I would like to suggest you to troubleshoot insolution digestion for your samples, insolution digestion is important for uniform elution and ion mobility.

Dear Yogesh,

thank you for your answer. I know in solution digestion followed by LC/MS/MS analysis could be an alternative strategy for characterizing the ubiquitinome however presently my problem is how to improve the fragmentation of the ubiquitinated peptides for MS/MS analysis. Thank you again for your interest and good luck with your research!

Hi Massimo,

I'm curious about what you mean by poor quality fragmentation - is the resulting spectrum too simple or too complex?

If you see the ubiquitin PMF, and the substrate PMF and the modified substrate peptide mass without having to allow too many miscleavages the digestion seems fine.

I've looked at fragmentation of SUMOylated peptides - SUMO leaves a longer piece of itself on the substrate peptide after tryptic digestion. The fragmentation spectra are then complicated by the "Y"- shape of the peptide.

It effectively has two n-termini and one c-terminus. If your fragmentation spectrum is too full this may be the reason. Then you have too annotate the b-ions for the LRGG and internal fragments by hand. The modifier also fragments like any other peptide and leads to other shifted ion series for the substrate peptide. These will all result in a poor match in a standard MSMS search, but lots of ions that can be assigned manually and make sense.

If your MSMS has too few peaks, then your peptide just isn't fragmenting under the "normal" PSD parameters.

So two questions:

1. Do you collect the LIFT spectra on the selected modified peptides by hand? If not, it could be that the laser energy boost set up for PSD of normal peptides is insufficient. Try adjusting the laser intensity by hand.

2. Does the Autoflex offer the possibility of CID as well? My experience was good with CID-based fragmentation.

Hi Thomas;

thank you for your answer. Well my problem is that peptides showing putative gly-gly modification do not fragment during LIFT experiment thus I have problems to assign the ubiquitination site.

I have tried to increase the intensity of the laser, but the peptides do not fragment the same. Yes Autoflex III allows CID analysis, I will try it! thank you, I will let you know if fragmentation is improved.

Hi Massimo,

One other question. When you look at the MALDI spectra (the PMF's) in which the target protein gets identified and you look at the mass deviations, the hypothetical Gly-gly peptide doesn't have an exceptionally high deviation does it? It shouldn't fragment that differently from a normal peptide. If you have a non-peptide being squeezed into matching the gly-gly peptide that could explain the non-fragmentation, and it would most likely show up as being further from the predicted peptide mass that it's "correct" neighbors are.

One last remark about the fragmentation - have you checked the MSMS by hand? Perhaps the S/N ratio is simply poorer. Did the other peptides of the protein and ubiquitin give you good MSMS?

Dear Colby,

all your considerations are correct. Generally the ppm deviation of the predicted ubiquitinated peptides is closer to that not having the modification so I think they are peptides. I was able to obtain fragmentation of other Ub-peptides and looking at the peptides that were not properly fragmented I must recognize that the S/N was quite poor thus I think this is the main reason that i could not obtain good MS/MS spectra.

I have a second question for you: when I perform PMF it happens that I can obtain protein ID. Then as usually I select the peptides that could be associated to the predicted protein (taking care about those having good S/N) then I perform MS/MS analysis. The first thing I do then is to perform a combined analysis, by mens of biotools, of the MS/MS spectra and to evaluate if some of the peptides can be assigned univocally to the protein identified by the PMF. This is successful sometimes but not always. The second option is to perform a combined analysis of the MS and the MS/MS spectra and to check how many MS/MS spectra can be assigned to the protein predicted by the PMF analysis. In general by means of this approach I have the situation in which for a certain proteins I have (for example) a 10% MS coverage plus 3% MS/MS. Do you think this strategy is good to improve the confidence about protein ID? thank you in advance for your assistance!

Massimo

Hi Massimo,

Are you trying to get a good ID, or are you trying to find the ubiquitinylation site (or both)?

With regard to your first statements about poor s/n. Did you mean that the precursors you selected for fragmentation were quite weak in terms of s/n in the PMF, or that the resulting fragmentation spectra had poor s/n? Indeed, a weak precursor will probably not yield very good fragmentation. Your processing might eliminate a lot of fragment ions.

As to the second question, we almost always use the second strategy you mention. In our case the acquisition is heavily automated. We use both the PMF and the combined PFFs to search independently on the same database. If both kinds of data point significantly to the same identification then we have high confidence.

We typically don't use biotools in automatic mode, so I'm not familiar with its criteria for scoring and significance, or indeed how it combines PMF and PFF data.

But just to return to your ubiquitinylation samples for a moment, in that case I really would use biotools to attempt a manual check of the possible branched peptides. If you have weak fragmentation you can't really expect to get a significant score for the modified precursor fragmentation spectrum in mascot. But if your PMF is well explained by ubiquitin peptides plus those from your identified protein and you have a few PFF matches to non-modified peptides, then you can look manually at the fragmentation spectra for the modified peptides - even at the raw data - and try to correlate ions to those predicted by biotools. The complete burden of proof of protein modification doesn't lie with the modified peptides, its mostly been taken care of by the presence of ubiquitin and unmodified target protein peptides in the gel spot. In my opinion you're then justified in lowering your S/N threshold (for example looking at raw data) in order to wring some reasonable hypotheses out of the remaining data.

Did I understand that you see both the modified and unmodified versions of a given peptide? That would imply multiple sites. Then one might also expect to see the same target protein appearing higher up in a multiply ubiquitinylated form. You do see the masses for ubiquitin peptides in all of your suspected ubiquitinylated protein spots, right?

Are the ubiquitylated tryptic peptides more basic? In other words, are these peptides rich in arginine (Arg), histidine (His) and lysine (Lys)? If so, though the basic peptides are known to ionize well since they can be readily protonated, they do not undergo proper fragmentation (MS/MS), because mostly the proton(s) are bound to the sidechain group of Lys, Arg and His. For obtaining a good fragment ion yield in MS/MS spectrum, it is essential to have good amount of backbone protonation. Therefore, if the ubiquitylated tryptic peptides (that you obtain from in-gel digestion) possess several basic amino acids, then you may try doing acetylation reaction. This can be done using acetic anhydride or acetyl chloride. Then, the basic moieties of Arg and Lys would be modified and their basicity property would get reduced. Hence, the chances of backbone protonation is enhanced, thereby better fragmentation (MS/MS) can be expected. A protocol for acetylation of peptides can be noted in www.ionsource.com. Let me know, when you see improvement in the fragmentation.

Hi Varatharajan,

thank you for your suggestions! I will try and let you know.

Regards,

Massimo