What endotoxin units mean on a peptide COA
A Certificate of Analysis reports endotoxin contamination as a number — for example, "<0.5 EU/mg" — without explaining what the unit actually means. This article unpacks the unit, the assay it comes from, and what numeric ranges look like in practice across research applications. It is the deeper reference for the endotoxin-result field on the COA.
What is an endotoxin unit?
An endotoxin unit (EU) is a defined-potency unit that ties an assay result to an FDA-recognized Reference Standard Endotoxin (RSE) — currently a lyophilized preparation of E. coli O113:H10 lipopolysaccharide. One EU corresponds to the activity of a defined mass of the reference standard in the LAL clotting cascade. The unit is what makes results comparable across labs, assay variants, and instruments: a 0.5 EU/mg result from a chromogenic LAL run in one lab and a 0.5 EU/mg result from a turbidimetric run in another reference the same biological activity.
EU is sometimes confused with IU (international unit), which is used for biological-activity calibration of compounds like hormones and vaccines. They are not interchangeable. Endotoxin is exclusively quantified in EU; biological potency for active compounds is quantified in IU. The two units share neither methodology nor reference standards.
How is EU/mg actually calculated?
EU/mg is the assay-detected endotoxin load divided by the mass of peptide tested. The lab dissolves a known mass of peptide in endotoxin-free water, runs the LAL or rFC assay against a calibration curve built with the reference standard, and reports the resulting EU value normalized to the peptide mass. The math is:
EU/mg = (EU detected in assay) / (mg of peptide in test sample)
A pass/fail entry on a COA — "endotoxin: <0.5 EU/mg" — means the lab spiked a known sample mass and confirmed the assay reading came in below the stated threshold. Numeric values ("0.18 EU/mg") carry more information than pass/fail entries because they let the researcher evaluate against application-specific tolerances. The methodology is codified in USP <85> for LAL and USP <86> for the rFC variant; both produce results in the same EU/mg unit.
What endotoxin levels are typical for research applications?
There is no single industry-wide acceptable EU/mg threshold for research peptides — the right number depends on what the material is being used for. A reagent destined for cell culture has a different endotoxin sensitivity than a compound for structural-binding work. The table below maps common research-application categories to the EU/mg ranges typically considered acceptable in the published literature; numeric specifics within each range vary by laboratory protocol.
| Research application | Typical EU/mg sensitivity | Why |
|---|---|---|
| Cell culture (general) | <1 EU/mg | Mammalian cells respond to endotoxin via TLR4 signaling; even sub-ng amounts can confound results |
| Sensitive cell culture (immune cells, primary lines) | <0.1 EU/mg | Macrophages, dendritic cells, and primary immune cells show responses at very low LPS concentrations |
| In-vitro biochemistry (binding, enzymatic assays) | <5 EU/mg | Most isolated-protein assays are not endotoxin-sensitive; the threshold is application-driven |
| In-vivo research (per FDA guidance for parenterals) | <5 EU/kg body weight (calculated) | Calculated dose-based threshold per FDA pyrogen guidance, not a per-mg threshold |
| Research-supplier batch release (general) | <1 EU/mg or batch-specific specification | Conservative default that supports most downstream applications without revalidation |
A peptide reported at "<0.5 EU/mg" by a reputable lab is suitable for most research workflows; a peptide reported at "<5 EU/mg" may be acceptable for binding work but require purification before sensitive cell-culture use. The deeper context for why endotoxin can pass through HPLC purity testing entirely undetected lives in the three-method release framework.
What does the LAL detection limit actually look like?
A LAL or rFC assay has a sensitivity range, not a binary detection capability. Most quantitative LAL formats (chromogenic, turbidimetric) reliably detect down to about 0.005 EU/ml in the test solution, which translates to ~0.05 EU/mg at typical peptide test concentrations. Below that, results are reported as "<X EU/mg" rather than as zero — the assay cannot resolve lower values, and reporting an artificially precise number would misrepresent the measurement.
Gel-clot LAL is the simplest format and is binary — pass or fail at a defined threshold. Chromogenic and turbidimetric formats produce continuous numeric readings within the assay's validated range. A reputable COA specifies which variant was used, which is what makes the precision of the reported number interpretable.
Frequently asked
- What does "<0.5 EU/mg" actually mean?
- It means the LAL or rFC assay detected an endotoxin level below 0.5 endotoxin units per milligram of peptide tested. The "<" indicates the result fell at or below the laboratory's reported precision boundary for that assay format and sample concentration. The true value may be much lower — the report only commits to "below the stated threshold."
- Is EU/mg the same as IU/mg?
- No. EU (endotoxin unit) is a potency unit specific to the LAL/rFC assay against an FDA-recognized reference endotoxin standard. IU (international unit) is a biological-activity unit used for compounds like hormones, enzymes, and vaccines. They share neither methodology nor reference standards and are not interchangeable.
- Can a 99.9% pure peptide still have high endotoxin?
- Yes. Endotoxin (lipopolysaccharide from Gram-negative bacterial cell walls) does not absorb at typical UV wavelengths used in HPLC peptide detection, so it is invisible to purity testing. A peptide can be 99.9% pure by HPLC and still carry endotoxin contamination from the synthesis water, glassware, or upstream reagents. This is why endotoxin testing is a separate release criterion.
- What is the difference between LAL and rFC endotoxin testing?
- LAL (Limulus Amebocyte Lysate) uses the blood-clotting cascade of horseshoe crabs and is codified in USP <85>. rFC (recombinant Factor C) uses a recombinant version of the active enzyme and is codified in USP <86>. The numeric output is in the same EU/mg unit and is interchangeable; rFC adoption is growing for animal-free supply-chain reasons.
- What endotoxin level should I look for on a research peptide COA?
- For most research workflows, <1 EU/mg is a reasonable conservative threshold. For sensitive cell-culture applications (immune cells, primary lines), look for <0.1 EU/mg. Reputable suppliers report numeric values rather than pass/fail, and the COA should specify the LAL variant used (gel-clot, chromogenic, turbidimetric, or rFC) so the precision of the result is interpretable.
Sources and further reading
- USP <85> Bacterial Endotoxins Test — methodology for the LAL assay across gel-clot, chromogenic, and turbidimetric variants.
- USP <86> Bacterial Endotoxins Test Using Recombinant Reagents — methodology for the rFC variant.
- FDA Guidance for Industry: Pyrogen and Endotoxins Testing — Questions and Answers — regulatory context on LAL methodology and acceptance criteria.