Capillary Electrophoresis and the Impurities a Purity Percentage Cannot See
Almost every purity figure printed on a research peptide certificate in the United States comes from one measurement, made on one axis. Reversed-phase HPLC with ultraviolet detection sorts molecules by how strongly they stick to a hydrophobic stationary phase, and the number reported is that one separation expressed as a percentage of total peak area. Our own certificates are explicit about it: the BPC-157 10mg batch certificate reads HPLC-UV at 214 nm, 99.41% purity, and carries no second analytical section at all.
That is the industry norm for US domestic suppliers, and for most questions it is the right method. But a single axis of separation has a predictable blind spot, and in peptide chemistry the blind spot is not obscure. It contains the three degradation products that a lyophilized peptide is most likely to actually contain.
Capillary electrophoresis is the method that separates on a different axis. This article is about what that second axis shows, what the compendia say about it, and what its absence from a certificate does and does not mean.
What a hydrophobicity number measures, and what it assumes
Reversed-phase separation resolves species that differ in how much non-polar surface they present to the column. It resolves truncated sequences well, because a missing residue changes hydrophobic surface. It resolves scavenger adducts and many synthesis by-products well, for the same reason. The area-percent calculation then assumes that everything present absorbs at the detection wavelength, and that everything present comes off the column as a resolved peak.
Both assumptions fail quietly. Material that co-elutes with the main peak is counted as main peak. Material that does not absorb at 214 nm is not counted at all. Neither failure produces an error message; it produces a slightly higher purity number.
The three impurities that live in the blind spot
Asparagine deamidation. An asparagine side chain loses its amide and becomes a carboxylate. The molecule gains one negative charge at neutral pH and gains one mass unit. Its hydrophobic surface barely moves. On a reversed-phase gradient the deamidated species frequently elutes as a shoulder on the parent, or under it.
Aspartate isomerization. The same chemistry that produces deamidation proceeds through a cyclic succinimide intermediate, and that intermediate can open two ways — back to normal alpha-linked aspartate, or to the beta-linked isoaspartate, in which the backbone runs through the side-chain carboxyl. Isoaspartate has the identical elemental composition and therefore the identical mass as aspartate. Mass spectrometry at the intact level cannot separate them. Neither can a hydrophobicity gradient, reliably.
Racemization. The succinimide intermediate racemizes rapidly, so the products include D-configured residues alongside the L. Bulk physical properties are close to unchanged.
Geiger and Clarke, working with a synthetic hexapeptide and reporting in the Journal of Biological Chemistry in January 1987 (PMID 3805008), put real numbers on how fast this happens. At 37 °C and pH 7.4 the peptide deamidated with a half-life of 1.4 days. The succinimide intermediate hydrolyzed with a half-time of 2.3 hours and racemized with a half-time of 19.5 hours. Their conclusion was that the net product of deamidation is a mixture of L- and D-forms of both the normal aspartyl and the isoaspartyl peptides, and that most of the racemization observed could be accounted for by the succinimide step.
Those are solution-phase numbers on a susceptible sequence, not a shelf life for a lyophilized powder in a sealed vial. The relevant point is narrower and more useful: these are not exotic degradation routes requiring harsh conditions. They are the ordinary chemistry of asparagine and aspartate in a peptide backbone, and they generate species that a hydrophobicity separation and an intact mass measurement are both, independently, poorly equipped to see.
What capillary electrophoresis does instead
In capillary electrophoresis a voltage is applied across a narrow fused-silica capillary filled with buffer, and species migrate according to their charge-to-hydrodynamic-size ratio, superimposed on the bulk electroosmotic flow of the buffer itself. Nothing partitions into a stationary phase. A change of one unit of charge — exactly what deamidation produces — is a first-order change in migration behavior rather than a marginal one.
The harmonized compendial text in the United States is USP General Chapter Biotechnology-Derived Articles — Capillary Electrophoresis, chapter <1053>. Reading the Revision Bulletin version official from 1 July 2009, the chapter sets out four modes: capillary zone electrophoresis, capillary gel electrophoresis, capillary isoelectric focusing, and micellar electrokinetic chromatography. It states that zone electrophoresis is applicable to molecules below 2,000 daltons and to the 2,000 to 100,000 range, which brackets essentially every research peptide sold in this category.
One operational detail in that chapter is worth pulling out, because it is a genuine difference from chromatography rather than a translation of it. In capillary electrophoresis, peak area is not proportional to amount as it stands. Species that migrate faster spend less time crossing the detection window and register smaller areas for the same quantity of material. The chapter therefore specifies that peak areas be divided by the corresponding migration time to give a corrected area, and that the percentage of a component be calculated from corrected areas as a fraction of the total corrected areas, excluding solvent and reagent peaks. A CE area-percent that has not been migration-time corrected is not the same quantity as an HPLC area-percent, and the two are not directly comparable.
The chapter’s system-suitability apparatus otherwise looks familiar: apparent theoretical plates as N = 5.54 (t_R / w_h)², resolution as 1.18 (t_R2 − t_R1) / (w_h1 + w_h2), a symmetry factor at 5% peak height, area and migration-time repeatability, and the conventional signal-to-noise thresholds of 3 for detection and 10 for quantitation.
What the published work shows
Two studies are worth naming because they bound the two questions a researcher would ask — is it sensitive enough, and does it actually resolve the species in question.
On sensitivity, Piešťanský and colleagues, reporting in Biomedicines in 2021 (9(10):1488, PMID 34680605), developed a capillary electrophoresis–tandem mass spectrometry method for the decapeptide triptorelin in both pharmaceutical and biological matrices. Using in-capillary field-enhanced sample stacking, they reported a detection limit of 5 ng/mL in a water matrix and 25 ng/mL in plasma, with the stacking step improving the detection limit fifty-fold over the conventional injection and a multisegment injection tripling throughput. Linearity was reported at r² ≥ 0.99, with intraday relative standard deviations of 1.5–9.4% and interday values of 2.3–11.9%.
On resolving power for the specific impurities described above, Gahoual, Beck, François and Leize-Wagner reported in the Journal of Mass Spectrometry in 2016 (DOI 10.1002/jms.3735, PMID 26889931) on the independent characterization of asparagine deamidation and aspartic acid isomerization by sheathless capillary zone electrophoresis coupled to tandem mass spectrometry. The word doing the work in that title is independent: the method distinguishes the two modifications from one another, which an intact mass measurement cannot do for the isomerization case at all.
Where capillary electrophoresis sits in US regulatory practice
It is a recognized compendial technique, harmonized internationally, and not a fringe one. ICH Q4B Annex 11 evaluated the pharmacopeial capillary electrophoresis texts and declared Ph. Eur. 2.2.47, JP General Information 4, and USP <1053> interchangeable across the ICH regions, reaching Step 4 in November 2009, Step 5 in September 2010, and coming into effect in December 2010. The annex records no acceptance criteria in the evaluated texts, and notes the FDA position that a company may be asked to demonstrate that a chosen method is suitable for a specific material or product regardless of where the method originated.
Two qualifications on that. USP <1053> is a chapter numbered above 1000, which in USP-NF convention makes it general information rather than a mandatory requirement. And the chapter has been through subsequent revision activity since the 2009 bulletin read here — anyone building a method to it should work from the current official text, not from this article.
What this means when reading a US supplier’s certificate
Nothing here says a certificate without a capillary electrophoresis section is deficient. Third-party HPLC-UV on a batch, published openly, is already well above what most of this market provides, and adding orthogonal methods costs money that ends up in the price of the vial. Domestic suppliers shipping from US facilities — ours go out of Santa Barbara, under a Wyoming entity — are competing on a certificate standard, and the honest framing is that a purity number is a measurement with a stated scope, not a verdict on a molecule.
What is fair to say is this. A certificate reporting HPLC-UV alone is reporting one axis of separation, and the reader should know which impurity classes that axis is weak on. Deamidation, isomerization and racemization are that list. If an experiment depends on charge-state homogeneity, on the absence of isoaspartate at a specific position, or on stereochemical integrity, the certificate as issued does not speak to it, and the honest response from a supplier is to say so rather than to imply the purity figure covers it.
Our certificates carry a sample-identification block and an HPLC-UV purity-and-content section. There is no capillary electrophoresis row on them. That is a scope statement, and it is one we would rather print than have a reader assume otherwise.
Where the method does not help
Capillary electrophoresis has real constraints. Injection volumes are nanoliter-scale, so concentration sensitivity with ultraviolet detection is modest and preconcentration steps such as the stacking used by Piešťanský are often necessary rather than optional. Migration-time reproducibility depends on the capillary wall condition, and peptides — basic ones especially — adsorb to bare silica, which is why the triptorelin method above ran at pH 1.88 specifically to suppress that adsorption. It is a complement to reversed-phase separation, not a replacement for it, and the compendial framing treats it that way.
All compounds discussed are supplied for research use only. Materials described here are laboratory reagents intended for in vitro and preclinical investigation by qualified researchers. They are not for human use, not for veterinary use, not foods, and not drugs, and nothing in this article describes or recommends administration to a person.
References
- Geiger T, Clarke S — succinimide-linked deamidation, isomerization and racemization at asparaginyl and aspartyl residues in peptides, with measured half-lives for the deamidation, hydrolysis and racemization steps. Journal of Biological Chemistry, 15 January 1987. PMID 3805008.
- Piešťanský J, Čižmárová I, Štefánik O, Matušková M, Horniaková A, Majerová P, Mikuš P — capillary electrophoresis–mass spectrometry with multisegment injection and in-capillary preconcentration for determination of the decapeptide triptorelin in pharmaceutical and biological matrices. Biomedicines, 2021; 9(10):1488. DOI 10.3390/biomedicines9101488. PMID 34680605.
- Gahoual R, Beck A, François YN, Leize-Wagner E — independent characterization of asparagine deamidation and aspartic acid isomerization by sheathless capillary zone electrophoresis coupled to tandem mass spectrometry. Journal of Mass Spectrometry, 2016. DOI 10.1002/jms.3735. PMID 26889931. (Volume and page numbers were not retrievable from an accessible host and are omitted rather than guessed.)
- United States Pharmacopeia — General Chapter <1053>, biotechnology-derived articles and capillary electrophoresis. Revision Bulletin official 1 July 2009. Modes, mobility and resolution equations, corrected-area normalization, and system-suitability criteria.
- ICH Q4B Annex 11 — evaluation and recommendation of pharmacopeial texts for use in the ICH regions, capillary electrophoresis general chapter. Step 4 November 2009; Step 5 September 2010; effective December 2010.