Mass spectrometry identity confirmation answers whether a peptide is built from the right amino acids. Peptide mapping answers whether those amino acids sit in the right order. Neither answers a third question that regulators evaluate separately: once the chain is correct, does it fold, or associate, into the right shape? That is higher-order structure (HOS), and on July 28, 2026, FDA gave it new prominence in how it evaluates peptide drug products.
What FDA changed
FDA published 17 revised draft product-specific guidances covering generic versions of peptide products including semaglutide, liraglutide, tirzepatide, teriparatide, calcitonin salmon, pegcetacoplan, dasiglucagon, and vosoritide, and withdrew its May 2021 guidance for ANDAs covering highly purified synthetic peptide drug products, which the agency said no longer reflects its current scientific thinking. Alongside submission pathways for recombinantly, synthetically, or semi-synthetically produced peptides, the revised guidances set updated expectations across four further technical areas: innate immune response testing, impurity thresholds, higher order structure assessment, and biological activity assessment. Placing HOS assessment next to impurity thresholds and bioactivity, rather than folding it into a single identity line, signals that regulators now treat shape as its own checkable attribute for this class of molecule, not a byproduct of getting the sequence and mass right.
What higher-order structure actually means
The working definition comes from ICH Q5E, Comparability of Biotechnological/Biological Products Subject to Changes in Their Manufacturing Process, which instructs manufacturers to determine that "higher order structure (secondary, tertiary, and quaternary structure) is maintained in the product" whenever a manufacturing process changes. The same guideline adds a fallback: "If the appropriate higher order structural information cannot be obtained, a relevant biological activity assay... could indicate a correct conformational structure." That sentence is worth sitting with, because it treats a bioassay as a substitute for structural data only when structural data cannot be obtained directly, not as an equivalent, interchangeable check. It also notes elsewhere that "physicochemical and in vitro biological assays might not be able to detect all differences in structure and/or function," which is the same limitation this site has flagged for identity and purity testing individually: a passing result on one method answers a narrower question than it appears to.
Secondary structure is the local shape a stretch of chain adopts, most often an alpha-helix or a beta-sheet, versus no fixed shape at all (random coil). Tertiary structure is how the full chain folds relative to itself. Quaternary structure applies only to peptides or proteins that assemble into multi-copy complexes. None of the three is visible in a mass number or an HPLC retention time.
Why a correct sequence does not guarantee a correct shape
The clearest failure mode is disulfide pairing. A peptide built from two or more cysteine residues can form its intended disulfide bond, or it can form the wrong one, producing a positional isomer with the identical molecular formula and the same intact mass as the correctly folded peptide. Mass spectrometry identity confirmation, which matches observed to theoretical mass, cannot distinguish the two, because swapping which cysteine bonds to which does not change the total mass. Aggregation and misfolding present a related problem: a peptide can retain its correct monomeric mass while a meaningful fraction of the sample has folded incorrectly or clumped into higher-order aggregates, neither of which a purity percentage or an identity match is built to catch.
A 2021 methods paper from FDA's own Center for Drug Evaluation and Research, published in Molecules (Wang et al., PMID 34299526, DOI 10.3390/molecules26144251), frames the stakes directly: protein and peptide "folding, heterogeneity, dynamic exchange, oligomerization, and aggregation," collectively HOS properties, are "typically critical for efficacy and safety," and because HOS is stabilized by comparatively weak interactions and is sensitive to formulation, a generic or biosimilar product has to demonstrate similar HOS to its reference product, not merely an identical sequence. The paper, which tested marketed peptide and protein products spanning roughly 3 to 145 kilodaltons, used two-dimensional NMR to build quantitative similarity acceptance criteria (a Mahalanobis distance threshold and defined chemical-shift tolerances) precisely because sequence identity alone left that comparison unaddressed.
How HOS is actually tested
Circular dichroism (CD) is the standard, most accessible method. A 2012 review in the International Journal of Molecular Sciences (Gopal et al., PMID 22489150, DOI 10.3390/ijms13033229) calls CD "the most widespread technique used for estimating the secondary structures of proteins and polypeptides in solution," and lays out the characteristic spectral signatures: alpha-helical structure produces two negative bands near 208 and 222 nanometers with a positive band near 192, beta-sheet structure produces a negative band near 217 nanometers and a positive band near 195, and unordered, disordered chain produces a single band below 200 nanometers. The same review demonstrates that an identical peptide sequence can shift between these signatures depending on its environment, adopting little fixed structure in water and folding into an ordered helix in a membrane-mimetic environment. That context-dependence is exactly why HOS is evaluated as a formulation-linked, conformation-specific attribute rather than assumed from sequence alone.
NMR, the method behind the FDA paper above, goes further, resolving structural detail at near-atomic resolution rather than a broad secondary-structure class, at the cost of more instrument time and expertise than CD requires. Neither method is part of the routine HPLC-purity-plus-mass-spec-identity combination that most research-peptide Certificates of Analysis report.
What this does and does not mean for a COA
None of this means a COA silent on higher-order structure has skipped a step it should have run. For a short, linear research peptide without disulfide bonds, in the 5 to 20 amino acid range typical of substances like BPC-157 or TB-500, there is little fixed three-dimensional structure to characterize in the first place, so the absence of a CD or NMR result is not informative about anything. HOS assessment earns its place for a different, narrower category: peptides with disulfide bridges, cyclic or otherwise constrained backbones, and the larger, chemically modified analogs, the categories that make up most of the drug substances named in FDA's July 2026 guidance package. For that category, a report that stops at identity and purity has simply not tested for the specific defect class, incorrect folding, incorrect disulfide pairing, or aggregation, that those two tests are structurally unable to see. Whether the laboratory that ran the available tests names every party involved has no bearing on which tests were actually performed; a report that documents a conformation-sensitive method for a peptide where that method matters has disclosed more of the data the FDA's own comparability framework treats as necessary than one that has not, independent of anything about who is willing to be named on the document.
Further reading
The intact-mass identity check that HOS assessment sits beyond is covered in mass spectrometry for peptide identity verification. The sequence-level check that closes a related but distinct gap is covered in peptide mapping under USP <1055> and ICH Q6B. The wider set of quality attributes a synthetic peptide should be characterized against, beyond a single purity number, is covered in synthetic peptide quality attributes under USP <1503> and <1504>.
For the primary sources: FDA, Revised Draft Product-Specific Guidances for Certain Generic Peptide Products (July 28, 2026); ICH Q5E, Comparability of Biotechnological/Biological Products Subject to Changes in Their Manufacturing Process; Wang et al., "NMR Spectroscopy for Protein Higher Order Structure Similarity Assessment in Formulated Drug Products," Molecules (2021), PMID 34299526, DOI 10.3390/molecules26144251; and Gopal et al., "Applications of Circular Dichroism for Structural Analysis of Gelatin and Antimicrobial Peptides," International Journal of Molecular Sciences (2012), PMID 22489150, DOI 10.3390/ijms13033229.