Hi, 

I think you missed the following.

What you measure in a iTRAQ (or in general, an isobaric tagging) experiment is the ratio between different reports ions.

Example: consider 3 cell cultures experiment. Everything is the same but the isobaric tag, which are iTRAQ 1, 2 and 3.

At the end of your sample prep procedure, you will combine the 3 samples and analyze them.

During the CID/HCD, the tag will be cleaved. And you want it to be cleaved. Since the tags have different masses, they can be resolved and measured. Their concentration ratio will reflect the concentration ratio of the original peptide--> protein. You will have an MS/MS spectrum where the lower part (generally between 100 and 200 m/z) will be used for the isobaric tagging detection and quantification, while the rest of the spectra will be used for the identification of the peptide. Your MS/MS identification will NOT consider the tag. It will simply add it to the parent ion mass for the final calculation.

Therefore you will be able to to a relative quantification between the 3 experiments by simply measuring the ratio of the reporter ions in a particular MS/MS spectrum.

You might want to check fig 2 of  Ross PL, Huang YN, et al. (2004) "Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents". Mol. Cell. Proteomics 3: 1154–69

Thanks for your detailed explanation. This is what I am concerned about. From the figure 2, the author demonstrated the MS2 from a peptide with the sequence: TPHPALTEAK. The molecular weight of amino acid Threonine is 119.12, however, estimated from Fig.2A, the b1 ion for this peptide is around 250. Also I attached a file with the iTRAQ peptide mascot MS2 identification in my result, you can see the b1 ion m/z is matched to 434.26 in a 10plex iTRAQ reagent. It seems that mascot engine assumes the iTRAQ modification acts in a way similar to phosphorylation modification. Yes, there are some of the tag cleaved off the peptide(neutral loss?), some of them must retain on the peptide to meet the identification criteria.

One more question, still do not know whether the bond between reacitve group and amino acid will be cleaved, since it is structurally similar to the peptide bond.

In your dataset I see that you selected a fixed modification for the cysteine, for it is alkylated.

Think about the iTRAQ the same way you think about an alkylated cysteine.

The only difference is, your iTRAQ will be partly fragment and give you a reporter ion. And here is probably the part you missed. 

Take the example I wrote in my previous answer.

Let's say that a particular peptide is present in all 3 cultures. Let's assume that the isobaric tagging was 100% for all the samples.

When you fragment the m/z equal to that peptide (plu the various modifications) you will see a series of y and b ions (mostly carrying modifications) and the 3 reporter ions.

The quantifications is possible because the amount of fragmented reporter ion is directly proportional to the concentration of its parent peptide.

If the peptide hypothesized in our experiment has a ratio distribution of 100:50:25 between the 3 cultures, the same ratio distribution will be reflected on the reported ions abundances.

To answer your last question: yes it will. Because otherwise you will not obtain the reporter ion. But it will not be a 100 % fragmentation.

After All, if you accumulate 100 parent ions and do a CID, the outcome will be a series of fragments which will represent a thermodynamic distribution directly correlated to the CID energy you used.

I would suggest you do this 1 day experiment (although it could be a little expensive): take some BSA. Process it. Create some aliquots. Label them with your iTRAQ, in triplicates. Pool together and analyze them.

This will give you an idea of your sample prep quality. Moreover, you could use them as QC. You might even use them to adjust the normalized collision energy. Afterall, you are interested in getting some more reporter ion rather then ALL of the b and y ions, right?

I would like to add one more things.

You might want to verify the labelling reaction conditions. before, I said "Let's assume that the isobaric tagging was 100% for all the samples".

Even considering that you label samples form the same type of experiment, this assumption could not be completely true. The only way out, in my opinion, is to play with biological replicates.