Written and maintained by CASRAI Editorial Board
Last updated
Most amino acids have no chromophore or fluorophore strong enough for sensitive HPLC detection on their own. Only tryptophan, tyrosine and phenylalanine absorb UV light appreciably, and even those are too weak for trace-level quantitation. Every practical amino acid analysis method therefore starts with a choice that has nothing to do with the column: how to attach a detectable tag to each amino acid, either before it goes on the column (pre-column derivatization) or after it comes off (post-column derivatization). That choice sets the sensitivity floor, decides whether proline gets measured at all, and determines how forgiving the method is of a slow autosampler queue.
This guide works through the three reagent chemistries actually used in practice — ninhydrin, OPA and AQC — what each does to detection limits and derivative stability, and how the acid hydrolysis step that precedes derivatization already limits which residues you can measure accurately regardless of which reagent you pick.
Pre-Column vs Post-Column: The Basic Trade-Off
Post-column derivatization adds the reagent to the column effluent after separation, via a dedicated reactor coil, on a system built for the purpose — the classic ion-exchange amino acid analyzer. Because derivatization happens after separation, reaction kinetics and derivative stability don’t affect resolution or peak shape; the chemistry only has to survive the few seconds between the reactor and the detector. The cost is hardware: you need a system with a post-column reactor and a second pump for reagent delivery, and separation happens on an ion-exchange column rather than a standard reversed-phase HPLC setup.
Pre-column derivatization reacts the reagent with amino acids in the sample vial before injection, then separates the derivatives by ordinary reversed-phase HPLC — see HPLC column selection: stationary phases & selectivity for the column-choice fundamentals that apply once you’re running derivatized amino acids on a standard C18. This runs on any conventional HPLC system, which is the main appeal, but now the derivatization reaction itself — completeness, side products, and how fast the derivative degrades before or during the run — is part of the method you have to control and validate.
Ninhydrin (Post-Column)
Ninhydrin reacts with primary amino acids to form Ruhemann’s purple, a strongly colored product read by visible-wavelength absorbance detection. Proline and other secondary amines (from an imino, not primary amine, nitrogen) give a different, yellow-orange product rather than the purple one — most ion-exchange analyzers run dual-wavelength detection specifically to capture both colors in one run rather than missing proline.
Ninhydrin’s sensitivity is modest by modern standards — practical quantitation generally needs sample amounts in the low-nanomole range per injection, well above what fluorescence-based pre-column methods need. What it has going for it is robustness and a long validation history: it’s the original amino acid analyzer chemistry, the post-column reaction is simple and doesn’t depend on getting a fast pre-injection reaction exactly right, and because the analyte never sits around as a derivative before separation, there’s no derivative-stability clock to manage. It remains the standard choice where the sample throughput doesn’t demand picomole sensitivity and the lab already runs a dedicated ion-exchange platform.
OPA (Pre-Column, Fluorescence)
OPA (ortho-phthalaldehyde), combined with a thiol reagent such as 2-mercaptoethanol, reacts with primary amino acids to form a fluorescent isoindole derivative. Fluorescence detection is inherently far more sensitive than UV/visible absorbance, and OPA derivatization typically pushes detection limits down into the low-picomole range — a substantial sensitivity gain over ninhydrin, which matters for limited sample volumes (a small tissue digest, a low-yield peptide hydrolysate) where you can’t just inject more.
OPA’s real limitation is chemical, not sensitivity-related: it only reacts with primary amines. Proline and hydroxyproline, both secondary amines, do not form a fluorescent OPA derivative and are effectively invisible to an OPA-only method — the same blind spot ninhydrin handles with a second detection wavelength, but here there’s no analogous workaround within OPA chemistry itself; labs that need proline alongside OPA typically add a separate oxidation or a secondary derivatization step, or switch reagents entirely. The second real limitation is derivative stability: OPA-isoindole derivatives are not chemically stable for long — fluorescence intensity decays within minutes of formation — so pre-column OPA derivatization has to happen right before injection, usually via automated, precisely timed inline derivatization on the autosampler rather than manual reagent addition to a batch of vials that then sit in a queue.
AQC (Pre-Column, Fluorescence)
AQC — 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate, marketed by Waters as the AccQ•Tag reagent — reacts with both primary and secondary amino acids, closing OPA’s proline gap without a second detection wavelength or a secondary derivatization step. The resulting derivatives are fluorescent and, unlike OPA’s, chemically stable: AQC-derivatized samples typically remain stable for days at room temperature, which means a full autosampler tray can be derivatized in a batch and queued for an overnight run without the timing precision an OPA method demands.
Sensitivity is comparable to or better than OPA — picomole-range detection limits are routine — and because AQC derivatives also absorb UV light, the same run can be detected by UV in addition to fluorescence, which is useful for labs without a dedicated fluorescence detector. The combination of full amino-acid coverage (including proline), derivative stability, and picomole sensitivity is why AQC-based kits have become a common default for labs replacing an aging ninhydrin ion-exchange analyzer, rather than replacing it with an OPA method.
What Acid Hydrolysis Already Decided Before Derivatization Starts
Derivatization chemistry only detects what survives sample preparation, and standard acid hydrolysis — 6 M HCl, 100-110°C, roughly 20-24 hours, used to break peptide bonds and release free amino acids for analysis — doesn’t survive all twenty intact:
- Tryptophan is destroyed outright by acid hydrolysis; its indole ring degrades under these conditions. No derivatization reagent recovers it after the fact — labs that need tryptophan use alkaline hydrolysis (e.g. with barium or lithium hydroxide) as a separate prep run specifically for that residue.
- Asparagine and glutamine are deamidated to aspartate and glutamate during acid hydrolysis, so standard acid-hydrolysis amino acid analysis cannot distinguish Asn from Asp or Gln from Glu — both are reported as combined Asx and Glx values, a hydrolysis limitation, not a derivatization one.
- Cysteine and methionine are vulnerable to oxidative loss and variable recovery during acid hydrolysis; labs needing accurate quantitation of these residues typically run a separate performic acid oxidation step before hydrolysis, which converts cysteine to cysteic acid and methionine to methionine sulfone — stable, quantifiable oxidation products — rather than trying to protect the native residues through hydrolysis.
- Serine and threonine are partially degraded during extended acid hydrolysis, which is why many labs hydrolyze in triplicate at different time points and extrapolate back to time zero, or apply a standard correction factor, rather than trusting a single hydrolysis run at face value.
None of this is a reason to prefer one derivatization reagent over another — ninhydrin, OPA and AQC all detect whatever survives hydrolysis equally poorly for these residues. It’s the reason “which derivatization reagent” and “which hydrolysis protocol” are two separate method-development decisions, not one: reagent choice controls sensitivity and which amines get tagged at all (the proline question); hydrolysis protocol controls which residues are still chemically intact by the time any reagent gets to them.
Choosing a Method
| Situation | Reasonable choice |
|---|---|
| Already running a dedicated ion-exchange amino acid analyzer; nanomole sensitivity is adequate | Ninhydrin (post-column) — simple, well-validated, no derivative-stability clock |
| Need picomole sensitivity, proline isn’t a required residue, and derivatization can be automated inline right before each injection | OPA (pre-column) — highest sensitivity per reagent cost, but proline-blind and unstable derivatives |
| Need full 20-amino-acid coverage including proline, want to batch-derivatize a full tray and queue an overnight run | AQC (pre-column) — stable derivatives, picomole sensitivity, UV or fluorescence detection |
| Tryptophan quantitation is required | Alkaline hydrolysis in a separate prep run, independent of derivatization reagent choice |
| Accurate cysteine/methionine quantitation is required | Performic acid oxidation before hydrolysis, independent of derivatization reagent choice |
For the general column and mobile-phase fundamentals that apply once derivatized amino acids are running on a reversed-phase system, see HPLC: columns, mobile phases, and a peak-problem troubleshooting table. Underivatized amino acids are also a textbook case of the retention problem HILIC exists to solve — very polar small molecules that barely retain on C18 — though in practice most labs solve it with derivatization rather than switching separation modes, precisely because derivatization also solves the detection problem HILIC doesn’t touch. For amino acid composition determined via mass spectrometry rather than a derivatization/fluorescence workflow, see LC-MS explained.
Frequently Asked Questions
Can I use OPA and still measure proline?
Not with OPA alone — it doesn’t react with secondary amines. Labs that need both OPA’s sensitivity and proline coverage typically add a second reagent step (commonly FMOC-Cl derivatization of whatever OPA leaves underivatized) rather than relying on OPA by itself. If proline is a required residue and you want to avoid a two-reagent workflow, AQC is the simpler single-reagent answer.
Why do some labs still use ninhydrin instead of a fluorescence method?
Mainly existing hardware and validation history. A dedicated ion-exchange amino acid analyzer with a ninhydrin post-column reactor is a substantial capital investment, and the chemistry is simple, robust and extremely well characterized. Where nanomole-range sensitivity is genuinely adequate for the sample type, there’s no analytical reason to replace a working ninhydrin system.
Does switching derivatization reagents fix tryptophan loss during hydrolysis?
No. Tryptophan is destroyed by standard acid hydrolysis before any derivatization reagent ever sees the sample — the fix is changing the hydrolysis protocol (alkaline hydrolysis), not the derivatization chemistry.
Are AQC derivatives detectable without a fluorescence detector?
Yes — AQC derivatives absorb UV light as well as fluorescing, so a system with only a UV/diode-array detector can still run an AQC method, just with less sensitivity than fluorescence detection would give.
