Quality and COAs

Amino acid analysis: the method that destroys the peptide to weigh it

Net peptide content is determined by hydrolysing the peptide back into free amino acids and counting them. The method is the quantitative reference, and it is also the reason the number carries more uncertainty than the purity figure printed beside it: acid hydrolysis destroys tryptophan outright, so quantitation is limited to 17 of the 20 amino acids.

Written for Analytical chemists, quality reviewers, and procurement · Updated August 13, 2026

Key points

Acid hydrolysis limits quantitation to 17 amino acids — tryptophan, asparagine and glutamine are lost
A NIST peptide reference material measured 94.0% pure by UV area and 69% peptide by mass
Content is calculated from seven well-recovered residues, not from all of them

Research-use boundary

This article is educational and limited to non-clinical research context, documentation practices, and terminology. It does not provide preparation, administration, dosing, treatment, diagnostic, cosmetic, or consumer-use guidance, and nothing here indicates that any material is suitable for human or animal use.

94% pure, 69% peptide — and NIST measured both

Reference Material 8327 is a peptide reference material issued by the National Institute of Standards and Technology. For its Peptide A, the report of investigation assigns a total UV absorbance area purity of 94.0% and a peptide mass purity of 69%. Peptide B: 86.5% and 73%. Peptide C: 97.1% and 67%. These are not competing estimates of the same quantity and neither number is wrong. The first says what fraction of the detectable material was the target sequence. The second says what fraction of the powder in the vial was peptide at all. Only one method answers the second question, and it works by destroying the peptide first.

The method in one line

Amino acid analysis hydrolyses the peptide back into its constituent free amino acids, separates those amino acids chromatographically, quantifies each one against calibrated standards, and converts the recovered quantities into a mass of peptide. It needs no reference standard of the peptide itself, which is precisely why it is the quantitative anchor: you cannot calibrate a peptide against a standard of that peptide whose content was never independently established. The composition it returns also corroborates identity, since a sequence predicts its own amino acid ratios.

The standard condition, and what it costs

United States Pharmacopeia General Chapter 1052 specifies hydrolysis in 6 N hydrochloric acid containing 0.1% to 1.0% phenol, at about 110 degrees for 24 hours, under vacuum below 200 mm of mercury or an argon headspace to limit oxidative destruction. Those conditions break every amide bond in the chain, which is the point. They also break things you wanted to count.

What the acid takes

Tryptophan is destroyed outright.
Serine and threonine are partially destroyed, and the loss grows with hydrolysis time.
Methionine can undergo oxidation.
Cysteine is typically recovered as cystine, and that recovery is usually poor because of partial destruction or reduction back to cysteine.
Asparagine and glutamine are deamidated to aspartic acid and glutamic acid, so they are reported as the combined values Asx and Glx.
The amide bonds of Ile-Ile, Val-Val, Ile-Val and Val-Ile are only partially cleaved in 24 hours, so those residues are under-counted rather than destroyed.

Seventeen, not twenty

USP 1052 states the consequence plainly: the loss of tryptophan, asparagine and glutamine during acid hydrolysis limits quantitation to 17 amino acids. The reference method for how much peptide is in a vial cannot see three of the twenty proteinogenic residues in its standard form. Tryptophan can be recovered by a modified hydrolysis using mercaptoethanesulfonic acid or thioglycolic acid as the reducing acid; cysteine can be measured by oxidising it to cysteic acid with performic acid, or by alkylation, before hydrolysis. Each of those is a separate determination with its own conditions, not something the standard run happens to include — and several of them modify other residues in the process, so a complete compositional analysis is not obtained from a single technique.

How the number is rescued: time-course hydrolysis

Two opposite errors are in play at once. Labile residues fall as hydrolysis proceeds; slow-cleaving residues rise. Running the analysis at 24, 48 and 72 hours separates them. Plot serine and threonine against hydrolysis time and extrapolate the line back to the origin, and you recover the starting concentration before the acid began consuming it. Plot isoleucine and valine and you look instead for a plateau, which is taken as the residue concentration. If the plateau starts to fall, hydrolysis has run long enough to begin destroying what it finished releasing. The cheaper alternative is to hydrolyse a free amino acid calibration standard alongside the sample, but USP notes the limitation: free amino acids do not degrade at the same rate as the same residues bound inside a chain, and that is especially true for the bonds that cleave slowly.

Content is calculated from seven residues, not twenty

Since some residues are recovered reliably and others are not, the content calculation uses only the dependable ones. USP 1052 lists the typically well-recovered set as aspartate-asparagine, glutamate-glutamine, alanine, leucine, phenylalanine, lysine and arginine. Divide the recovered quantity of each by the number of times that residue appears in the known sequence, and each well-recovered amino acid yields an independent estimate of the peptide content. Average them. Any estimate deviating from the mean by more than about 5% is discarded and the mean recalculated. A published validation of a reverse-phase amino acid analysis method used exactly that seven-residue set and reported an absolute error under 7% for each residue, under 2.5% on average, with recoveries of 97% to 108% per residue and 102% overall against bovine serum albumin.

Why the seven agreeing with each other is the real evidence

That discard rule is doing more work than it appears to. Seven independent estimates of the same quantity, derived from seven chemically different residues, either converge or they do not. When they converge, hydrolysis went to completion, the calibration held, and the sequence is what it was claimed to be. When one estimate sits far off the others, something specific is wrong — a co-eluting peak, a contaminated tube, an incorrect residue count in the assumed sequence. A single number reported without that internal spread has thrown away the diagnostic and kept only the result.

The contamination that ruins it is ordinary dust

Glycine and serine are common laboratory contaminants, which is a problem for a method that identifies material by finding glycine and serine. USP 1052 is specific about the source: glove powders and fingerprints on hydrolysis tubes cause erroneous results. The remedies are correspondingly blunt — boil the glassware for an hour in 1 N hydrochloric acid, or pyrolyse it at 500 degrees for four hours. Keratin from skin is protein; hydrolyse it alongside the sample and its amino acids are counted as the sample's.

The number is not portable between laboratories

In a published inter-laboratory comparison of amino acid analysis, between-laboratory coefficients of variation reached 23.7% for tryptophan, 17.6% for cystine and 16.1% for methionine, against up to 10% for most essential amino acids — while within-laboratory variation stayed below 5%. The residues that vary most between laboratories are exactly the ones hydrolysis handles worst, which is what you would predict if the variation lives in the sample preparation rather than the chromatography. The practical reading: a content figure is far more comparable to another figure from the same laboratory and method than to one from a different laboratory, and lot-to-lot comparisons should hold the laboratory constant where possible.

The uncertainty is real, and NIST published it

Return to Reference Material 8327. Its UV area purities carry expanded uncertainties of roughly 3.6 to 12.6 percentage points across five laboratories. Its peptide mass purities carry 11 to 16 percentage points across three. Those are the intervals a national metrology institute reported for a purpose-built reference material with participating expert laboratories. The honest conclusion is not that the method is unreliable — it is the best available answer to the question and there is no substitute — but that peptide content is intrinsically a harder measurement than peptide purity, and a content figure quoted to more precision than that deserves a question. Note also that the reference values in that archived report expired in September 2015; a reference material has a shelf life like anything else.

Why weighing the powder is not an alternative

The obvious shortcut is to weigh the solid and call that the peptide. It is not, because counter-ion salt, residual water and residual solvent are weighed along with it. Lyophilised peptides are hygroscopic and take up atmospheric moisture on the balance, so the error runs in one direction: weight-based figures overstate content. An evaluation of standard peptides used in quantitative proteomics found that approximately 30% of peptide concentrations established by weight were not accurate. That is a systematic error, not noise — it does not average out across replicates, because every replicate prepared from the same powder inherits the same offset.

One design detail worth noticing

The three peptides in Reference Material 8327 are DAEPDILELATGYR, KAQYARSVLLEKDAEPDILELATGYR and RQAKVLLYSGR. None of them contains methionine, cysteine or tryptophan. NIST states the sequences were designed for long-term stability, a range of purities and a range of molecular masses. Whatever the reasoning, the effect is that the reference material against which peptide measurements are evaluated contains none of the three residues that acid hydrolysis handles worst. That is a reasonable choice for a stable standard, and it is worth remembering when a real sequence does contain them.

What to ask about a content figure

Was content determined by amino acid analysis, by elemental analysis, or estimated from the sequence and assumed counter-ion stoichiometry?
Was a hydrolysis time course run, or a single 24-hour hydrolysis?
Which residues were used in the calculation, and did any get discarded?
Was water content determined separately, by Karl Fischer rather than by difference?
Does the reported mass on the vial refer to gross powder weight or to net peptide weight?

The point of the exercise

Purity by chromatography and content by amino acid analysis are orthogonal measurements: one uses a detector to compare peaks against each other, the other destroys the molecule and counts the pieces against external standards. They fail in unrelated ways, which is exactly why agreement between them is evidence and either alone is not. A certificate reporting only purity has answered a real question honestly. It has not answered how much peptide is in the vial, and the method that does answer it has to take the peptide apart to do so.

Common questions

What is amino acid analysis used for?

Amino acid analysis determines how much peptide is present in a sample and what its amino acid composition is. The material is hydrolysed back into its constituent free amino acids, those amino acids are separated and quantified against calibrated standards, and the recovered quantities are converted into a mass of peptide. It answers a different question from chromatographic purity, which describes the proportion of the peptide-related material that is the target sequence rather than how much of the sample is peptide at all.

Why is tryptophan destroyed during acid hydrolysis?

The standard hydrolysis condition is 6 N hydrochloric acid at about 110 degrees Celsius for 24 hours, and tryptophan does not survive it. United States Pharmacopeia General Chapter 1052 states that tryptophan is destroyed, that serine and threonine are partially destroyed, that methionine can be oxidised, and that asparagine and glutamine are deamidated to aspartic acid and glutamic acid. The combined loss of tryptophan, asparagine and glutamine limits quantitation to 17 amino acids. Modified hydrolysis procedures using mercaptoethanesulfonic acid or thioglycolic acid can recover tryptophan, but they are separate determinations rather than part of the standard run.

Which amino acids are considered well recovered in amino acid analysis?

United States Pharmacopeia General Chapter 1052 lists the typically well-recovered amino acids as aspartate-asparagine, glutamate-glutamine, alanine, leucine, phenylalanine, lysine and arginine. Peptide content is calculated from these residues because they survive hydrolysis reliably and are not common laboratory contaminants. Content values derived from individual well-recovered amino acids that deviate from the mean by more than about 5 percent are discarded and the mean recalculated.

Why do peptide content figures vary between laboratories?

Because the number is the output of a destructive chemical procedure rather than a direct reading. Hydrolysis time, vacuum quality, glassware cleanliness and calibration all shift recovery, and the residues most affected are the ones that survive worst. In a published inter-laboratory comparison, between-laboratory coefficients of variation reached 23.7 percent for tryptophan, 17.6 percent for cystine and 16.1 percent for methionine, against up to 10 percent for most essential amino acids, while within-laboratory variation stayed below 5 percent. A content figure is therefore most comparable when it comes from the same laboratory and method as the figure it is being compared with.

Can peptide content be determined by weighing the powder?

Weighing gives the mass of the whole solid, which includes counter-ion salt, residual water and residual solvent alongside the peptide. Lyophilised peptides are hygroscopic and take up atmospheric moisture readily, so a weight-based figure tends to overstate how much peptide is present. In a published evaluation of standard peptides used for quantitative proteomics, approximately 30 percent of peptide concentrations were found not to be accurate when established by weight.

Sources

General Chapter 1052, Biotechnology-Derived Articles — Amino Acid Analysis (Stage 6 harmonisation text, official 1 May 2018) United States Pharmacopeia Compendial source for the hydrolysis conditions, the residues destroyed or partially cleaved, the 17-amino-acid quantitation limit, the time-course extrapolation, the well-recovered residue list and the 5% discard rule.
Report of Investigation, Reference Material 8327: Peptide Reference Material for Molecular Mass and Purity Measurements National Institute of Standards and Technology (archived report, reference values expired 30 September 2015) Assigns percent peptide mass purity from amino acid analysis and UV absorbance; reports 69%, 73% and 67% peptide by mass against UV area purities of 94.0%, 86.5% and 97.1% for the same three peptides.
Validation of a reversed-phase HPLC method for quantitative amino acid analysis Journal of Biomolecular Techniques 2006;17(2):131-137 (ABRF) Validated method restricted to seven best-recovered amino acids; absolute percent error under 7% per residue and under 2.5% on average, recovery 97-108% per residue and 102% overall.
Accurate quantitation of standard peptides used for quantitative proteomics Proteomics 2009 (PubMed 19637239) Reports that approximately 30% of peptide concentrations were not accurate when based on weight quantitation, and attributes the error to hygroscopicity and salt content.
Comparison of interlaboratory variation in amino acid analysis and rat growth assays for evaluating protein quality Journal of the Association of Official Analytical Chemists 1985;68(1):52-56 (PubMed 4038977) Cited only for its amino acid analysis variability figures: between-laboratory coefficients of variation up to 23.7% for tryptophan, 17.6% for cystine and 16.1% for methionine, against within-laboratory variation up to 4.7%.