Dear Fatin,

Although I'm not familiar with the BioRad machine, looking at your graphs (also your other question here on RG) my first question is: what did you plot on the vertical axis? It looks like the first derivative of the fluorescence signal? If that is true, can you re-phrase your question?

Klaus

Dear Klaus Doering , the y axis is the rfu or the raw fluorescence reading. the attached figure is my raw data not the derivatives. Ive lowered the concentration and prepare a new run using a fresh enzyme but the pattern persist. ive attached the slight optimization that i made. as you can see, the pattern is roughly similar, high pre transition rfu, a dip at 60 degree celsius and a small peak right after

Dear Fatin, do you have enough material (& access to a fluorescence spectrometer) to record spectra? Which excitation & detection wavelengths do you use on the BioRad machine?

Do you have an idea why the fluorescence intensities you show in your second graph don't change linearly with respect to the different concentrations of your sample?

Fatin Safiudin before I saw your question here I originally posted a response to your other question about this that you put on RG (and I have repeated my response below this one as I feel that it gives quite relevant context). I am now pretty confident that the reason you are getting these results is that the protocol which you are using is designed for use with a fluorescent marker dye such as Sypro Orange. Furthermore, I am assuming that you are working off the thermal shift assay protocol which Biorad provide for the CFX series (see https://www.bio-rad.com/webroot/web/pdf/lsr/literature/Bulletin_7180.pdf), but can you give a link to the exact protocol which you are using? I can find no protocol online mentioning the CFX96 which is label free.

The Biorad CFX96 can only give a minimum excitation wavelength of 450 nm, which is nowhere near the wavelength that will cause peak excitation of tryptophan/tyrosine/phenylalanine residues (~280 nm). Therefore, I am pretty confident that what you are measuring is not Trp/Tyr fluorescence and any results that you are getting are meaningless in the context of a thermal shift assay as other components of the protein etc will be reacting to the excitation. I suggest that you either obtain a dye such as Sypro Orange which is suitable for the filter range of your QPCR machine or find an alternative method for a thermal shift assay. The only truly label free thermal shift assays to my knowledge are those which are offered by Nanotemper via their series of dedicated machines. They do give excellent results and are very rapid (results in less than 5 minutes), but the up front costs are very expensive compared to using a QPCR machine that can also be used for other purposes.

Would you agree with my reasoning, Klaus Doering?

--------(Response to other question)--------

I am assuming that you used a ~280 nm excitation filter to excite the tryptophan and tyrosine residues in your protein, but what emission filter setting did you use? Be aware that excitation at 280 nm will also react with a lot of other molecules other than Trp/Tyr and giving off fluorescent emissions at many other wavelengths than Trp/Tyr residues do. Therefore, you need to use a filter which cuts the recorded signal to read that at ~340 nm to only capture the Trp/Try fluorescence. I had the misfortune once to try reading intrinsic Trp/Tyr fluoresence of protein products with platereader and the data seemed completely nonsensical every time I tried it and I eventually gave up. It wasn't until months later when I had to open up the filter wheels (which nobody had opened in years) that I discovered that somebody had removed the 340 nm emission filter at some point (but not changed the settings in the software to reflect this), so where I thought I had only capturing the Trp/Tyr fluorescence signal, I was actually capturing all fluorescence signals.

What QPCR machine exactly did you use, as looking at current ones offered by Biorad none of them can use an excitation frequency of below 450 nm. Of course, they may have offered previous machines that did, but I can't see a reason why they would sell a QPCR machine that could go to such a low wavelength when no common fluorphores are excited by it (plus there would be many problems with other molecules like those in plastics being excited).

How have you processed the data? For data like this a standard way to make it easier to visualise is to take the first derivative of the fluorescence, as that can make it easier to see a definite peak.

Perform a series of dilutions of your protein sample to determine the concentration that gives the best signal-to-noise ratio. Sometimes lower concentrations can reduce background fluorescence.

Verify that the gain and other settings on your Bio-Rad CFX96 are properly configured for your specific assay. You might need to experiment with different settings to minimize background noise while maintaining sensitivity.

Use the instrument's software to correct for baseline fluorescence before the transition.

For your buffer, test different ionic strengths to find optimal conditions.

Samuel Davis , 100% agree, if the machine doesn't work at wavelengths relevant to the intrinsic (trp) fluorescence, i.e. exc. 280--300nm, em. 330--360nm, then all other observations & troubleshooting steps are wasted.

In which case either change the protocol (external dyes), or the instrument; I've run thermal denaturation of proteins in a standard cuvette-based fluorescence spectrometer equipped with a temperature controlled sample holder. Mind you, that likely requires more sample and is quite slow.

Fatin Safiudin you've had a lot of very insightful answers from several people to your question, perhaps you would like to give them the respect that they deserve by responding to their suggestions and updating on how your thermal shift assay is going?