One possibility is that the native protein is aggregated. Rayleigh light scattering of aggregated protein causes an apparent increase in absorbance at shorter wavelengths. Denaturation with SDS would break up these aggregates.

I think @Adam's suggestion of scattering is a good one. Red-shifts of aromatic residues in folded globular proteins (compared to unfolded) in solution is generally only on the order of a few nanometers. It seems that the blue-shift you show is larger in the samples with larger absorbance. Do these samples show similar spectra at lower concentration?

Thank you for your suggestions!

I will definitely learn more about Rayleigh scattering. Although I always centrifuged samples before analysis, we noticed globular forms/aggregates in all of the samples, including native VLP without fused proteins with TEM (link: TEM photo).  So I guess that if blue-shift is made because of aggregates, it would also be seen in sample with native VLP?

Also, the absorbance of denatured proteins is seen past 300 nm, is that normal with denatured proteins?

At the moment I don't have spectra with lover concentrations, but I know that spectra at higher concentrations are the same, if that helps any.

http://s21.postimg.org/dftu59nsn/PVY2_142676_2_J_12.png

These are not spectra of (pure) proteins. In your native spectra, there is an anomalous high absorbance superimposed to all the spectra, probably due to scattering or/and other substance(s). In the denatured spectra, the situation is identical even if a bit better (I hope you have denatured by using concentrated guanidine HCl solutions). However, this seems to discard the aggregation hypothesis. In fact, a protein spectrum (native or denatured) must be almost flat from about 300 – 305 nm to higher wavelengths. In the tails of your spectra, as you can check, there is something absorbing up to about 350 nm. What is surprising is the total absence of protein morphology in your spectra. Is your instrumentation adequate?

I denatured with DTT, SDS and heating.

I can't find anything that would absorb at 250nm...

It is not the instrumentation, I got the same results on spectrophotometer and Nanodrop.

Your problem is the anomalous absorbance at 310 - 320. It should be flat. This suggests that ther'is something in the solution or sticked to protein. Do you know if the protein contains tyr and/or try? Check also the spectra of pure proteins in the literature to have a comparison and to check the characteristic spectral features that are missing in your spectra.