As Ankur said, photodegradation can be a problem, especially with fluorescein and long light exposures. This will be demonstrated as a loss in total fluorescence intensity (parallel + 2 x perpendicular). I don't think it would appear as a reduction in anisotropy, however.

The first thing to check is whether the DNA is being degraded during the experiment. The reduced size of the oligonucleotide resulting from degradation could cause the fluorescein fluorescence anisotropy to decrease.

Another thing that may be happening is that the fluorescein is interacting with the DNA bases, reducing its mobility and elevating its anisotropy. When the protein binds, it could block this interaction, increasing the mobility of the fluorescein and decreasing its fluorescence anisotropy. It may be possible to avoid this by putting a longer DNA spacer between the fluorescein and the protein binding site.

It is indeed important to check for degradation during the experiment but it is also important to compare fluorescence intensities before and after binding. Are there maybe unexpected quenching effects? Time-resolved fluorescence intensity and anisotropy measurements may shed more light on this issue.

I'm unable to get consistent reading. Anisotropy with time was also not consistent 

Do you have a strong enough fluorescence signal from the fluorophore to make good anisotropy measurements? The polarizers reduce the detection sensitivity significantly. Stray light in the instrument causes a problem with anisotropy measurements if the signal from the sample isn't strong enough. Increasing the fluorophore concentration is the solution.

If you are having trouble because of photobleaching of the FAM fluorophore, consider switching to a fluorophore that is less prone to this problem.

Aside from instrumental or sample problems, anisotropy can decrease if:

1) the region including  the probe becomes more flexible.

2) the quantum yield of the probe increases.

Do you need controls to test your fluorephore. For example just DNA adding buffer instead protein

Hi Harsha,

Did you solve the problem?

I also came cross the same issue now. When I searched it on line and happened to see your posted question 3 years ago.

I don't think mine is due to photo degradation since I am using 96 well plate reader. Every sample was scanned once at the same time.

Also, according to Adam's comment on the interaction between fluorescein and DNA base. I know the FAM on the 5'-DNA end is far away from the protein binding site since I have the structure on the protein-DNA complex.

Again, Adam's comment on the quantity of the fluorescent signal, I do am working at very low concentration of labeled DNA though, since the expected binding affinity is quite high (nM range).

If any of you happen to see my problem here, could you please give me some suggestions?

Thanks

Susu

Sherwin Lehrer wrote that the anisotropy can decrease if the quantum yield of the probe increases. I do not think this correct.

For a two component system, bound (b) and free (f), the observed anisotropy is r is given by

r = (cf*Ff*rf + cb*Fb*rb)/(cf*Ff+cb*Fb)

where cf and cb are the concentrations of free and bound fluorophore, Ff and Fb are the florescence intensities of free and bound fluorophore, and, rf and rb are the anisotropies of the free and bound fluorophore.

If we consider rb > rf (as is the case in binding of a small fluorophore to a larger receptor), and If Fb/Ff > 1 (quantum yield of the probe increases upon binding) , r can never decrease (see equation above). In that case the variation of r will not reflect the true binding isotherm.

One reason why anisotropy can decrease upon binding is if the electronic nature of the fluorophore changes upon binding (charge transfer complex). That would change the excitation and emission transition dipoles and thus affect intrinsic anisotropy.

Consider: If the quantum yield increases because its lifetime increases, the probe has more time to depolarize and its anisotropy will decrease.

Sam Lehrer is absolutely correct! If the lifetime increases the anisotropy can certainly decrease. Remember that the lifetime of fluorescein probes are very dependent on the local pH - if the binding region provides a more basic environment the lifetime can certainly increase. Also, as Adam says, the local flexibility could affect the anisotropy. I studied a protein once with a single tryptophan and when it bound to another protein (without tryptophans) the anisotropy decreased since the local motion of the tryptophan increased so much. In this case we first checked the lifetime to be sure that it did not increase.

Thanks, Gautam, Sherwin and David, for all your comments.

The major message I got from all of you is the local pH / charge could affect the probe quantum yield. I am not from physics background, so I am not able to go too deep into quantum.

In my situation, the fluorescein is on the extremity of one DNA strand (The substrate is a double strand 36bp DNA). DNA by itself is always highly negatively charged. So I assume the probe now is more in a negatively charged (acidic) environment. Since I have the structure of this protein-DNA complex, I know the binding interface on protein is highly positively charged. But it only covers the central 26bp DNA. That means, the probe is still of 5bp DNA in distance (approximately 17 Å ) to positively charged protein. In this particular case, do you still think it could be possibly due to the charge change?

To David: When you suggested to check the lifetime. Do you mean test both the lifetime of fluorescein in DNA alone sample and life time of fluorescein in DNA-protein mixture sample. And then compare these two values to see the difference?

Thanks

Susu

Sherwin Lehrer's statement that the anisotropy can decrease if the quantum yield of the probe increases upon binding, was correct. I am sorry to have wrongly negated it. I made the statement without too much of a thought and I apologise. I guess the new anisotropy will depend on the new fluorescence life time as well as the new rotational correlation time.

No worries, Gautam. We wouldn't get to the point of fluorescence lifetime if you did not say out your questioning comments.

We learn better by hearing different voices since it leads to deeper discussion:)

Susu

Hi all,

Great topic for discussion.

I have few queries to Susu and Harsha:

1. How pure is the dye labelled DNA sample?

2. From which value to which value the anisotropy decreases?

3. Did you compare the anisotropies of free dye, dye labelled DNA and dye-DNA-protein complex?

4. For biophysical experiments samples should be prepared in buffer at a fixed pH. What is the experimental pH value? Did you measure the pH of samples after each addition of protein?

5. How does the fluorescence intensity profile looks like during the titration experiment?

If possible provide plots. Data speak better than human :).

The simplest experiment is to measure the fluorescence intensity because intensity is proportional to lifetime, with and without the perturbation, i.e., the protein, under identical conditions, when there is no instrumental reasons for intensity changes, like light scattering. In that case any change would be dependent on lifetime changes . A more direct experiment is to measure the lifetime change which is not affected by light scattering. From the lifetime change one could estimate the change in anisotropy for no change in flexibility from the Perrin equation:

ro/r = 1 + tau/theta, where ro is the rigid anisotropy, tau is the measured lifetime and theta is the correlation time (related to flexibility) of the probe.