A purity percentage and a peptide content figure sound like they should be the same number. They are not, and amino acid analysis is the assay that produces the second one. Where HPLC purity reports how much of the detected signal belongs to one dominant peak, amino acid analysis (AAA) breaks the peptide down to its constituent amino acids and weighs them, giving a mass-based answer to a question purity testing was never built to answer: how much actual peptide is in the vial.
What amino acid analysis is for
Amino acid analysis refers to the methodology used to determine the amino acid composition or content of proteins, peptides, and other pharmaceutical preparations. The compendial reference is USP General Chapter <1052>, Biotechnology-Derived Articles: Amino Acid Analysis, which is harmonized with the corresponding chapters in the European and Japanese pharmacopoeias (USP General Chapter <1052>, full chapter text). The chapter states four distinct uses: to quantify protein and peptides, to help determine identity based on amino acid composition, to support structure analysis, and to detect atypical amino acids that might be present.
That list is worth reading slowly, because it is where the content-versus-purity confusion usually starts. Our guide to what a purity percentage means draws the same line from the other side: purity is the target peptide measured against peptide-related impurities, while content, sometimes called net peptide content, is actual peptide mass measured against non-peptide material such as counterions, residual salts, and water. Amino acid analysis is one of the standard ways that second figure gets produced, alongside UV spectrophotometry.
Hydrolysis: the destructive first step
A peptide cannot be analyzed as amino acids until it is broken into them, and that means hydrolysis. USP <1052> describes acid hydrolysis as the most common approach: the sample is dried, combined with 6N hydrochloric acid containing a small amount of phenol (which prevents halogenation of tyrosine), sealed under vacuum or an inert atmosphere, and heated to about 110C for 24 hours.
Hydrolysis is not clean. The chapter is explicit that this step destroys or alters several residues: tryptophan is destroyed outright, serine and threonine are partially destroyed, cysteine is typically recovered as cystine with poor efficiency, methionine can oxidize, the peptide bonds joining isoleucine and valine residues (Ile-Ile, Val-Val, and related pairs) cleave slowly and incompletely, and asparagine and glutamine deamidate to aspartic acid and glutamic acid, which is why results are reported as combined Asx and Glx values rather than the original amide residues. Because acid hydrolysis loses tryptophan, asparagine, and glutamine as distinct residues, routine quantitation is limited to 17 amino acids, and a time-course hydrolysis (parallel runs at 24, 48, and 72 hours) is often used to extrapolate back to the true starting concentration of the labile residues.
Separating and detecting the amino acids
Once hydrolyzed, the liberated amino acids are separated chromatographically and made visible to a detector through derivatization, since free amino acids do not absorb UV light or fluoresce strongly enough on their own. USP <1052> groups the available methods into two families. Postcolumn methods separate free amino acids by ion-exchange chromatography first, then react the column effluent with a reagent such as ninhydrin (read at 570 nm, or 440 nm for the imino acid proline) or o-phthalaldehyde (read by fluorescence). Precolumn methods derivatize the amino acids before separation, using reagents including phenylisothiocyanate (PITC), 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate (AQC), or related fluorescent tags, followed by reversed-phase HPLC. Precolumn methods are generally more sensitive, requiring less sample, while postcolumn methods are less affected by run-to-run performance variation.
Turning peak areas into a content figure
The calculation that ties this back to a certificate of analysis is the part most readers never see. USP <1052> sets out how a laboratory converts amino acid peak data into an estimate of protein or peptide content. Not every amino acid is equally trustworthy for this purpose: because of the destruction and contamination described above, the chapter directs analysts toward a shortlist of well-recovered residues, typically aspartate/asparagine, glutamate/glutamine, alanine, leucine, phenylalanine, lysine, and arginine, for the actual quantitation.
For each of those residues, the measured quantity (in nmol) is divided by the expected number of residues per molecule to yield a content estimate, and the estimates from every well-recovered amino acid are averaged. Any individual result that deviates from that mean by more than roughly 5 percent is flagged and typically discarded before the mean is recalculated. The agreement between the measured composition and the expected composition, expressed as a relative compositional error, is also used to corroborate the identity and purity of the sample, not just its content. A peptide that hydrolyzes to the wrong amino acid ratios is not the peptide the label claims, regardless of what an HPLC purity trace shows.
How this fits with identity testing
Amino acid analysis is not only a content assay. ICH Q6B, the guideline on specifications for biotechnological and biological products adopted by FDA in 1999, lists amino acid composition alongside amino acid sequence, terminal amino acid analysis, and peptide mapping as part of structural characterization. The guideline states plainly that "amino acid composition analysis provides some useful structural information for peptides and small proteins" and that "quantitative amino acid analysis data can also be used to determine protein content in many cases" (ICH Q6B, FDA Guidance for Industry, August 1999). The same document notes that peptide fragments generated during peptide mapping are commonly identified using amino acid compositional analysis, N-terminal sequencing, or mass spectrometry together, which is a reminder that identity confirmation is rarely a single-method exercise.
Why it isn't interchangeable with purity or mass spec
None of this replaces the other assays on a complete certificate. HPLC purity reports how much of the detected chromatographic signal is the main peak versus other peptide-related species; it says nothing about how much non-peptide mass, salt, counterion, or water, is diluting the actual peptide content. Mass spectrometry confirms the molecular weight of that main peak, answering an identity question purity testing cannot. Amino acid analysis answers a third question: given everything in the vial, how much of it, by mass, is actually peptide. A peer-reviewed comparison of protein quantitation methods found that amino acid analysis remained the most reliable approach for accurate protein and peptide content determination in complex or chemically modified samples, where colorimetric assays such as Bradford and BCA could overestimate content by an order of magnitude and simpler UV-based methods broke down when the sample composition deviated from the calibration standard (Reinmuth-Selzle et al., "Determination of the protein content of complex samples by aromatic amino acid analysis, liquid chromatography-UV absorbance, and colorimetry," Analytical and Bioanalytical Chemistry, 2022, 414(15):4457-4470, PMC9142416).
What this means for a buyer
A certificate that reports only an HPLC purity percentage has not addressed net peptide content, and a certificate that reports only content has not confirmed identity. Amino acid analysis under USP <1052> is the compendial route to that content figure, built on a hydrolysis step that destroys some residues by design, a derivatization and separation step that makes the survivors visible, and a calculation that leans on the residues known to survive hydrolysis intact. Reading a COA critically means checking which of these questions, purity, identity, or content, each stated number is actually answering, since none of the three stands in for the others.