It happens due to the presence of DNA in your sample.I also had the same problem before. However, now I have changed the composition of my washing buffer (Phosphate buffer) a bit. I added 300mM NaCl and 30mM imidazole instead of 50mM NaCl and 300mM imidazole It worked like magic. Hopefully, it will also work for you.-----good luck 

It happens due to the presence of DNA in your sample.I also had the same problem before. However, now I have changed the composition of my washing buffer (Phosphate buffer) a bit. I added 300mM NaCl and 30mM imidazole instead of 50mM NaCl and 300mM imidazole It worked like magic. Hopefully, it will also work for you.-----good luck 

try low concentration of imidazole in wash buffer and high concentration in elution buffer. 10mM in wash buffer and 400mM in elution buffer. see if this helps.

Same as two previous answers; I have even tried 3.6 M NaCl, if your protein can withstand it; it won't make it come off the  Ni column

Hi Gustavo, probably what you are seeing is due to the higher imidazole concentration. Did you use the elution buffer as your blank? DNA complex are greater than 30 kDa, I don't know which cutoff membrane you used for the dialysis, but I don't think it is your problem.

Hi, people. Thanks for your answers.

Even if the bands I'm seeing are due to DNA presence, why are the two usual absorption bands gone? Could the DNA concentration be high enough to "hide" the protein absorption? The same goes for imidazole. Or is there some kind of interaction between the imidazole and protein that could shift these bands?

Hi, Lucas.

Yes, I have used the elution buffer as a blank. If I'm not wrong, I have used a 10 kDa cutoff membrane for dialysis. So, DNA complex should not be my problem, since it would continue to cause interference after dialysis.

Hey, do one thing for me: blank with water and check the absorption of elution buffer alone. You shall find something interesting. Imidazole powder is light-sensitive, so it is kept in brown bottles. However, i have seen the Imidazole with a pale color and also resulted in pale-colored solution (1M) instead of colorless solution with good quality Imidazole. This paled-colored solution also showed UV absorption. This is why, after dialysis with no Imidazole buffer, yours extra peaks disappeared. 

Hi Gustavo, 

I have a question for you. What kind of molar absorptivity are you seeing from this substance...or if you don't know it's concentration, what kind of absorbance difference is it giving above you regular sample? This would be very tell-tale in addressing your question. What does the absorption peak look like. You are saying that you are getting changes in the lower wavelengths as well. I would be interesting to see an absorbance spectrum where you have a standard concentration of the molecule you are trying to look at.

Can you do this other experiment? Using the same concentration of your material in the PBS  and Imidazole systems. In a dual beam spectrophotometer, look at the difference spectrum of the solvents alone over the wavelength range of 190-300 nm being sure you are using Quartz cells. If you have a fluorometer around, look at the Quantum Yield difference between the two solvents alone...there is a sensitivity reason of 4-6 fold reason for doing this...in the event that your Imidazole has a slight fluorescent contaminant. Now, with your samples at the same concentrations in the two solvents, do the same thing for both the absorbance and fluorescence spectra...just zero the fluorometer with you PBS system. What would be interesting to know is what the absorbance spectral differences look like and if there are like peaks of absorbance between your sample and the solvent.

Otherwise, Zaigham has the answer. One of the nifty things about our salt solutions is that we tend to think that because the bottle looks pretty and says it's pure, the fact is there are impurities which will show themselves up at higher concentration because of their absorbance characteristics. You might try a NEW bottle of Imidazole.

There is an old saying. "When all else fails, drop back and punt." Unless it is more than an idol curiosity, I would just make new solvents with brand new salts and see that the problem just disappears. You can always play" later.

DNA is not the problem in your sample. This in an effect of the imidazole at high concentration. As a consequence of that, when you dialised your sample you removed the imidazol and so that, you got the expected results.

I don't know which cutoff membrane you used for the dialysis, but I don't think it is your problem, because in mos of the standard dialysis membranas DNA remains retained.

Another important aspect is that proteins redox state couls change due to the pH value and you are changing from 8 to 6. Different redox states display differect UV-Vis features.

Are you working in a quality controlled environment? (GMP, GLP etc?) If not, I would strongly suggest to check your spectral photometer first. This is usually performed without much efford and often saves time and also your preciuos sample. If you don't have a routine procedure in place, you can easily prepare a reference solution containing only standard protein (or Vitamin B12) and make a wavelength scan and compare it with theoretical scans or scans from an alternative spectrometer.

I agree with Rosa Maria and Lucas, it is impossible to remove DNA by dialysis, you are removing imidazole and it change the UV abosorption.

Agree with answers above.  Dialysis does not remove nucleic acids that bound tightly to protein.  Had that problem before: had to use a 1M NaCl with buffer before eluting my protein. 

I think this link is useful, 

http://www.instanano.com/characterization/theoretical/concentration-calculation-from-uv-vis-absorbance/

Dear Gustavo:

Agree with previous comments. Proteins absorb at 280 nm. Absorbance at 260 nm is due to nucleic acids.