Sequence-Confirmed Synthesis — Why Mass Spec Alone Isn't Enough

Most peptide lot reports list one line under "Identity" — ESI-MS measured mass within 0.5 Da of theoretical. The implication, often unstated, is that the molecule in the vial matches the sequence on the label. For most research workflows that implication is good enough. For others it quietly fails. This Note explains where the failure happens, why mass spec alone cannot fix it, and the two add-on tests Lyochem ships when sequence-level assurance is non-negotiable.

What ESI-MS actually tells you

Reversed-phase HPLC separates the peptide from its synthesis impurities. The eluate is sprayed into the mass spectrometer (electrospray ionization, ESI), the instrument measures the mass-to-charge ratio of the ionised molecule, and the result is reported as an observed monoisotopic mass. A modern high-resolution instrument hits ±0.005 Da on a peptide under 5 kDa; a standard ion-trap or single-quad sits closer to ±0.5 Da, which is still tight enough to discriminate against most truncation impurities.

What the measurement actually proves: the molecule eluting at the labelled retention time has a total elemental composition that sums to the expected molecular formula. Within the resolution of the instrument, the labelled and observed masses agree.

What the measurement does not prove: that the residues are in the correct order, that two residues of the same residual mass are the ones the synthesis intended, that no permutation of the sequence is present that happens to share the formula. Mass is a property of the molecule's empirical formula. Sequence is a property of its connectivity.

The isobaric residue problem

The standard amino-acid building blocks divide into pairs and small sets that the mass spectrometer cannot distinguish at the residue level:

These intrinsic-residue ambiguities are well known. Less obvious — and the larger practical risk in routine peptide release — is the dipeptide isobaric problem: two-residue motifs that sum to the same nominal mass:

A recent (2025) review on therapeutic-protein peptide mapping reported that misidentifications "resulted from isobaric and near-isobaric dipeptides (e.g., SA vs. GT)" even in regulated high-resolution workflows ([Hopper et al., PMC12563827](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12563827/)). The substitution does not change the total measured mass of the peptide. It changes the bioactivity, the receptor binding, the immunogenicity profile — every property the bench scientist cares about — but the COA still reads "ESI-MS confirmed."

The class of failure that hides best inside an ESI-MS pass: a deletion-plus-insertion event in solid-phase synthesis that swaps one isobaric pair into another. A 25-mer with an SA at positions 8-9 that becomes a GT at the same positions through a coupling-cycle misfire has the same observed mass to the limit of any single-quad instrument.

The chain-order permutation problem

Even when no isobaric pair is involved, a chain-order permutation conserves total mass. A peptide labelled VPRSGE has the same empirical formula and observed mass as VPRGSE, VPSRGE, or PRSGVE — six permutations of the same six residues. ESI alone cannot tell them apart. RP-HPLC retention time will usually differ between the permutations (so a clean HPLC trace plus mass match is a stronger combined assertion than mass alone), but retention-time identity is only useful when a reference standard of the labelled sequence exists in the same run.

The actual sequence-verification tools

Two analytical methods cross the gap from "mass agrees" to "sequence proved." Each has a clear use case; in regulated peptide release, the two are complementary.

### LC-MS/MS fragmentation

The peptide is selected in the first mass filter, accelerated into a collision cell, and fragmented along its peptide-bond backbone. The fragment ions — usually b-type (N-terminal fragments) and y-type (C-terminal fragments) — are mass-analysed in the second filter. A complete b/y ion ladder reveals each successive residue mass; reading the ladder produces a sequence.

For a 20-mer the analysis runs in minutes per spectrum. Software walks the spectrum, scores candidate sequences against the observed ions, and reports a sequence match with a confidence metric. In regulated environments only USP- or vendor-validated software is acceptable; the academic peptide-identification scorers commonly cited in literature usually fail the validation bar ([Hopper et al., 2025](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12563827/)).

LC-MS/MS resolves both the isobaric-pair problem and the chain-order problem. Its remaining edge cases: - Some leucine-isoleucine assignments still require auxiliary information (immonium ion ratios, retention-time matching against a reference standard, or a separate Edman run on a co-eluting reference) - Modifications that block backbone fragmentation (N-terminal acetylation can suppress a-ions; D-amino acid substitutions need chiral-stationary-phase chromatography to resolve) - Highly cross-linked peptides (intra-chain disulfides) require a reduction step before fragmentation

### Edman degradation

Edman is the stepwise N-terminal sequencing method: phenyl isothiocyanate reacts with the free N-terminal α-amino group to form a phenylthiocarbamoyl-peptide, the N-terminal residue is cleaved off as a phenylthiohydantoin, identified by RP-HPLC against a reference, and the cycle repeats. A practiced laboratory reads 25-30 residues from one application; longer peptides require enzymatic fragmentation first ([AltaBioscience method overview](https://altabioscience.com/protein-sequencing/edman-degradation-vs-mass-spectrometry/)).

Edman has two non-negotiable prerequisites: 1. Free N-terminus. Acetylation, formylation, pyroglutamate formation, and other N-terminal modifications block the reagent. About 60-70% of native eukaryotic proteins are N-terminally blocked; synthetic peptides as released by Lyochem are usually free, but a customer using a modified peptide must specify. 2. No proline at the residue immediately following the labelled one. Pro disrupts the chemistry of the cleavage step; the cycle stalls or misreads.

Where Edman wins: unambiguous residue-by-residue assignment with chemically distinct reference standards. There is no isobaric ambiguity because each residue is identified by its phenylthiohydantoin derivative on a separate HPLC trace, not by mass. Where Edman loses to MS/MS: throughput (one peptide per several-hour run vs minutes per peptide), sensitivity (Edman needs ~10 pmol; MS/MS reads femtomoles), and any N-terminally modified target.

What USP and ICH expect

The USP General Chapter ⟨1055⟩ Biotechnology-Derived Articles — Peptide Mapping became official 1 December 2024. The chapter defines peptide mapping — a digestion-plus-LC-plus-MS workflow — as the chemical identification test for peptide and protein identity. The December 2024 revision tightened the section on development of identity tests, pretreatment, digestion conditions, separation, and specificity, with explicit guidance on when high-resolution MS is required vs when standard quadrupole resolution is acceptable ([USP Peptide Mapping page](https://www.usp.org/harmonization-standards/pdg/biotechnology/peptide-mapping)).

ICH Q6A and Q6B treat peptide identity as a release-test requirement; both expect a sequence-confirming method beyond mass-match for any peptide entering a regulated workflow (clinical, GMP API, biotech-product release). For pure bench research the bar is lower — a single ESI-MS pass plus HPLC purity is the practical norm — but the bar has moved as regulators have aligned around peptide-mapping methodology.

When to ask for the add-on

Sequence-level verification matters most when the answer to one of these is yes:

• First lot of a new sequence into your lab. Establish identity once with a sequence method; subsequent lots can ride on mass-match plus retention-time identity against the qualified first lot.
• A bioactivity readout is sensitive to single-residue substitutions. Receptor-binding studies, structure-activity relationships, residue-scan SAR — anything where the conclusion depends on one named residue being the labelled one.
• The peptide is going into an animal or human study. Regulatory expectation pushes the identity bar up regardless of who pays for the work.
• Cross-lab reproducibility is critical. A method that fails to reproduce on a second supplier's material is sometimes a real-biology negative; it is more often a sequence-level supplier difference that mass alone missed.
• The sequence contains a known isobaric pair (SA/GT, KQ/AN, DG/EA, CQ/GM) or an L/I pair in a functionally important position. Don't trust mass alone there.

For routine in-vitro cell-line work where one residue is unlikely to matter and the peptide is well-established in the literature, mass match plus HPLC purity is usually proportionate. The cost of an unnecessary Edman or LC-MS/MS sequence run is mostly the customer's analytical budget; the cost of an unnecessary mass-only release of a peptide that turned out to be a sequence variant is months of confused data.

What Lyochem ships by default and on request

Every Lyochem reference-grade lot ships with: - RP-HPLC trace at 214 nm + integrated main-peak purity (release spec ≥ 98.5%) - ESI-MS observed monoisotopic mass + Δ to theoretical (target ≤ 0.5 Da on standard instrument, ≤ 0.05 Da on the high-resolution rig) - Amino acid analysis (AAA) confirming hydrolysed-residue composition against theoretical (the third leg that catches gross errors mass alone can miss)

Available on request and noted on the COA when run: - LC-MS/MS sequence confirmation with b/y ion ladder and per-residue confidence (recommended for first lot of any new sequence) - Edman degradation for N-terminal sequence verification up to 25 residues (recommended for SAR work and any release where a downstream regulator will ask) - Chiral analysis for D-/L-amino-acid composition (rare but specifiable)

If your lot report says "ESI-MS confirmed" and nothing more, you have a mass match. If your study needs a sequence, request the sequence test, name it on the PO, and the per-lot data shipped with the next batch will say so explicitly.