Protein Thermal Shift Assay

This protocol automates the thermal shift assay on the Transcriptic platform. By detecting changes in a protein’s thermal stability, this assay can be used for a variety of applications during drug development. This includes evaluating protein stability for quality control, screening for protein/ligand binding, identifying the effect of mutations on protein folding, and characterizing other functional properties of a protein.

OVERVIEW

The protein thermal shift assay is an important technique for studying protein thermodynamics. In principle, this assay works by using a protein’s melting temperature as a readout for its thermal stability. By screening for changes to the melting temperature under varying conditions, this technique can be used to identify previously unknown intramolecular and intermolecular interactions involving a protein of interest.

To implement this assay on the Transcriptic platform, we used the Protein Thermal Shift Starter Kit offered by Thermo Fisher Scientific1. This kit offers several key advantages over comparable kits, including low reagent costs, minimal reaction volumes, and compatibility with high-throughput applications in 96-well format. To assess thermal stability, the kit uses a dye that emits a fluorescent signal when bound to hydrophobic amino acid residues. As a protein is heated in the presence of this dye, the fluorescent signal increases with protein unfolding, allowing fluorescence as a function of temperature to be used as a readout of melting temperature. The simplicity and versatility of this method makes it well-suited for high-throughput screening, especially when alternative methods of thermodynamic measurements are unavailable or cost-prohibitive.

At Transcriptic, we validated this protocol using a control reaction included in the commercial kit as well as bovine serum albumin (BSA) samples to demonstrate two applications: protein quality control and ligand binding screening. In conjunction, we have also developed a data analysis tool to simplify data interpretation. Once raw data has been generated from a run, this tool calculates the average melting temperature and temperature of maximum fluorescence of each sample, and uses statistical analyses to compare these samples to a reference sample.

The Protein Thermal Shift Starter Kit is a versatile kit that been designed for general use. In addition to the applications described here, it can be used with a variety of aqueous buffers and water soluble proteins. For other applications of this kit, please contact [email protected] to make a special onboarding request.

KIT VALIDATION - EVALUATING PROTEIN STABILITY USING KIT CONTROLS

METHODOLOGY

Protein stability was measured using the Protein Thermal Shift Starter Kit (Thermo Fisher Scientific)1. Reactions were prepared in 96-well format by combining thermal shift dye, buffer, and a control protein in combination with a known ligand binding partner (note that acceptable protein stock concentrations for this kit range from 2.5 ng/µL – 250 ng/µL). Both a protein-only and a buffer-only sample was included as well. Final reaction volumes for each sample came to 20 µL. Fluorescent signal was detected as a function of temperature using a CFX96 Real-Time PCR Detection System (Bio-Rad)2. Raw data were retrieved using the Transcriptic webapp, and melt curves were generated by plotting the fluorescent signal or the first derivative of the fluorescent signal as a function of temperature. Rightward shifts in these curves represent an increase in thermal stability.

RESULTS

As expected, adding a ligand known to bind the control protein increased its thermal stability, resulting in a rightward shift of the melt curve by approximately 5℃ (see Figures 1-2, Protein + Ligand as compared to Protein Only). The flat signal in the buffer-only control verifies the purity of the reagents used. Together, these data demonstrate that this protocol, as executed on the Transcriptic platform, can accurately identify conditions that change the thermal stability of a protein of interest. This method can be easily extended for use with any protein of interest, and under any condition(s) hypothesized to affect the thermal stability of a protein.

Figure 1.(A) Raw fluorescence (RFU) of kit controls was measured as a function of temperature. (B) The addition of ligand to the control protein caused a statistically significant shift in the temperature of maximum fluorescence (p<0.05).

Figure 2. (A) The first derivative of the fluorescence of kit controls was measured with respect to temperature. (B) The addition of a ligand to the control protein caused a statistically significant shift in the temperature of the melt peak (p<0.05). This rightward shift represents an increase in thermal stability.

APPLICATION 1 - PROTEIN QUALITY CONTROL USING BOVINE SERUM ALBUMIN (BSA)

METHODOLOGY

To validate this core thermal shift protocol as a tool for protein quality control, we tested a common protein, Bovine Serum Albumin (BSA, Sigma Aldrich)3, subjected to three different environmental conditions: incubated at 4°C (control group), incubated at 100°C for 30 minutes (boiled group), and subjected to 3 freeze-thaw cycles, in which samples were incubated at -80°C for 10 minutes followed by thawing at room temperature (freeze-thaw group). Samples were assayed using the Protein Thermal Shift Starter Kit, as described above. As different environmental conditions, including the ones mentioned above, have been reported to affect protein thermodynamics, we anticipated that this kit would be able to detect these changes.

RESULTS

As hypothesized, boiled BSA samples could be distinguished from control samples using the protein thermal shift assay. Compared to the control reference sample, boiled samples exhibited a reduction in the temperature of maximum fluorescence (Figure 3A) and a notable lack of detectable peaks in the derivative plots (Figure 4A). In contrast, none of the control or freeze-thawed BSA samples demonstrated a significant difference from the control reference sample. The analysis module validated this qualitative observation, showing that only the boiled samples had significantly different temperatures as compared to the reference protein (Figures 3B and 4B). These results show that the thermal shift assay can be used on the Transcriptic platform as a high-throughput tool to screen different proteins for quality control.

Figure 3. (A) Raw fluorescence (RFU) of 4 uM BSA samples was measured as a function of temperature. (B) While freeze-thawed samples were indistinguishable from control, boiled samples showed a statistically significant difference in the temperature of maximum fluorescence (p<0.05). All samples were compared to a control reference sample (aggregated here with additional control samples that were measured in parallel).

Figure 4. (A) The first derivative of the fluorescence for 4 uM BSA samples was measured with respect to temperature. (B) While melt peaks were detected for freeze-thawed as well as control samples, none of the boiled samples produced a detectable peak. All samples were compared to a control reference sample (aggregated here with additional control samples that were measured in parallel).

APPLICATION 2 - SCREENING FOR LIGAND BINDING USING BSA AND TREHALOSE

METHODOLOGY

To further validate this protocol for ligand screening applications, we tested BSA in combination with the disaccharide trehalose. Trehalose was previously found to bind BSA and increase its thermal stability4, so we expected the addition of trehalose to cause an increase in thermal stability that could be detected using this assay. We tested 4 uM BSA in combination with 3.5% w/w trehalose (TCI Chemicals)5 compared to vehicle-only controls. Samples were assayed using the Protein Thermal Shift Starter Kit, as described above.

RESULTS

Addition of trehalose did not result in a detectable change in the temperature of maximum fluorescence (Figure 5). However, as expected, the addition of Trehalose resulted in a large and statistically significant change in thermal stability, with a shift of approximately 20 ℃ in the melt peak (Figure 6). These results demonstrate that the thermal shift protocol, as executed on the Transcriptic platform, can be used in high-throughput to screen for ligands or other binding partners, based on their effect on the thermal stability of a protein.

Figure 5. (A) Raw fluorescence data (RFU) of BSA/trehalose samples as a function of temperature. (B) There was not a statistically significant difference in the temperature of maximum fluorescence for BSA + trehalose compared to vehicle-only controls. All samples were compared to a control reference sample (aggregated here with additional control samples that were measured in parallel).

Figure 6. (A) The first derivative of the fluorescence of BSA/trehalose samples was measured with respect to temperature. (B) The addition of trehalose resulted in a statistically significant 20 ℃ shift in the melting temperature of BSA (p < 0.05). All samples were compared to a control reference sample (aggregated here with additional control samples that were measured in parallel).

REFERENCES
  1. https://www.thermofisher.com/order/catalog/product/4462263

  2. http://www.bio-rad.com/en-us/product/cfx96-touch-real-time-pcr-detection-system

  3. http://www.sigmaaldrich.com/catalog/product/sigma/05470?lang=en&region=US

  4. Hédoux, A., Willart, J. F., Paccou, L., Guinet, Y., Affouard, F., Lerbret, A., & Descamps, M. (2009). Thermostabilization mechanism of bovine serum albumin by trehalose. The Journal of Physical Chemistry B, 113(17), 6119-6126.

  5. http://www.tcichemicals.com/eshop/en/us/commodity/T0832/