Can you post a view of the spectrum for us to have a look at?

Does your protein have a ligand, cofactor, or prosthetic group?

Hi,

Here are the files. Sample 008 is tagless version ~7mg/ml, 009 is Hist taged at ~5mg/ml. Sample buffer is 10mM Tris, 150mM NaCl, 0.2% sodium lauroyl sarcosinate which was used as blank and automatically subtracted. These are just preliminary raw data without repeats and smoothing applied. My protein doesnt have any known ligand and was purified with HIC chromatography to more than 80% purity (it is insoluble membrane associated protein). I was trying to find any clues about the negative ~230nm peaks in CD, but no alas. The only closely related information I have found:

Wustneck, R.; Buder, E.; Wetzel, R.; Hermel, H. The modification of the triple helical structure of gelatin in aqueous solution 1. The influence of anionic surfactants, pH-value, and temperature.

Colloid Polym. Sci. 1988, 266, 1061–1067.

"It has been reported that upon complete denaturation of gelatin, the positive

peak at 222 nm disappears completely while the negative band shifts to nearly 230 nm"

A few comments and possibilities:

-  The two spectra are nearly identical -  some minor differences 205 and maybe 210 nm, but you'd have a hard time making a big story out of that given the dominant (and consistent) major trough.  Many years ago my thesis work dealt with a somewhat similar CD spectrum -- at the time there was a debate that 'denatured' and 'randomly structured' were not the same states.  One problem with CD is that it's fantastic when substantial ordered (and well-understood) structures are present, which gives you the unique features of helices and well-ordered sheets.  But if your protein folds into a stable structure that doesn't have those features, the CD spectrum may be unusual and maybe not very helpful.

-  My approach to that situation during my thesis work was to try to denature the protein with guanidine HCl.  The idea is that if you can denature a structure (seeing a change in the CD as you increase denaturant concentration), then you must have had some sort of 'native' structure in the first place.  You could certainly try that.  In my case, there was a clear transition across a relative high but narrow guanidine concentration range. 

-  I'm going to assume you are working with a clone/expression system, since you mention His tags.  There is no reason to assume beforehand that expression in one system results in a native folded protein.  Especially with the his tags -- usually benign, but occasionally they participate in folding with the body of the protein, which ultimately would be very bad.  Another possibility could be to introduce a pinch of nickel salt, so force the His tags to self-fold, which would reduce the likelihood of them 'contaminating' your native fold.  The problem with that idea is that it could be "too late" in the sense that you may already have some stable structure and the nickel will not induce a structural rearrangement.  It should be easy enough to try, though.

-  Try changing to a 'simpler' buffer system.  It seems to me a good possibility that that you could be solublizing the protein in a sort of micelle (or micelle-like) state with your current buffer.  Simple phosphate buffer at pH 7.2, with 100 or 150mM NaCl might give you a different result.  If this is supposed to be a soluble protein, simpler buffers shouldn't be a problem.  You didn't mention what pH you are working at -- even that could have an effect.

Will be curious to know what new data look like.  Good luck.

Thank you for the input. Yes, the denaturing conditions are the first thing I wanted to try.  Which is actually what I am doing right now. To provide you with more details, the protein native function is to polymerize in association with membranes. Already confirmed by DLS and AFM I know my protein is present at a very broad distribution of sizes. From monomer to MDa. This might be a big trouble for the CD, as I read somewhere that CD doesnt work well with polymers.

I can exchange any component of the buffer except the detergent. I was using 0.1mm path length without any substantiation interference with far UV readings. With a very little CD experience I was just baffled with the looks of the spectrum.

The complication with membranes is definitely going to be part of the problem, along with the variability in size distribution that you are seeing in DLS.  In reality, the only way that CD will actually be useful is if all of the proteins are in a relatively homogeneous structural state -- from your DLS comments they probably are not.   I think the 'structured yet non-regular' idea is probably consistent with the CD data you are getting --> the spectra would be an average of a bunch of different folding states among individual protein molecules.  Could be that if you can achieve a more homogeneous distribution of particles, detected by DLS, that you may indeed start to get a clearer picture in the CD as well.

Since your protein is aggregated, and aggregates often form amyloid-like quaternary structures, the 230 nm feature may be an indication of such a structure. Try looking at CD spectra from research on amyloids to see if there is anything like this.

Considering that we are dealing with a membrane protein, have you thought about some lectin or other polysaccharides, covalently bound to the protein ? They often have spectra in far UV CD with strong minima around 225 - 230 nm.

Hi,

I simply don't know about your protein, but as you are saying you are getting a strong negative peak around 230 nm and i guess you have taken care of the nickel, other any drug or polysaccharide, then I can surely predict that your protein is forming higher order structure, oligomers and most likely it is forming fibrils or any filament type of things, so I suggest to lower the concentration and repeat the same experiments and take care of the oligomerization. And your DLS data also suggested the oligomerization formation.

Please follow the link below 

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3317744/

thanks

Just an update. Increasing Urea and GuHCl is significantly decreasing ellipticity up to 220 nm. The 230 nm negative peak remains, and is only decreased to half when boiling sample in 8M urea, which is very interesting indeed. Will now try to find answer whats behind all this.

Everything  suggests that in solution there is something different from the protein. The two denaturants, in almost all cases, destroy the structural organization of many biological macromolecules. As the technique sees the molecular asymmetry, in the solution there is something that yet possess asymmetry/dissymmetry, that is, a chiral center, despite the loss of organization due to the denaturation.

I had also experienced such data. I am sending you two of my publications, that may help you. check fig.1. in TFE paper and 4th Fig of IJBM paper.

Dear Priyankar,

At alkaline pH, globular proteins denature but almost never completely. They often form intermediate with residual nucleation centers which often tend to interact generating aggregation. I saw the two papers, and I largely agrees with the findings on the presence of polyproline coils. These coils, if the sequence is suitably hydrophobic, are also formed by increasing the temperature. However, the permanence of organized structure, found by Matej,  in the presence of very high concentrations of chaotropic agents is surprising. Similar phenomena take place also in the IDPs that under certain conditions tend to collapse into compact globules, not organized but dynamically rich in fluctuating hydrogen bonds, and often also form chimeras with the presence of coils of polyproline.

Hello,

I just wanted to get back and share my revelations. The negative peak was the result of "light leaking" in the cell holder, affecting HV response on PMT. Therefore I could not recognize, the Sarkosyl was interfering up to ~220nm, thus creating a negative peak at 230nm. So nothing related to protein spectrum. Thats about precision in engineering nowadays. Hope this information can help to anybody struggling with CD.

Well, I told you that in solution, there was something else different from the protein. "Light leaking" is precisely what was different in the solution. I imagine it was a puzzle and it is good that you came out of. Thank you for sharing the solution of the problem.

Good luck with your research.

Hello Matej,

I also have found a strong peak at 230 nm in a protein CD spectra, and reading your comments I saw you experienced it earlier, could I have your opinion about it? 

Hello Eliane,

If you can describe in detail what were you measuring and show us the result, I can try to help.

Hi,

The far UV CD spectrum of alpha-chymotrypsin has a large -ve feature at 230nm (this is seen in the literature - Manavalan et al 1983 - and we have measured it more recently at our site).  The beta sheet in this protein is quite distorted and this 'may' cause the unusual CD spectrum.

Also, antibodies produce CD spectra dominated by signal assigned to beta sheet (+ve near 200nm and -ve near 215nm).  They also show some relatively fine structure around 230-240nm.  We associate this with the disulfide bonds.

Does this help?

Hi,

thanks, if you have published your data, could you send me the reference?

Hi, I want to whether you figured out the negative peak at 230 nm. Can you share the data if you have published? I am facing the same issue.

Thanks

Kamal Malhotra Have a look at a paper we recently published on the 230 nm peak in CD spectra..

Article C-Terminal Truncated α-Synuclein Fibrils Contain Strongly Tw...

Thanks for sharing Aditya