Thermal shift assay

A thermal shift assay compares a protein's temperature-dependent unfolding transition between experimental conditions, often with and without a candidate ligand. The measured shift reports altered thermal behaviour. It is commonly used to investigate ligand binding without first requiring an assay of the protein's biological activity.^[ref]

In differential scanning fluorimetry, an environment-sensitive fluorescent dye reports changes during heating. A temperature-controlled instrument measures fluorescence from matched control and compound-containing samples. The resulting apparent protein unfolding midpoint is a feature of the measured transition; it need not be a reversible equilibrium melting temperature.^[ref]

For an invented example, suppose a consistently defined midpoint changes from $48{}^{\circ }\mathrm{C}$ in the control to $52{}^{\circ }\mathrm{C}$ with a compound. ThenThis indicates thermal stabilization under the tested conditions. It does not by itself give a room-temperature dissociation constant or establish inhibition of enzyme activity. The 2004 study compared the method with isothermal titration calorimetry and activity assays, illustrating the value of an independent readout.^[ref]

Protein denaturation, protein aggregation and cleavage of peptide bonds are different events. A fluorescence trace alone must not be assumed to identify all three.

An apparent protein unfolding midpoint is an operationally defined temperature marking the middle of an observed unfolding-associated transition. In a thermal shift assay, it is often written $T_m$ or $T_{m,\mathrm{app}}$. “Apparent” keeps the measured reporter transition distinct from an established reversible equilibrium melting point.^[ref]

Consider a deliberately simple two-state reporter model. If $f_U$ is the unfolded fraction and the folded and unfolded signals are constant values $F_F,F_U$, thenAt $F=(F_F+F_U)/2$, algebra gives $f_U=1/2$, provided $F_U\ne F_F$. If the signal changes from $100$ to $800$ arbitrary units, its midpoint is $450$ units. This arithmetic identifies half-unfolding only under the stated reporter assumptions.

Real traces can have drifting baselines, irreversible changes or protein aggregation. A fitted midpoint or derivative peak must therefore be defined consistently when comparing samples. A higher apparent midpoint supports greater thermal stability under those conditions; it is not by itself a dissociation constant or proof that the protein retains biological activity.

Differential scanning fluorimetry follows fluorescence while a protein sample undergoes controlled heating. Changes associated with protein unfolding provide a transition curve and an apparent protein unfolding midpoint. Comparing matched conditions turns the method into a thermal shift assay.^[ref]

In a common implementation, an environment-sensitive fluorescent dye reports changes as normally buried regions become exposed. Compare otherwise matched protein-plus-dye samples with and without a candidate ligand, rather than comparing measurements made with different buffers or heating programs. Fluorescent compound-only and dye-only controls can help diagnose reporter interference; these are proposed controls, not specific settings claimed for every published assay.

For an illustrative result, midpoints of $320\mathrm{K}$ and $324\mathrm{K}$ give a thermal shift of $+4\mathrm{K}$. The measured quantity is a change in a heating-dependent transition. An independent ligand binding or activity measurement is needed to assess the proposed molecular mechanism, as illustrated by the comparisons with isothermal titration calorimetry in the 2004 evaluation.^[ref]