Follow The White Rabbit – Interpreting Individual Peaks – MS2 spectra

' Armel Nicolas

Hi everyone, and welcome to a glorious New Year! 2017 was fun, wasn’t it? I for one so much look forward to 2018 being slightly more normal. Oh what the hell, who am I kidding, I’ll get some popcorn already.

Today, we will be talking about MS2 spectra, that is, the spectra resulting from the fragmentation of a precursor, which can be used to find out what said precursor is.

But first, let’s start with an apology. Reviewing my blogging activity from last year, I realised I am guilty of a filthy sin, which for someone supposedly trained as a teacher is near unforgivable. I have been talking theory and more theory but not actually showing the real stuff. So let’s get ourselves grounded in reality by looking at some real spectra!

Below you can see an example of the Total Ion Current (TIC) chromatogram (top) over the whole length of a run:

an example of the Total Ion Current (TIC) chromatogram (top) over the whole length of a run

And this a single MS1 spectrum #18618 acquired at 31.81 min:

single MS1 spectrum #18618 acquired at 31.81 min

Now, if we zoom in on the very narrow red rectangle, we will see an isotopic envelope that corresponds to a peptide:

isotopic envelope

In this experiment, we isolated for fragmentation a 1.6 Th window (in red) centred on the monoisotopic peak, at 515.30. As you can see, this window will also isolate the next peak in the envelope, as well as some “parasite” peaks we can see in the same area. These will also get fragmented, and there’s nothing we can do about it. Below, you can see the resulting MS2 spectrum:

MS2 spectrum

That is a lot of peaks! So how do we make sense of them?


The ideal case: a single precursor

In order to understand observed MS2 spectra, we will imagine a single peptide with a known sequence and describe what happens to it during fragmentation.

  • Precursor selection

We described in the previous entry why peptides exist as isotopic envelopes of multiple peaks in MS1. For MS2, we ideally want to only isolate and fragment one of the envelope’s peaks. Which one? Well, let us think for a moment. A monoisotopic peak has two advantages. First, it is for all but the longest peptides the most intense peak in the envelope, so it is faster to accumulate enough amounts of it to fragment. Second, it has a known isotopic composition. By contrast, the next peak is a collection of different peptides with 1 additional neutron located somewhere in one of the many atoms that make up the molecule. Thus, if we fragment this peak we will get twice more complex spectra, as each time we would expect a single peak from a monoisotopic precursor it will have a twin with one more neutron. For the next peak in the envelope, there will be three peaks to each monoisotopic fragment. Thus, what we really want is to isolate the monoisotopic peak only.

However, this is in practice not strictly possible. Too narrow isolation boundaries would result in too many losses, the typical isolation window has a width of between 1.4 and 2 Th. A peptide usually has a charge of at least +2 (most frequent charge) or higher; this means the distance between peaks of its envelope is 0.5, 0.333 or less Th. Thus, even in the most favourable case (2+charge), an isolation window centred on the monoisotopic peak will inevitably include also the next one.

  • Fragmentation

There are several methods of fragmentation available out there, each with their own specificity for fragments. Here we will only talk about CID/HCD, the one most commonly used.

Peptide fragments which are most relevant to identifications are those which break the peptide’s backbone, thus releasing fragments which can be identified as losses of a series of amino acids. These are called sequence ions. In the example below we are showing the accepted nomenclature for the sequence ions resulting from fragmentation of a 4 amino acids peptide:

accepted nomenclature for the sequence ions resulting from fragmentation of a 4 amino acids peptide

Thus, each peptide bond can fragment in three ways to generate a pair of a/x, b/y or c/z ions:

Peptide Fragment

Shown here with just one charge. As said before, a peptide usually has 2 or more positive charges. A fragment will only be detectable if it retains at least one positive charge.

Importantly, “sequence” ions do not form “sequentially” but sort of randomly (this is affected by sequence), which means it is difficult to predict relative fragment peak intensity, or which, of all of the possible fragments, will be identified. This means that full peptide de novo sequencing from MS2 spectra is rarely possible, and is the main reason why MS-based Proteomics is radically different from Genomics or Transcriptomics, for which true sequence can be unambiguously determined.

In CID/HCD, the main ions observed are b/y series ions (mostly y), as well as some a ions.

In addition, other forms of peptide fragmentation can occur:

  • Internal fragmentation can also generate fragments (e.g., here, R2-R3).
  • Internal fragments that form by the combination of a y-type cleavage to the left of an amino acid and an a-type cleavage to the right are called immonium ions. Their general structure is the same as that of the a1 ion shown above. Immonium ions are important because observing tells you that the corresponding amino acid must be part of the sequence of the peptide.
  • At low collision energies, in some contexts neutral loss of ammonium or water can occur.
  • At collision high energies, some of the amino acid side chains may fragment.


The reality: multiple precursors

In reality, very often MS2 spectra correspond to several mixed precursors. This means that they will be an even worse mess than what was described above, since they will correspond to the mix of two or more different fragmentation spectra.


The extreme case: Data Independent Acquisition (DIA)

In DIA, MS2 spectra are derived from fragmentation of all precursors within a few scores of wide isolation window (SWATH) collectively spanning the whole MS1 M/Z range, or even in some variants, of all precursors in the whole MS1 M/Z range. I won’t even try to describe the mess, just look at the number of precursors in the MS1 shown at the top of this entry and imagine fragmenting all these precursors together… Suffice to say, there are ways these extremely complex, mixed MS2 spectra can be made sense of to identify precursors, but we will discuss these another day.


In the next post, we will discuss how the MS2 spectra can be used to identify their parent peptide.


Point of View
Related Posts: