments; supplied as a 5000× stock and used at about 2–10×.

First-derivative Tm

Tm taken as the temperature of the steepest rise of the melting curve, the standard analysis for DSF and nanoDSF.

Ligand depletion

Loss of free ligand to binding when protein and ligand concentrations are similar, which shrinks the observed shift.

Target engagement

Evidence that a compound physically binds its intended target in the relevant setting, such as intact cells.

Aggregation onset

The temperature at which scattering from aggregating protein starts to rise; reported by nanoDSF alongside Tm.

When to use it

Screening a fragment or compound library for binders to a purified protein, where hundreds of conditions must be tested with little protein and no labelled ligand.
Finding buffers, salts and additives that stabilise a protein before crystallisation, cryo-EM or long-term storage.
Comparing the stability of mutants or constructs, or checking batch-to-batch consistency of a purified protein.
Showing that a drug candidate reaches and binds its target in intact cells or tissue (CETSA), without modifying the compound or the protein.

Compared with alternatives

Differential scanning calorimetry →

Measures the heat of unfolding directly, giving ΔH and ΔCp as well as Tm and resolving multiple domains; prefer it when thermodynamics matter more than throughput.

Isothermal titration calorimetry →

Gives KD, stoichiometry and binding enthalpy at constant temperature; prefer it to turn a thermal-shift hit into an affinity.

SPR & BLI →

Measures binding kinetics and KD directly; prefer it to confirm hits and to catch binders that do not stabilise.

Circular dichroism →

Follows unfolding through secondary structure with much less throughput, but tells you what changes structurally.

Intrinsic fluorescence →

The cuvette version of the nanoDSF signal, used for chemical denaturation and equilibrium binding at higher precision than a plate.

Sub-techniques

Choosing one here switches the whole page, including the protocol and the lab, to it.

Typically 1–5 µM protein and 2–10× SYPRO Orange; the protein must not already expose large hydrophobic patches, or the baseline starts high.
Detergents, lipids and some compounds bind the dye and raise the background; compound autofluorescence can also distort curves.
The curve falls after the peak because of aggregation, so only the rising part is analysed.
Absolute Tms depend on protein concentration, dye and ramp rate; compare ΔTm within one plate and one set of conditions.