Hi there,

One can't reply formely to your question! The major issue is that you don't know if the complex is formed in the first instance, and then if it's present when diluted and finally if it's in its native form. What I would suggest first is to try to characterize the formation of the native complex (coelution of protein and DNA during SEC experiment and increasing of protein MW in presence of DNA would be a good indication).

What you mean by that is confirming whether DNA and Protein make the complex or not is first thing to check?

If that is right, I've already confirm the complex formation by my DNA and Protein.

Thanks for the comments.

My point is that precipitation is not proving specific interaction. Therefore my question is how have you confirmed that your protein and DNA form a complex?

I've done the ITC, EMSA and size exclusion. All those results show that they make a complex

OK then you know how to form the soluble complex from your analytical SEC experiment (protein and DNA concentrations, buffer composition). So I suggest to start from these diluted conditions and to scale up volume:

either to concentrate the sample slowly (up to the point of possible precipitation) and to pass it on gel filtration to get rid of extra DNA (or dialysis with a cutoff between complex MW and DNA)

or to concentrate the complex on affinity column (if protein is tagged) and to pass eluate on gel filtration to get rid of competitor (or dialysis).

Hi,

Since you want to do NMR, it is important to have a stable complex and at relatively high concentration (0.3 to 0.5 mM). Dominique made some very good suggestions. You may also consider the following:

1) Try buffer screen to find good conditions that stabilize your complex. Screen for pH and salt additives. E.g. Mg ions tend to stabilize protein-DNA complexes. In our lab we use 96-well plates to do such screens.

2) You have done ITC, so I guess you know the affinity of you complex. If it is microM or less, you are okay. Also do you see a nice 1:1 binding stoichiometry? If not then something funky is going on. May be you are missing some essential component.

Goodluck!

Hi

We encountered something similar when preparing protein:DNA complexes at NMR concentrations (see PMID 16813837) but found that preparing them under dilute conditions, adding the DNA drop-wise on ice, and only then then concentrating kept them in solution.  Our theory at the time was that due to the fact most DNA-binding proteins can also bind DNA non-specifically, we were getting non-specific, non-stoichiometric complexes forming competitively at the higher concentrations.

Good luck!

Hi,

Depending on your experiment you may be able to manage with a protein concentration as low as 50 uM, although you will need to do a lot of NMR scans though, thus increasing your experiment time.  A higher concentration of protein is always preferred by an NMR spectroscopist but this is offset by high protein concentrations inducing protein precipitation.  Assuming you are monitoring unlabelled or 15N labelled protein by 1D proton or HSQC experiments; I would first run a spectrum of the protein alone, at say 150 uM, see what the spectrum looks like.  Then run an identical experiment with double the protein concentration.  This will tell you if the protein is stable or is precipitating on its own at these concentrations.  Then as recommended above, titrate DNA into the sample in a step wise manner, running NMR experiments at each point.  This will tell you a lot about the binding dynamics and equilibrium.  NMR spectroscopy, with careful analysis will quantify any loss of protein to precipitation, by a loss of signal intensity (line broadening) across the protein, and can do the same for analysing the DNA state.  DNA is of course strongly ionic, so may induce protein precipitation at high concentrations.  There could be a chance that you are measuring another binding process in your ITC experiments, e.g. precipitation.  

Good luck!

Hi,

I agree with a lot that's been said.  I've done a number of protein:DNA and protein:GAG titrations in the past and this is what I've found.

Often the complexes are much less soluble than the free proteins - and they may be more soluble at different pHs or salt concentrations.  It is therefore important to think about what you want to get out of the complex as the final concentration you require will be very dependent on what NMR experiments you want to do on the complex.  If you just want to do chemical shift mapping then for a 16 kDa well behaved protein on a spectrometer with a cryoprobe, you might be able to go as low as 10uM.

As Janice Bramham has already said, forming the complex slowly can be a very good idea. I like to form the complex slowly with continuous stirring - I do this by making micro-stirrer bars by cutting small fragments of wire (paper clips) and sealing them in glass by dropping them into the ends of glass Pasteur Pipettes and sealing either end by melting.  This then means that you can stir the sample at a reasonable rate whilst adding the ligand.  Add the ligand in small (1 ul) aliquots.  Think about and check the pH during the formation of the complex.  Also consider the temperature you use (both for the experiments and complex formation) - I've had a number of instances when I've formed the complex cold, only for it to precipitate when I put it in the magnet at a 5K higher temperature.

Concentrating the complex can be a very hit or miss thing in my experience, but if you need to get to higher concentrations than you can use for making the complex then you don't have too much choice - you need to think about the effect your method of concentration will have on the equilibrium in your complex.

Hope this helps

If the protein is well-behaved and you can have a 15N labelled sample i would suggest doing an ITC, with the labelled sample and stopping after it his reaching saturation.

You can control the volume of injection, temperature, stirring rate... Also formation of aggregates would give "characteristic-abnormal " peaks.

This depends on the calorimeter you have. Micro-type ones would allow you to titrate 300 or 350 uL of protein, roughly the same amount as needed for a shigemi.

This does not invalidate the  NMR controls at different concentrations, as formerly suggested.

good luck