Endotoxin testing for research peptides: LAL, recombinant Factor C, and what a COA actually certifies
Endotoxin is what remains of the outer membrane of Gram-negative bacteria after the cell is dead: lipopolysaccharide, resistant to autoclaving, active at picogram concentrations, and capable of confounding a wide range of in vitro and in vivo research through cytokine induction, complement activation, and fever-response modeling in animal studies. A lyophilized peptide produced by solid-phase synthesis is not intrinsically a high-endotoxin material — the synthesis chemistry does not involve living cells — but the finished vial passes through water systems, reconstitution buffers, stopper materials and container closures at which endotoxin can be introduced. Whether the vial that reaches a laboratory carries a quantifiable endotoxin burden depends on process controls the researcher cannot see, and on whether the lot was released against a certified endotoxin limit.
This article walks through the biology the endotoxin assays exploit, the two pharmacopeial families in current use — the traditional Limulus amebocyte lysate (LAL) assay and its recombinant successors — and what a research-grade certificate of analysis does and does not certify about the endotoxin content of a peptide vial. Our own certificates (our TB-500 10 mg COA is used as the worked example) are HPLC-UV purity reports; the article closes on what that scope means and what a researcher whose assays are endotoxin-sensitive should ask separately.
The biology behind the assay
The Limulus amebocyte lysate test rests on an observation Jack Levin and Frederik Bang made in the early 1960s: horseshoe crab (Limulus polyphemus) hemolymph clots in the presence of Gram-negative endotoxin, and the clotting reaction can be reproduced in vitro using the lysate of the crab’s amebocytes. The reaction is a serine-protease cascade — Factor C recognizes the endotoxin, activates Factor B, which activates a proclotting enzyme, which cleaves coagulogen into a gel. The cascade amplifies, so picogram quantities of endotoxin drive a detectable end point.
The test came into routine pharmaceutical use through the 1970s and was codified in USP General Chapter <85> Bacterial Endotoxins Test in 1980. Three read-out formats developed:
- Gel-clot — the qualitative or limit test; a positive reaction is judged by inversion of the reaction tube and gel formation.
- Turbidimetric — quantitative; the developing turbidity of the coagulogen gel is measured spectrophotometrically over time.
- Chromogenic (kinetic) — quantitative; a synthetic chromogenic substrate replaces coagulogen, and the released chromophore is read at 405 nm.
All three read Factor C’s response to endotoxin. All three are calibrated against endotoxin reference standard preparations calibrated in Endotoxin Units (EU) — an activity unit rather than a mass unit, reflecting that different LPS structures show different potencies in the assay.
The pharmacopeial framework: <85>, and the arithmetic
USP <85> requires that every parenteral product carries an endotoxin specification below a defined limit. The limit is calculated as K/M, where K is the threshold pyrogenic dose per kilogram per hour (5 EU/kg/hr for non-intrathecal parenteral administration; 0.2 EU/kg/hr for intrathecal) and M is the maximum dose of the drug substance per kilogram per hour. The limit is inversely proportional to dose — a small-dose material has a higher permitted EU/mg than a large-dose one.
Compendial acceptance is by validated assay against endotoxin reference standard, with sample interference addressed by spike-recovery within 50–200%. A material that inhibits or enhances the assay (through chelation, protease activity, or matrix effects) must be diluted to the point at which recovery is quantitative; the maximum valid dilution (MVD) defines how much dilution can be tolerated before the limit becomes unmeasurable at the endotoxin threshold.
For a research peptide, three practical points follow. First, the endotoxin limit that would apply to a hypothetical parenteral use depends on the intended dose, which is why a certified EU/mg is more informative on a research COA than a compliance-with-a-single-threshold statement. Second, interference varies by peptide — chelating sequences, cysteine-containing peptides, and any material carrying trace divalent cations can shift the assay — so an endotoxin specification without an interference validation is weak evidence. Third, LAL is not a sterility test. A material can be endotoxin-negative and still carry viable Gram-positive contamination, and endotoxin can persist in an autoclaved solution long after the source bacteria are dead.
The recombinant successors: rFC, rCR, and USP <86>
The horseshoe crab supply chain has been under pressure for two decades. Ding, Ho and collaborators cloned Factor C from Carcinoscorpius rotundicauda in the late 1990s and demonstrated that recombinant Factor C, expressed heterologously, binds endotoxin and can be used as the sole assay reagent (Ding JL, Chai C, Pui AWM, Ho B, J Endotoxin Res 4(1):33-43, 1997). Removing the downstream cascade removes a specific interference: Factor G, which is activated by 1,3-β-D-glucans (from cellulose or fungal sources) and generates false positives in the LAL cascade. A recombinant-Factor-C-only assay reads endotoxin without that glucan cross-reaction.
Two independent product comparability studies, both published in the PDA Journal of Pharmaceutical Science and Technology, established that rFC and LAL give equivalent endotoxin readings across a wide range of pharmaceutical matrices while diverging on interference. Bolden and Smith reported the Eli Lilly experience across a large validation portfolio (Bolden JS, Smith KR, PDA J Pharm Sci Technol 71(5):405-412, 2017) and Bolden’s later CASSS presentation summarized the outcome: approximately 50 products validated, 40-plus additional products qualified, over 92,000 samples tested, with the recombinant reagent performing at least equivalently to LAL. Marius, Vacher and Bonnevay ran a four-way comparison across four vaccine formulations with complex matrices — a live attenuated viral vaccine with proteases, an inactivated viral vaccine with minimal interference, an inactivated bacterial vaccine with natural endotoxin and glucans, and an inactivated colored viral vaccine — and reported that rFC results were “at least equivalent to those for the LAL-based assays,” with the interference pattern splitting predictably: proteases confounded rFC in one product, glucans confounded LAL in another (Marius M, Vacher F, Bonnevay T, PDA J Pharm Sci Technol, DOI 10.5731/pdajpst.2019.010389, 2020).
USP has codified the recombinant approach in a separate chapter. USP General Chapter <86> Bacterial Endotoxins Test Using Recombinant Reagents was published for early adoption in November 2024 and became official in May 2025, covering both recombinant Factor C (rFC) and recombinant cascade reagents (rCR). It sits alongside <85> rather than replacing it; a laboratory may use either, provided the assay is validated for the product.
What “endotoxin-tested” means on a certificate
A COA that carries an endotoxin row should carry three things to be independently interpretable: the assay method used (LAL gel-clot, turbidimetric, kinetic chromogenic, rFC, or rCR), a reported EU value with units (EU/mg is the useful reporting unit for a lyophilized peptide; EU/mL requires a reconstitution volume to convert), and evidence that the assay was validated on the specific product — spike-recovery in the 50–200% window at the working dilution. Any of the three missing means the row reports what the lab looked at, not what it demonstrated.
An additional consideration for peptides: many synthetic peptides contain residues (cysteine, histidine, aspartate-rich sequences) that can interact with the endotoxin assay’s cofactors, and some carry residual trifluoroacetate counterion from purification that can shift the assay’s calibration. A validated endotoxin specification on a peptide COA therefore represents more work than the equivalent specification on a simple small molecule — which is one reason most research-grade peptide certificates in this category do not carry one.
What our certificates carry — and what they do not
Our certificates are HPLC-UV purity reports with a stated content mass and a fill accuracy calculated against expected vial mass. For our TB-500 10 mg vial (report DBAV-TB500-10-051226, tested 23 May 2026 by Testides Analytical), the analysis table records expected 10.00 mg, content 9.84 mg, purity 98.62% by HPLC-UV, fill accuracy 98.4%. That is what the certificate covers. It does not carry an endotoxin row, a sterility row, or a bioburden row. It is a compositional certificate on the peptide content of the sampled vials, not a microbiological release certificate.
The distinction is important because the two questions are separate. HPLC-UV purity says: of the material that eluted from the column with UV-detectable absorbance, 98.62% was the target sequence. Endotoxin content is a lipopolysaccharide activity assay on a completely different scale, calibrated in EU rather than mass percent, run on a completely different platform. A high HPLC-UV purity number is not evidence of low endotoxin, and low endotoxin is not evidence of high peptide purity. Any research design where endotoxin contamination would confound the outcome — most immunology, most cell-culture work involving TLR4-competent cell lines, any in vivo pyrogenicity-sensitive rodent model — needs a separate endotoxin specification, and the honest answer for research-grade peptides is that if the certificate does not report one, the material was not released against one.
Practical implications
Three things follow.
First, for in vitro work where endotoxin contamination would confound the read-out — RAW 264.7 macrophages, THP-1 differentiation assays, any NF-κB reporter line — either request an endotoxin-tested lot from a supplier who runs the assay routinely on that SKU, or run the LAL/rFC assay in-house on the reconstituted stock before the experiment. A LAL gel-clot limit test at 0.25 EU/mL is a low-effort, low-cost first-pass check.
Second, for in vivo work in animal models, the endotoxin limit that applies scales with the intended administration route and dose. USP <85>’s K/M calculation is the framework; institutional animal-care committees may require a specific endotoxin certification separately from the peptide COA.
Third, treat “sterile” and “endotoxin-tested” as separate claims. Filter-sterilization of a reconstituted solution removes viable organisms but does not remove endotoxin already present in the powder. The two rows have to be certified — or performed — separately.
Bottom line
USP <85> and <86> define what an endotoxin test result must demonstrate and what units it must report in. The choice between LAL and recombinant Factor C is largely a supply-chain and interference-profile decision; multiple comparability studies show equivalent endotoxin detection with different failure modes on glucans and proteases. A COA that carries an endotoxin row with method, EU value and validation evidence is doing real work; a COA that does not carry one is silent on that question rather than reassuring on it. Research protocols that depend on low endotoxin should ask separately, and — for material that will be applied to endotoxin-sensitive cell or animal models — should verify in-house before use.
Materials described here are supplied for research use only. Not for human consumption.
References
- Levin J, Bang FB. The role of endotoxin in the extracellular coagulation of Limulus blood. Bull Johns Hopkins Hosp 115:265-274, 1964.
- Ding JL, Chai C, Pui AWM, Ho B. Expression of full length and deletion homologues of Carcinoscorpius rotundicauda Factor C in Saccharomyces cerevisiae: immunoreactivity and endotoxin binding. J Endotoxin Res 4(1):33-43, 1997. DOI 10.1177/096805199700400105.
- Bolden JS, Smith KR. Application of Recombinant Factor C Reagent for the Detection of Bacterial Endotoxins in Pharmaceutical Products. PDA J Pharm Sci Technol 71(5):405-412, 2017. Available at https://journal.pda.org/content/71/5/405.
- Marius M, Vacher F, Bonnevay T. Comparison of Limulus Amoebocyte Lysate and Recombinant Factor C Assays for Endotoxin Detection in Four Human Vaccines with Complex Matrices. PDA J Pharm Sci Technol, 2020. DOI 10.5731/pdajpst.2019.010389.
- U.S. Pharmacopeial Convention. General Chapter <85> Bacterial Endotoxins Test. USP-NF, current revision. Available at https://www.uspnf.com.
- U.S. Pharmacopeial Convention. General Chapter <86> Bacterial Endotoxins Test Using Recombinant Reagents. USP-NF, official May 2025. Announcement at https://www.usp.org/news/expert-committee-approves-endotoxin-testing-using-non-animal-derived-reagents.
Research use only. Not intended for human diagnostic or therapeutic application.