Thermo Stability Assay For Tnfr 1 Protein Stability Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

The mechanism by which tumor necrosis factor receptor (TNFR) cause the disease TNFR associated periodic syndrome (TRAPS) is thought to be via protein misfolding, aggregation and ligand independent signaling. This study aimed to develop a thermo stability assay that will allow the investigation of the differences in structural stability of wild-type TNFR 1 and mutant forms C33Y and R92Q. Protein unfolding was measured using SYPRO Orange protein dye which fluoresces when exposed to hydrophobic residues. Lysozyme, Carbonic Anhydrase and IgG1 were used to optimise the condition prior to testing of TNFR 1. We observed unusual results for TNFR 1 and its mutants, the results showed an inverted curve recording a decrease in fluorescence. We found that a large number of variables affect the results of this type of assay and as a consequence reaction conditions are required to be tailored each individual protein being tested.


Tumor necrosis factor receptor (TNFR) associated periodic syndrome (TRAPS; OMIM 142680) is an autoinflammatory periodic fever syndrome. It is associated with autosomal-dominant mutation in TNFR1, the gene that encodes tumor necrosis factor receptor (TNF R) [1]. There are more than 50 different mutations ( that have been identified in association with TRAPS within the TNFR superfamily 1A (TNFRSF1A). The majority of these mutations are likely to significantly affect the structure, conformation and stability of the receptors. Studies supporting this hypothesis suggest that these mutations TNFR may result in a conformational changes in the protein that leads to receptor misfolding, aggregation and ligand independent signalling, which explains the functional and clinical differences between the TRAPS-associated mutations [2, 3]. The propensity of the mutants to aggregate and independently signal suggest a confirmation that is less thermodynamically stable, this may results in prolonged exposure of hydrophobic residues and unfolded states.

TNFR1 is a transmembrane glycoprotein, with a conserved intracellular death domain (DD) that is responsible for receptor signalling, a transmembrane region and four tandem-repeat cysteine-rich domains (CRD 1-4) which contain three disulfide bonds. CRD1 is known as the preligand assembly domain (PLAD) which mediates ligand independent trimetric receptor interactions. All the TNFR 1 mutations associated with TRAPS are located within the CRD ectodomian of the receptor.

This study uses thermo stability assays to investigate differences in structural stability between wild type TNFR 1 and two key mutants, C33Y and R92Q. C33Y and R92Q are located in the CRD1 and CRD2 respectively. C33Y is likely to be highly disruptive to the overall conformation of the receptor as it abolishes one of the disulfide bonds found in the extracellular region of the receptor which is therefore expected to destabilize the protein. There is strong evidence to suggest that as well as a change in conformation [4] and lack of TNF binding, this mutations lead to lack of TNF binding and abnormal receptor function [3, 5]. R92Q is a mutation with low penetrance, and is predicted to no major impact on the protein structure [3], arginine is a basic amino acid which is replaced by a glutamine which has an uncharged polar side chain. It behaves very similarly to the WT receptor in terms of TNF binding and cell surface expression [6-8].

To date there are no publications which show experimental differences in the stability of TNF R1 and its mutations, although there are some computational studies which predicting the structural effects of theses mutations. All previous work focuses on the functional and behavioral difference of the mutant receptors. There are a growing number of diseases which are being recognized as conformational disorders such as Alzheimer's disease, Huntington's disease and Parkinson's disease. The pathological effect of these diseases is the result of misfolded proteins therefore understanding structural stability and the effect of mutations on a protein conformation is an important tool.

The structural stability of a protein can have a fundamental effect on its function at a physiological level. The importance of amino acid interactions within the tertiary structure of a protein means that a single residue change can result in a significant decrease in thermodynamic stability of the protein structure which can lead to aggregation, proteolysis or unwanted interaction. It would be possible, using thermostability profiles, to investigate difference in stability and unfolding between wild type and protein mutants. If a mutation is involved in the structural stability of the protein it can significantly affect the protein structure as well as the rate of folding and unfolding, this in turn can affect how the protein functions at a biological level resulting in disease.

Previously thermo stability assays have been applied as a high-throughput screen for protein engineering [9], structure determination[10], structural optimization of protein clones [11] and monoclonal antibodies[12] . It has also been used for crystalization screening [10] screening of stability promoting ligands [13] and to study interaction between target proteins and small molecules [13, 14]. Other techniques used to investigate protein stability and determine the melting temperature of a protein included differential scanning calorimetry (DSC), spectroscopy and circular dichroism but most of these techniques require large amounts of protein and can be difficult to modify for a high-throughput system.

The method being used in this study has been adapted from [15] this study used ANS (1-anilino-8-napthalenesulfonate) but because the ANS has an emission and excitation maxima that is not accessible by standard PCR filters, this study uses SYPRO orange, which has been successfully used in other thermo stability assays [9, 16]. SYPRO orange is highly fluorescent when exposed to hydrophobic regions in proteins. As the protein starts to unfold due to the increase in temperature, the hydrophobic regions of the protein become exposed binding to the dye. This increase in fluorescence can be measured using a Real-time PCR machine to produce a profile of protein unfolding. This profile should be distinct for each protein and as a result this technique can be used to investigate differences between mutated proteins.

Material & Methods

Optimization of Thermal Stability Assays

SYPRO Orange dye was obtained from Sigma. It was stored in dimethyl sulfoxide (DMSO) solution at 5000 times concentration. The dye was stored at -20OC. Before use the dye stock is diluted 1:20 to 250 times concentration in 20mM Tris.Cl (pH 7.1) (Sigma-Aldrich, St. Louis, MO), 150mM NaCl buffer. The final concentration of SYPRO orange during the experiment was 5 times. Protein solution contained either; Lysozyme (Sigma), Carbonic Anhydrase (Sigma) or mouse IgG1 isotype control (R&D systems Minneapolis, MN) at concentration between 1mg/ml - 0.005mg/ml, the proteins were diluted in one of three buffers; 25mM MES pH 6.1 (sigma) 50mM NaCl 2% DMSO, 100mM HEPES pH 7.0(Sigma) 150mM NaCl 2% DMSO or 20mM Tris pH 7.1(Sigma) 150mM NaCl, as described in figure legends. The protein solutions (5 and 20µl) were dispensed into 8x 0.2ml format strip tubes with optically clear caps (Stratagene, La Jolla, CA). The assays were performed using an MX4000 series multiplex quantitative Real-Time PCR system (Stratagene) the proteins were heated at a rate of 1oC/min and three data point were collected at the end points of each cycle. The fluorescence emission was collected, initially using standard Stratagene MX4000 filters; FAM, HEX and ROX as well as SYPRO orange specific bandpass filters purchased from Chroma (excitation 470/50 nm and emission 590/55 nm). Assays were performed from a starting temperature of 30oC and ending at 95oC. The results were analyzed using the Mx4000 software.

Thermal stability assays of TNF R1 wild type and mutations

TNFR 1 was supplied by Paul Radford from within the group. It was column purified FLAG tagged, full length TNFR 1, solubilised from the membrane using NP40 detergent and stored in Tris (Sigma) buffer at -20oC until use. The proteins were diluted to a concentration of 0.005mg/ml in 20mM Tris (Sigma) 150mM NaCl buffer, dye was added to a final concentration of 5 times. The MX4000 set up was as described above.

Flourescence excitation and emission spectra of SYPRO orange


Assay optimization

The thermostabilty profiles of Lysozyme, Carbonic Anhydrase and IgG 1

Low availability of TNFR 1 meant initial experiments used Carbonic Anhydrase, Lysozyme and IgG1 to optimize the reaction conditions and assess the limitations of the assay. It also allowed us to test the compatibility of the dye with the standard real-time filters.

Figure 2 (a-c) shows the thermo stability profiles of the Lysozyme, Carbonic Anhydrase and IgG1. Samples were heated at a steady rate and the fluorescence change was measured over time. Rapid rise in fluorescence intensity seen in all three profile suggest an increase in hydrophobicity in the environment of the dye this indicates protein unfolding. The dissociation points of each on the proteins can be seen in figure 2 (d-f), which represent derivative plots of the thermal fluorescence profile (change in fluorescence, DF, between temperature points). These are shown as negative peaks because the set up of the software is for real-time PCR reaction which uses decrease in fluorescence to indicate DNA melting and dye dissociation, in this study this function has been used to look at rapid changes in emission relating to temperatures where the most protein unfolding occurs. This temperature is referred to in this study as the TD (dissociation temperature) it relates to the steepest part of the melt curve. The melt curve and dissociation profile of all three protein are distinct, with different TD and curve shapes. It was observed that in all three profiles the fluorescence drops off slightly after the maximum point.

Lysozyme is a simple globular protein with only a single domain; Carbonic Anhydrase and IgG1 are more complex with defined domain structure. This can be seen in the melt curves, Lysozyme produces a simple smooth curve, whereas the results for the other proteins look more complex with a less rapid decline in fluorescence after the first rise in fluorescence.

The effect of assay buffer on thermo stability assay

Originally the proteins were tested in Tris buffer but we were unable to observe any results using Carbonic Anhydrase, because we were aware that there had been previous success with this protein we chose to perform further experiment using MES pH 6.1 and HEPES buffers pH 7.0. Using these reaction buffers we were able to gain results for Carbonic Anhydrase. All three proteins were repeatedly tested using all three buffers but no results were obtained for Lysozyme in HEPES. We assessed the effect of the buffer by looking at the TD of the protein under the different condition. Slight variations can be seen in the results (table 1).









Carbonic Anhydrase




IgG 1 isotype




Table 1. The TD (o C) of Lysozyme, Carbonic Anhydrase and IgG1 under various buffer conditions. Buffers are as stated in the material and methods. n=3

N/A indicates not clear TD was obtained during this study.

The effect of filter sets on thermo stability assay

Fluorescence intensity varied greatly depending on the filter set used to collect the melt curve data. The range of the filters can be seen in figure 1, the maximum fluorescence which was seen by each of the filters set is summarized in table 2. Initial experiments used existing ROX and HEX filters for data collection, the band pass for these filters fell within the excitation and emission spectrum of SYPRO orange (see figure 1) and required no internal adjustment to the Mx4000. Unfortunately the maximum fluorescence collected was 376 relative fluorescence units (rfu). Because of this low output alternative filters were tested. Previous studies [9] have used a combination, FAM(ex)/HEX (em), of the standard filters to increase the fluorescence intensity collect during their experiments. Although the FAM/HEX filter combination did improve the level of fluorescence collected to some extent, it was decide that to improve the sensitivity of the assay further we would purchase SYPRO orange specific filters from Chroma. This saw the most drastic improvement in maximal fluorescence. There was an almost 17 time increase in fluorescence when comparing the FAM/HEX filter set and specific filters in experiments using Lysozyme, and a 5 times increase when looking at the same condition using Carbonic Anhydrase. IgG1 showed a 14 times increase in maximal fluorescence using specific filters.

Filter sets




Specific SYPRO orange






Carbonic Anhydrase





IgG 1 isotype





Table 2. The maximum fluorescence (rfu) for Lysozyme, Carbonic Anhydrase and IgG1 using standard (ROX, FAM/Hex and HEX) filter set. In comparison with SYPRO orange specific filters. All reaction volumes 20µl. Lysozyme(0.6mg/ml) and IgG1 (0.5mg/ml) in 20mM Tris, 150mM NaCl, Carbonic Anhydrase (0.6mg/ml) tested in 25mM MES, 50mM NaCl, 2% DMSO

*No experiments were performed looking at the maximum fluorescence of IgG1using HEX filters

The effect of concentration on thermo stability assay

The concentration of the TNF R1 available to us was quite low and for this reason we chose to test the limits of this method by decreasing the concentration of protein used. Decease in protein concentration resulted in lower levels of fluorescence. Below a concentration of 0.01mg/ml signal became distorted and we were unable to obtain clear results using proteins at lower concentrations. Table 3 shows the maximum levels of fluorescence decreasing with decreasing concentration of protein.

Protein concentration mg/ml







Carbonic Anhydrase







Table 3. The maximum fluorescence for Carbonic Anhydrase using varing concentrations of protein. Reaction volume 5µl, dilution made using 25mM MES, 50mM NaCl, 2% DMSO.

* 20µl volumes

The effect of volume on the outcome of the thermo stability profile

It was observed that reaction volume played a large part in optimization of the assay. Increasing the reaction volume resulted in an increased the maximum fluorescence, Figure 3 show the effect of increasing the reaction volume on the shape and height of the curve. The results were obtained simultaneously and all three reactions were performed using the same reaction conditions. For Lysozyme we observed an almost 14 fold decrease in fluorescence which resulted from halving the reaction volume. For Carbonic Anhydrase the reduction in fluorescence was less drastic but we still observed that the fluorescence was decrease by a quarter as a result of the reaction volume being halved.

Thermo stability assay for TNF R1 WT, R92Q and C33Y

From the optimization experiment we selected conditions which would provide the best chance of producing a successful melt curve of TNF R. Unfortunately the concentrations of the wild type and mutant TNFR1 which were available were much lower than the concentration used in the initial experiments with test proteins. To overcome this we used larger volumes and the specific filters to maximize fluorescent output. Tris buffer was used as this was the buffer the protein was stored in and therefore seemed most suitable.

Despite this the levels of fluorescence are significantly lower when compared with the other proteins used in this study. We unable to obtain a clear TD from these results, and although the three curves were reproducible some variation can be seen between the three repeats. We were unable to identify any clear difference between the thermo stability profiles of wild type and mutant TNF R (figure 4).

The shape of the thermo stability profiles of TNF R mutants we collected is unlike any of the others seen in earlier experiments (figure 2) the curve appears to be inverted with fluorescence decreasing over time and as the temperature increases. This is further discussed in the next section.


The ability to study protein stability is an important tool when investigating the effect of protein mutations. The aim of this study was to develop a method which would allow the study the thermo stability of wild type and mutant TNF R 1. The concentration and volumes of TNFR 1 available to us were low therefore we decided to optimize the method using Lysozyme, Carbonic Anhydrase and IgG1. Previous studies have used thermal stability assays to study the effects of ligands on the protein stability of Carbonic Anhydrase [15]. This study showed Carbonic Anhydrase to be an ideal protein for this type of assay. Lysozyme and IgG1 are both proteins which were freely available and have well known defined structures.

The thermostability plots for all three proteins appeared distinct, giving different profiles; the Lysozyme plot was, as expected, the simplest of the three with a smooth increase to a peak fluorescence. The profiles for the other two proteins appear to be more complex with a rapid initial increase in fluorescence with some fluctuations, emphasized in the derivative plot which may be the results of specific domain unfolding. All three profiles showed a decrease in fluorescence towards the end of the run (ie higher temperature), this is mostly likely due to protein aggregation; an alternative explanation is that at the higher temperatures the dye becomes less stable, therefore reducing the level of fluorescence.

The effect of reaction buffer is an important factor to consider when interpreting these results. The difference between results which we observed when changing buffer means that in order for comparison to be draw between the two proteins both need to be assayed using the same buffer to prevent unwanted variation which resolved from the buffer. It may be that salt concentration and pH of the buffer has a large effect on the solubility of the protein, buffers with a pH near the pI of the protein may result in the protein coming out of solution. To ensure that this method will work for any given protein it is important to find a suitable buffer that will stabilize the protein and that is compatible with this technique.

There were similarities in the TD of Lysozyme and IgG1, both proteins gave a TD of 70.5-72.5 (buffer dependent), therefore it is important to consider whether the result obtained in the study were a true reflection on individual protein unfolding. Despite these similarities in dissociation of these two proteins the fluorescence profile obtained during these experiments were distinct (figure 2) the profile for IgG1 seems to plateau after the sharp rise in fluorescence whereas the profile for Lysozyme shows a rapid decrease in fluorescence after the peak in intensity. The shapes of these profiles were consistent thought out the study suggesting that they are in fact a product of the protein rather than the dye or an external contaminant.

It is worth noting that although Lysozyme, Carbonic Anhydrase and IgG1 were compatible with this method and gave clear profiles, we tested a wide range of proteins. Ribonuclease A and BSA were both used in initial experiments but neither gave clear or consistent results. We were able to obtain some results using Nuclear Histone protein and anti TNFR2 antibody but they were inconsistent and required further optimization to improve the definition of the curves. Anti TNFα antibody, Anti TNF1 antibody and TNFα were also tested but the profiles were unclear which is most likely the result of low protein concentrations. This method was first designed for high throughput screening but the inconstancy of the assay in regard to both protein and buffers used, means that in order for this assay to be successful the correct experimental environment need to be tailored to each individual protein being used.

We found that with more specific filters the concentration of protein required could be lower. The observation was also made that higher volumes of protein solution resulted in less interference and allowed the concentrations protein to be lowered even further. Due to the effect of the filter set on the reaction sensitivity we believe it is worth using optimal filters for this type of study.

It may be that this technique is not sensitive enough to differentiate between different domains unfolding. Higher resolution machines to provide a more detailed thermal profile and higher concentration of proteins may be enough to define the melt curve at a high enough resolution which will allow us to identify individual domain unfolding events.

The TNFR wild type and mutants gave surprising results with the fluorescence decreased as the temperature increased resulting in an inverted profile, which is unlike the results obtained when using the other proteins. Because these proteins are membrane associated, they must be solubilised using a detergent to allow purification before this type of experimental testing. It is possible that it is the interaction between SYPRO orange and the detergent at the start of the assay that resulted in the unexpected decrease in fluorescence. For this reason it may be that SYPRO orange is an unsuitable dye for this type of assay when using membrane proteins, therefore future experiment may require a change of dye to an alternative such as 1-anilino-8-naphthalenesulfonate (ANS) [15] which fluoresces in response aqueous environment, or N-[4-(7-diethylamino-4-methyl-3-coumarinyl)phenyl]maleimide (CPM)[17] which is a thio-specific dye which can reveal thermal stability of disulphide bonds within a protein.

Another explanation for the unusual results seen by TNFR1 is the protein itself. If the hydrophobic resides normally buried in the plasma membrane were exposed because of the solubilisation this would result in a high initial florescence. The conformation the receptor takes up after it's been removed from the membrane is unknown. Carbonic Anhydrase acts as a monomer whereas TNFR 1 is found to form trimers via the PLAD domain in the absence of TNF. It may be that the decrease in fluorescence accounts for the initial breaking of these interactions as the temperature increases. Further studies using site directed mutagenesis could look at the effect of mutation in the PLAD domain on the thermo stability profile of TNFR 1.

It may be that the concentration of TNFR 1 were too small for it to be assessed using this method, as a result further purification may be required. Although this may not be easily possible as other members of the research group have found that TNFR 1, and in particular C33Y, has a tendency to come out of solution when extensive purification is performed. This may be because WT TNFR 1 is itself is an unstable protein. Studies have shown that if highly expressed within cells, WT receptor also has a propensity to form aggregates [18]. For this reason it may be that this assay system is unable to cope with the instability of the protein as well as the size of the receptor.

Further studies could look at simplifing TNFR 1 for the purposes of the assay, performing experiments using the individual domains rather than the whole protein (figure 5). As the mutation for TRAPS are all located in the ectodomain it would be reasonable to study only this section of the protein, this would require no solublisation from the plasma membrane and therefore may be easier to purify. We could also use this technique to investigate the structural stability of the intracellular death domain (DD) and TNFR 1- TNF association. Alternatively we could simplify the model further by studying the local effects of the mutation on individual CRD.

Current treatments for TRAPS have had varying success, some patients respond to anti-TNF treatments but for others this treatment is ineffective and can lead to increased disease flares[19]. This is due to the differences in behavior between different TRAPS mutations. If we were able to successfully modify this technique so that it would allow us to study TNF R1 it would be possible to this assay to screen for ligands which are able to stabilize the mutant forms of the protein, this may allow us to investigate novel treatment for TRAPS. This would allow us to study mutations as individual entities and from this treatments could be tailored to the patient depending on the specific mutation they have.

There are many variable to consider with this technique. The set up for this method required extensive testing to define its optimal conditions and limitations. We had limited success using TNFR 1 and in order for this testing to progress any further simplification of the protein may be required.


I would like to thank Patrick Tighe for all his support and guidance throughout this project. I would also like to thank Paul Radford for supplying the TNFR 1 and Sue Bainbridge for assisting me with protocols and other laboratory work.