Hi,

I'd troubleshoot by increasing protein concentration to 0.2 - 0.5 mg/ml, although 0.1 mg/ml should be enough in most cases with your volume. Sometimes the SYPRO orange is a problem - I suppose your stock is still nice and orange, has been stored light-protected, and everything dissolves if you make for example a 10x dilution in 50% or so DMSO? I've used final concentrations of 1 - 10x, depending on the age of the stock. Typically 1x has been fine in a 50 µl reaction, but for a lower volume I've usually increased the concentration. I've heard people using as high as 50x. Something else you could try is swapping the tubes and caps to something else. I've typically had the best results using a 96-well plate and sealing tape suited for such an assay (purchased from Applied Biosystems, but definitely worth the price). This depends of course whether your instrument reads the signal from above or below.

You experimental set up looks good to me, so I guess I should ask is your FRET channel suitable for the experiment? Everything depends on the excitation and emission wavelengths, anyway, and your signal is unusually low. Is anyone else in your lab having the same problems as you? I typically run everything using a TAMRA filter, which is close enough for the use for SYPRO orange. With the TAMRA filter, at 1x SYPRO orange and 0.1 mg/ml protein in a 50 µl reaction, I typically get a starting signal at around 5,000 - 20,000 RU, and after melting the peak reaches up to 100,000 - 250,000 RU (using an ancient Applied Biosystems 7500 RT-PCR machine).

You also may want to refer to the manual of the instrument to see if you can run any control experiments to assess the condition of the instrument at different wavelengths.

For control DSF you might want to try other proteins also, like BSA, ovalbumin, or something similar. Lysozyme is relatively small and as such has a small expected binding surface for SYPRO orange. Try a few proteins with known, but different Tm's to check whether you see *any* increasing signal in the expected melting region(s).

Also, ligands: make sure that these don't interfere with the assay by binding to SYPRO orange or quenching its signal otherwise at the given conditions. I guess this is a problem with relatively hydrophobic compounds at concentrations similar to the fluorophore itself.

And finally... well I suppose this is kind of futile to ask but, I suppose you have checked that your protein is folded and not aggregated using CD, IR, DLS, SEC and/or other methods? If your DSF doesn't work, you could always try other methods like DSC (assuming you have enough protein) or CD (assuming you can substitute your buffer and ligands with something that don't absorb or have a massive CD signal that overwhelms the protein).

Hope this helps,

Arne

Hi,

I'd troubleshoot by increasing protein concentration to 0.2 - 0.5 mg/ml, although 0.1 mg/ml should be enough in most cases with your volume. Sometimes the SYPRO orange is a problem - I suppose your stock is still nice and orange, has been stored light-protected, and everything dissolves if you make for example a 10x dilution in 50% or so DMSO? I've used final concentrations of 1 - 10x, depending on the age of the stock. Typically 1x has been fine in a 50 µl reaction, but for a lower volume I've usually increased the concentration. I've heard people using as high as 50x. Something else you could try is swapping the tubes and caps to something else. I've typically had the best results using a 96-well plate and sealing tape suited for such an assay (purchased from Applied Biosystems, but definitely worth the price). This depends of course whether your instrument reads the signal from above or below.

You experimental set up looks good to me, so I guess I should ask is your FRET channel suitable for the experiment? Everything depends on the excitation and emission wavelengths, anyway, and your signal is unusually low. Is anyone else in your lab having the same problems as you? I typically run everything using a TAMRA filter, which is close enough for the use for SYPRO orange. With the TAMRA filter, at 1x SYPRO orange and 0.1 mg/ml protein in a 50 µl reaction, I typically get a starting signal at around 5,000 - 20,000 RU, and after melting the peak reaches up to 100,000 - 250,000 RU (using an ancient Applied Biosystems 7500 RT-PCR machine).

You also may want to refer to the manual of the instrument to see if you can run any control experiments to assess the condition of the instrument at different wavelengths.

For control DSF you might want to try other proteins also, like BSA, ovalbumin, or something similar. Lysozyme is relatively small and as such has a small expected binding surface for SYPRO orange. Try a few proteins with known, but different Tm's to check whether you see *any* increasing signal in the expected melting region(s).

Also, ligands: make sure that these don't interfere with the assay by binding to SYPRO orange or quenching its signal otherwise at the given conditions. I guess this is a problem with relatively hydrophobic compounds at concentrations similar to the fluorophore itself.

And finally... well I suppose this is kind of futile to ask but, I suppose you have checked that your protein is folded and not aggregated using CD, IR, DLS, SEC and/or other methods? If your DSF doesn't work, you could always try other methods like DSC (assuming you have enough protein) or CD (assuming you can substitute your buffer and ligands with something that don't absorb or have a massive CD signal that overwhelms the protein).

Hope this helps,

Arne

I would agree that checking the wavelengths would be ideal. Generally the ROX filter set works well for Sypro Orange: the manufacturer's website for the instrument that you have isn't great, but it looks like you might have filters for that?

Sypro Orange usually excites best at about 490-500 nm, and emits at 590-600 nm. The excitation and emission spectra are quite broad, so it's possible that your FRET filters aren't really picking up the fluorescence well.

I think that you are on the right track by trying a positive control with BSA or a similar protein. It's best to troubleshoot with something that you know will work, and then you can move over to your unknown protein.

Good luck,

Nic

Hi,

I can only add a few comments to the suggestions provided by Arne and Nicholas:

1. You should verify that your protein does not aggregate and precipitate as a result of heating

2. Some proteins are resistant to heat denaturation, I had instances in which I had to go above 100C (verified by DSC)

3. What is really important is the AMOUNT (mass) of protein, not its concentration: the standard value of 75 ug/ml is for a 35 kDa protein, you will need to adjust and optimize for your specific protein.

4. One last comment, probably not applying in your case considering the low initial signal: some proteins are already in an "extended" state and don't show much signal during heating.

5. Did you try molecular rotors instead of SYPRO orange? Those are more sensitive to change in viscosity and could give you a signal when you don't see anything with SYPRO orange (due to a limited change in hydrophobic residues exposure, etc.).

Good luck.

Carlo