When is LC-MS/MS peptide mapping worth paying for instead of a plain intact mass?

I have been quoted roughly four times the price of a purity-plus-identity panel for something described as "tryptic digest with LC-MS/MS peptide mapping and sequence coverage report". I do not want to pay four times for a longer PDF.

My understanding is that intact mass tells me the whole molecule weighs the right amount, and mapping cuts it into pieces and weighs each piece, so mapping can localise a defect that intact mass can only detect as a total. That much makes sense. What I cannot judge is when that localisation is worth money. If the intact mass is right to within 0.5 Da and the HPLC says 99.2%, what specific realistic defect is still hiding that mapping would find?

Concretely, I am looking at semaglutide from a source I have not used before. If someone can tell me what the tryptic map of semaglutide is actually supposed to look like — what fragments, what masses — I would find that far more useful than a general explanation, because then I can read a report rather than trust one.

Mapping earns its money on exactly one class of defect: something that is present at a level intact mass cannot resolve, or that intact mass can see but not attribute. For an acylated GLP-1 analogue there is a specific, common, potency-relevant example, and the semaglutide tryptic map shows it beautifully.

The semaglutide tryptic map

Trypsin cleaves C-terminal to Lys and Arg, but not when the residue is followed by Pro and not when the lysine side-chain amine is acylated. Semaglutide is GLP-1(7-37) with Aib at 8, Arg at 34, and the AEEA-AEEA-gamma-Glu-C18-diacid side chain on Lys26. So Lys26 is blocked and the only cleavage sites are Arg34 and Arg36:

X is Aib, and the asterisk marks the acylated Lys26. Residue coverage is 31 of 31 on paper, 30 of 31 in practice because a free glycine at 75 Da is below the low-mass cutoff of any normal method.

Why that map is diagnostic

Now suppose the acylation step failed on some fraction of the batch. The des-acyl peptide has a free Lys26, so trypsin cleaves there, and the map changes qualitatively rather than quantitatively:

Two new peptides appear at 2110.98 and 1032.58 that have no business being in a clean digest. That is a much easier thing to see at low abundance than a 715 Da shift on a 4111 Da intact mass, because you are now looking for the presence of a peak rather than for a small satellite next to a large one. The side-chain addition is 715.43 Da, so des-acyl semaglutide has a monoisotopic mass of 3395.69 — visible on an intact scan if it is at a few percent, invisible at 0.3%, and always visible in the digest.

The higher-resolution version: Glu-C

Trypsin is a poor first choice here because T1 carries 93% of the molecule in one 3841 Da fragment. Localising anything within it means MS/MS on a large multiply charged precursor, which works but is not free. Endoproteinase Glu-C in phosphate buffer cleaves after both Glu and Asp, and on semaglutide that gives five clean fragments:

Now the entire side chain question lives in one 1317.7 Da hexapeptide, and every side-chain defect is a clean mass difference on a small, well-behaved peptide:

• Correct: 1317.73
• Des-acyl (whole side chain missing): 602.30, a difference of 715.43
• One AEEA spacer missing: 1172.65, a difference of 145.07
• gamma-Glu missing: 1188.69, a difference of 129.04

Each of those is a distinct molecule with plausible receptor binding and an altered albumin affinity, which means altered duration of action. None of them is reliably visible on an intact scan at low abundance. This is the answer to your question about what is still hiding at 99.2% purity and a correct intact mass: side-chain variants that co-elute closely with the parent and carry a small enough mass difference to hide in the isotope envelope, or that resolve on HPLC but were integrated as part of the main peak.

When it is not worth it

Be honest about the cases where mapping adds nothing you will act on:

• A short, unmodified peptide. If the whole molecule is 10 residues with no side-chain modification and no disulfide, intact mass plus a good HPLC purity is close to a complete characterisation.
• A repeat purchase from a source you have already mapped. Map once to establish the manufacturer makes the right molecule; after that, purity and content per batch is the sensible ongoing spend.
• When you have not yet bought a content assay. If your budget is one test, quantification tells you more about whether you were defrauded than a coverage map does. Mapping tells you what the molecule is; content tells you how much of it there is, and shortfalls are far more common than wrong molecules.

My ordering for a new source of an acylated analogue: content first, purity second, map third. For a source you already trust, content every batch and nothing else unless something looks wrong.

edited 3 Sept 2025 by n_takahashi — added a caveat about sampling

Adding the digest artefact problem, because it determines which findings on a map you should believe.

A tryptic digest is a chemical reaction run for hours at 37 degrees, usually in ammonium bicarbonate or Tris at pH 7.8 to 8.0. Those are close to optimal conditions for manufacturing two of the impurities the map is supposed to detect:

• Deamidation. Asn deamidation is base-catalysed and its half-life at pH 8 and 37 degrees is hours to days depending on the following residue. An 18-hour digest can generate a percent or more of deamidated peptide that was not in the vial. Any "deamidation detected" finding from a long, high-pH digest is suspect.
• Aspartimide and succinimide chemistry. Same pH range, same direction.

How a competent lab handles it: digest at pH 6.5 to 7.0 rather than 8, use a high enzyme-to-substrate ratio (1:20 rather than 1:100) to shorten the incubation to two hours or less, keep the sample cold between steps, and run a mock digest of a reference standard in parallel so that any artefact appears in both. If the report does not state the digest pH and duration, you cannot evaluate a deamidation result from it at all.

Two other artefacts to know:

• Missed cleavages. Incomplete digestion gives peptides spanning an uncleaved site — for semaglutide, a 7-36 fragment from a missed cleavage at Arg34. These are normal, they should be assigned as missed-cleavage products, and a map with many of them indicates under-digestion and therefore unreliable coverage.
• Non-specific cleavage. Commercial trypsin carries a little chymotryptic activity unless it has been TPCK-treated. Semi-specific peptides in a map are usually this, not a real backbone defect.

The practical filter: a finding is credible if it is a presence that cannot be made by the digest (a des-acyl fragment, a deletion, a wrong mass on the modified-Lys peptide) and less credible if it is a modification the digest chemistry could have created (deamidation, dehydration). Ask for the digest conditions and the parallel-standard control before you accept the second category.

On the disulfide part of the question that has not come up: for semaglutide and tirzepatide it does not exist, and it is worth saying so because people import the concept from antibody and insulin work.

Neither peptide contains cysteine. Semaglutide is the GLP-1(7-37) backbone with Aib8, Arg34 and an acylated Lys26; tirzepatide is a 39-residue GIP-based sequence with two Aib, an acylated Lys20 and a C-terminal amide. No Cys, no disulfides, no free thiol, nothing to scramble. So a "disulfide bond confirmation" line on a COA for either of these is either a template artefact or a misunderstanding.

Where it does matter is the cyclic and bridged peptides people buy alongside them — somatostatin analogues, oxytocin, vasopressin analogues, insulin and its analogues, and anything with two or more Cys. For those, the mapping approach is:

1. Digest non-reduced at low pH. pH 6.0 to 6.5 with a Lys-C or trypsin digest, because thiol-disulfide exchange is base-catalysed and a pH 8 digest scrambles the very bonds you are trying to locate.
2. Look for the linked species. A disulfide-linked pair of fragments has a mass of the sum of the two peptides minus 2.0157 Da (two hydrogens). That mass, at a retention time where neither free peptide elutes, is the evidence.
3. Confirm by reduction. Reduce with DTT or TCEP and re-run: the linked peak disappears and the two free peptides appear, each 1.0079 Da heavier than its contribution to the linked mass.

The important limitation, which mirrors the theme of this whole tag: a scrambled disulfide is isobaric with the native one. If a peptide with Cys at positions 3, 7, 11 and 15 has bonds 3-11 and 7-15 instead of 3-7 and 11-15, the intact mass is identical. Only the digest, with fragments that separate the pairings, can tell them apart — and only if the digest sites fall between the cysteines, which for a tightly bridged peptide they often do not. In that case you are down to partial reduction with sequential alkylation, which is a real project rather than a routine test.