What mass spectrometry tells you about a peptide

A mass spectrometer does three things: it converts molecules into gas-phase ions, separates those ions by mass-to-charge ratio (m/z), and counts them. The output is a spectrum of intensity against m/z. For a peptide of known sequence you can calculate the expected molecular weight from first principles, so the comparison between calculated and observed mass is a direct identity test.

Electrospray ionisation (ESI) sprays the sample from solution through a charged needle. It is gentle, couples directly to an HPLC outlet, and characteristically produces multiply-charged ions — the same molecule appearing at several m/z values carrying two, three or four protons.

MALDI (matrix-assisted laser desorption/ionisation) co-crystallises the sample with a light-absorbing matrix and fires a laser at it. It tends to produce singly-charged ions, tolerates salts better, and is usually paired with a time-of-flight analyser — hence MALDI-TOF.

Neither is inherently superior. ESI's advantage for a COA is that it can be run in line with the HPLC, which links the mass directly to the peak being integrated. That linkage is exactly the gap discussed in identity vs purity.

An ESI spectrum of a 3,000 Da peptide will not show a peak at 3,000. It shows a series: the singly-protonated ion at roughly m/z 3,001, the doubly-protonated at roughly 1,501, the triply-protonated at roughly 1,001, and so on. Each charge state adds the mass of a proton and divides by the charge.

The relationship is m/z = (M + nH) / n, where M is the neutral molecular mass, n the number of protons added, and H approximately 1.008.

Software deconvolutes this envelope back to a single neutral mass. A report may quote the deconvoluted mass, the observed m/z of one charge state, or both. If a spectrum shows a peak at an unexpectedly low m/z, check whether you are looking at a higher charge state before concluding anything is wrong.

Two different numbers are both correctly called the molecular weight, and a report should say which it is quoting.

• Monoisotopic mass uses the mass of the most abundant isotope of each element — carbon-12, hydrogen-1, and so on. It is the mass of the lightest peak in the isotope cluster.
• Average mass weights each element by its natural isotopic abundance. It is what you get from a standard molecular-weight calculator.

For a small peptide the two differ by well under a mass unit. For a large one the gap widens and comparing a monoisotopic observation against an average calculation can look like a failure when nothing is wrong. High-resolution instruments resolve the isotope pattern and quote monoisotopic; lower-resolution TOF work often quotes average.

Four limitations, in rough order of practical importance:

1. Sequence order. Any rearrangement of the same residues gives an identical mass. Resolving order requires MS/MS fragmentation.
2. Leucine versus isoleucine. Structural isomers, identical mass, not distinguishable even by MS/MS under standard conditions.
3. Quantity. Ion intensity depends on ionisation efficiency, which varies enormously between molecules. A tall peak does not mean an abundant species.
4. Stereochemistry. D- and L-amino acids have identical masses. A racemised residue is invisible.

The third is worth dwelling on because it is counter-intuitive: mass spectrometry is not quantitative without a matched internal standard. Reading relative abundances off a spectrum as if they were proportions is a common misreading.

Minimally: the theoretical mass, the observed mass, and a statement of which mass convention is used. Better: the spectrum itself, with the axis labelled and the relevant peak annotated. Better still: an indication that the sample analysed was the collected main HPLC peak rather than crude material.

An image of a spectrum with no axis labels, no annotation and no stated theoretical value carries essentially no information — it is decoration. That failure mode is covered further in red flags on a peptide COA.